Include:
Excludes:
Outputs resulting form creative practice as research, including the following subtypes.
Artefacts, objects or craftworks, exhibited, commissioned or otherwise presented or offered for distribution or sale in the public domain, for example, visual arts, craft and cultural creations. Specific examples are: illustration, sculpture, media installations, ceramics, jewellery, metalwork, whakairo, taonga, raranga, or cultural artefacts such as large permanent public sculptures.
A collection of artworks displayed together can be entered as Exhibition/Curatorial Exercise.
A published/ publicly available score, first performance or first recording by a record label (on CD or DVD) of a musical composition.
Includes (but not limited to):
Excludes:
A creative research/problem-solving output in the form of design drawings, books, models, exhibitions, websites, installations or build works.
This can include (but is not limited) to:
A work of creative prose, poetry, dramatic text or a literary essay.
Includes (but not limited to):
A display of a researcher's objects/artworks in a public place (museum, art gallery or other public place) or curatorial work undertaken by an academic to form an exhibition (including catalogue). The objects may have historical, cultural or scientific importance, or alternatively possess aesthetic qualities or extraordinary characteristics.
Includes:
Film/Video
Research, creative or scholarly works in audio-visual form and likely to be first presented in a cinema, on television or online.
Includes (but not limited to):
1. introduction, 2. limitations of previous research, goals, and research questions, 3. the austrian science fund, 6. discussion, 7. conclusions.
Rüdiger Mutz, Lutz Bornmann, Hans-Dieter Daniel, Types of research output profiles: A multilevel latent class analysis of the Austrian Science Fund’s final project report data, Research Evaluation , Volume 22, Issue 2, June 2013, Pages 118–133, https://doi.org/10.1093/reseval/rvs038
Starting out from a broad concept of research output, this article looks at the question as to what research outputs can typically be expected from certain disciplines. Based on a secondary analysis of data from final project reports (ex post research evaluation) at the Austrian Science Fund (FWF), Austria’s central funding organization for basic research, the goals are (1) to find, across all scientific disciplines, types of funded research projects with similar research output profiles; and (2) to classify the scientific disciplines in homogeneous segments bottom-up according to the frequency distribution of these research output profiles. The data comprised 1,742 completed, FWF-funded research projects across 22 scientific disciplines. The multilevel latent class (LC) analysis produced four LCs or types of research output profiles: ‘Not Book’, ‘Book and Non-Reviewed Journal Article’, ‘Multiple Outputs’, and ‘Journal Article, Conference Contribution, and Career Development’. The class membership can be predicted by three covariates: project duration, requested grant sum, and project head’s age. In addition, five segments of disciplines can be distinguished: ‘Life Sciences and Medicine’, ‘Social Sciences/Arts and Humanities’, ‘Formal Sciences’, ‘Technical Sciences’, and ‘Physical Sciences’. In ‘Social Sciences/Arts and Humanities’ almost all projects are of the type ‘Book and Non-Reviewed Journal Article’, but, vice versa, not all projects of the ‘Book and Non-reviewed Journal Article’ type are in the ‘Social Sciences/Arts and Humanities’ segment. The research projects differ not only qualitatively in their output profile; they also differ quantitatively, so that projects can be ranked according to amount of output.
Research funding organizations have shown increasing interest in ex post research evaluation of the funded projects ( European Science Foundation 2011a ). For instance, the Austrian Science Fund (FWF), Austria’s central funding organization for the promotion of basic research and the subject of this article, has conducted ex post research evaluations for some years now ( Dinges 2005 ). By collecting and analysing information on the ‘progress, productivity, and quality’ ( European Science Foundation 2011b : 3) of funded projects, research funding organizations hope ‘to be able to identify gaps and opportunities, avoid duplication, encourage collaboration, and strengthen the case for research’ ( European Science Foundation 2011b : 3). As stated succinctly in the title of a 2011 working document by the European Science Foundation (ESF), a central topic in this connection is ‘The Capture and Analysis of Research Outputs’ ( European Science Foundation 2011a ). This involves the issues of what research outputs are actually important for ex post research evaluation, how they can be classified (typology) and how the data can be analysed. The ESF document provides the following definition of outputs: ‘Research outputs, as the products generated from research, include the means of evidencing, interpreting, and disseminating the findings of a research study’ ( European Science Foundation 2011a : 5).
But opinions differ on what research output categories should be included in ex post research evaluation. Without doubt, publication in a scientific journal is viewed in all scientific disciplines as the primary communication form ( European Commission 2010 ). For assessing the merits of a publication, bibliometric analyses are favoured. In the humanities and social sciences, however, the use of classical bibliometric analysis ( Glänzel 1996 ; Nederhof et al. 1989 ; Nederhof 2006 ; Van Leeuwen 2006 ) is viewed critically in the face of different forms of research outputs (e.g. monographs) and limitations of the databases ( Cronin and La Barre 2004 ; Hicks 2004 ; Archambault et al. 2006 ). For these disciplines, other forms of quantitative evaluation are under discussion ( Kousha and Thelwell 2009 ; White et al. 2009 ).
A number of authors have made a plea for extending classical biblio analysis and for broadening the concept of ‘research output’ generally ( Bourke and Butler 1996 ; Lewison 2003 ; Butler 2008 ; Huang and Chang 2008 ; Linmans 2010 ; Sarli et al. 2010 ): ‘A fair and just research evaluation should take into account the diversity of research output across disciplines and include all major forms of research publications’ ( Huang and Chang 2008 : 2018). Huang and Chang (2008) looked at an empirical analysis conducted of the publication types of all publications in the year 1998–9 across all disciplines at the University of Hong Kong and found that journal articles accounted for 90% and 99% of the total publications produced only in the disciplines medicine and physics. The other disciplines produced output in the form of very different types of written communication, such as books, book chapters, and conference and working papers. Huang and Chang’s (2008) comprehensive review of the literature on the characteristics of research output showed that especially in the humanities and social sciences, books, monographs, and book chapters are important forms of written communication.
The German Research Foundation (DFG), Germany’s central funding organization for basic research, carried out a survey in the year 2004 on the publishing strategies of researchers with regard to open access ( Deutsche Forschungsgemeinschaft 2005 ), and 1,083 DFG-funded researchers responded (response rate of 67.7%). When the researchers were asked to name their preferred form of traditional publication of their own work, they mentioned articles in scientific journals (on the average about 20 articles in 5 years). Life scientists published the largest number of journal articles (23.6 articles in 5 years) and humanities scholars and social scientists the fewest (12.7 articles in 5 years). Papers in proceedings were published far more often by engineering scholars than by researchers in other disciplines. Social scientists and humanities scholars had a greater preference for publishing their work in edited volumes and monographs than researchers in other disciplines. However, big differences in the numbers reported (e.g. number of books, number of journal articles) were found within disciplines. This study and the Huang and Chang study made it clear that not only the sciences and humanities differ greatly from other disciplines in their preferred form of written communication. There are great differences also within the natural sciences and humanities. The Expert Group on Assessment of University-Based Research set up by the European Commission came to similar conclusions ( European Commission 2010 : 26). In the opinion of the expert group, the peer-reviewed journal article is used as the primary form of written communication in all scientific disciplines. In addition, engineering scientists primarily publish in conference proceedings, whereas social scientists and humanists show a wide range of research outputs, with monographs and books as the most important forms of written communications.
The broadest concept of research output is used by the Research Council UK (RCUK) (see www.rcuk.ac.uk ), the United Kingdom’s (UK) central funding organization, and the Research Assessment Exercise (RAE) ( www.rae.ac.uk ), which in 2014 will be replaced by the new system, Research Excellence Framework (REF) (ww.ref.ac.uk). RAE and REF have the task of assessing the quality of research in higher education institutions in the UK. Whereas the RAE focuses on scientific impact, the performance measurement by the REF in addition includes societal impact—that is, any social, economic or cultural impact, or benefit beyond academia. As research output, the RAE and REF include different forms of research products (journal article, book, conference contribution, patent, software, Internet publication, and so on). The Research Outcome System (ROS) of RCUK distinguishes a total of nine categories of research outputs: publication, other research output, collaboration, communication, exploitation, recognition, staff development, further funding, and impact. The new REF is planned to extend the currently peer-supported RAE with a quantitative, indicator-based evaluation system that includes bibliometric and other quantitative methods. Butler and McAllister ( Butler and McAllister 2009 , 2011 ) spoke generally of a metric as opposed to peer review that would capture more than the classical bibliometric analysis based on journal articles does. RAE and REF are based on a research production model ( Bence and Oppenheim 2005 ) that differentiates between inputs (personnel, equipment, overheads), research generation processes, outputs (paper, articles, and so on), and utilization of research (scientific and societal impact). This kind of structuring in input, process, output, outcome/impact is also found in other frameworks for research evaluation, such as in the payback approach ( Buxton and Haney 1998 ; European Commission 2010 ; Banzi et al. 2011 ) and other national and international evaluation systems ( European Commission 2010 ).
Previous research on research outputs has had the following limitations:
As the databases for the empirical analysis, studies up to now used mainly literature databases ( Glänzel 1996 ; Nederhof et al. 1989 ) and (survey) data from researchers ( Deutsche Forschungsgemeinschaft 2005 ; Huang and Chang 2008 ). Therefore, the unit of analysis was people and not projects (European Science Foundation 2011). But the different research outputs and also inputs (e.g. human resources, funding) are tied with the research projects.
For the individual disciplines, the frequencies of certain research outputs were presented mostly in totals and separately without any closer examination of the combination of different research outputs in the form of a core profile. For example, some disciplines focus more on monographs and conference contributions and not so much on journal articles, whereas for other disciplines it is just the opposite. Beyond that, the variability of research output within a discipline, such as that found in a study conducted by the DFG ( Deutsche Forschungsgemeinschaft 2005 ), was hardly considered.
The studies often did not describe the research output comprehensively, as the RAE, REF, and RCUK do, for instance, and instead restricted the study to a specific research output category, such as journal articles. This can lead to an inadequate treatment of some disciplines. Technical sciences can be at a disadvantage, for instance, if patents are not included in the study. Moreover, mostly only selected disciplines were included in the analyses, such as social sciences and humanities, so that comparative analysis of various disciplines was not possible. But research projects in different disciplines can be very similar in the profiles of research output categories (abbreviated in the following as ‘research output profiles’).
The studies did not distinguish between quality and quantity of research outputs. For example, life sciences are similar to natural sciences in research output profiles, but life sciences have a higher volume of journal articles than the natural sciences do ( Deutsche Forschungsgemeinschaft 2005 ).
The goals of our study are:
Based on a secondary analysis of data in final project reports ( Glass 1976 ) at the FWF, Austria’s central funding organization for basic research, the goals of this study were (1) to find, across all scientific disciplines, types of funded research projects with similar research output profiles; and (2) to classify the scientific disciplines in homogeneous segments (e.g. humanities, natural sciences, engineering sciences) bottom-up according to the frequency distribution of these research output profiles. We aimed to establish the types of funded research projects using multilevel latent class analysis (MLLCA) ( Vermunt 2003 ; Kimberly and Muthén 2010 ; Mutz and Seeling 2010 ; Mutz and Daniel 2012 ).
The research questions are:
Are there any types of FWF-funded projects that have different core profiles of research outputs?
Do types of research output profiles vary across scientific disciplines? Can disciplines be clustered into segments according to the different proportions of certain types of research output profiles?
How does the probability of being in a particular type of research output profile depend on a set of project-related covariates (e.g. requested grant sum)?
Is there any additional variability within types of research output profiles that allows for a quantitative ranking of projects according to higher or lower research productivity?
The FWF is Austria’s central funding organization for the promotion of basic research. It is equally committed to all scientific disciplines. The body responsible for funding decisions at the FWF is the board of trustees, made up of 26 elected reporters and 26 alternates ( Bornmann 2012 ; Fischer and Reckling 2010 ; Mutz, Bornmann and Daniel 2012a , 2012b ; Sturn and Novak 2012 ). For each grant application, the FWF obtains at least two international expert reviews (ex ante evaluation). The number of reviewers depends on the amount of funding requested. The expert review consists (among other things) of an extensive written comment and a rating providing an overall numerical assessment of the application. At the FWF board’s decision meetings, the reporters present the written reviews and ratings of each grant application. In the period from 1999 to 2009 the approval rate of proposals was 44.2%. Since 2003, all funded projects are evaluated after completion ( Dinges 2005 ) (see www.fwf.ac.at/de/projects/evaluation-fwf.html ). The FWF surveys the FWF-funded researchers, asking them to report the outputs of their research projects using a category system that is akin to the research output system of RCUK. Additionally, referees are requested to provide a brief review giving their opinions on aspects of the final project report. They are also requested to assign a numerical rating to each aspect. The final reports were used for accountability purposes and to improve the quality of FWF’s decision procedure ( Dinges 2005 ).
The data for this study comprised 1,742 FWF-funded research projects called ‘Stand-Alone Projects’ across all fields of science (22 scientific disciplines classified into six research areas), which contributed to 60% of all FWF grants (‘Stand-Alone Projects’, ‘Special Research Programs’, ‘Awards and Prizes’, ‘Transnational Funding Activities’) and finished within a period of 9 years (2002–10). The labelling of the scientific disciplines and the research areas was adopted from the FWF ( Fischer and Reckling 2010 ). Each project head was requested to report the results of his or her research project by completing a form (final project report) containing several sections (summary for public relations; brief project report; information on project participants; attachments; collaboration with FWF).
Of the 1,742 completed FWF-funded research projects ( Table 1 ), most were in the natural sciences (31.6%), and the fewest were in the social sciences (6.0%) and technical sciences (4.5%). The finished projects (end of funding) were approved for funding in the period 1999–2010, one-third of them in 2003–4 alone. Due to still ongoing research projects, projects approved for funding in 2007–8 make up only 3.9% of the total database of 1,742 FWF-funded research projects. The average duration of the research projects was 39 months. In 84.5% of the projects, the project heads were men. The average age of the project heads was 47.
Sample description ( N = 1,742 completed FWF-funded research projects)
Variable . | . | Per cent . | . | SD . | Range . |
---|---|---|---|---|---|
Research area | |||||
Biosciences | 399 | 22.9 | |||
Humanities | 339 | 19.5 | |||
Human medicine | 269 | 15.4 | |||
Natural sciences | 551 | 31.6 | |||
Social sciences | 105 | 6.0 | |||
Technical sciences | 79 | 4.5 | |||
Time period of the approval decision | |||||
1999–2000 | 210 | 12.1 | |||
2001–2 | 433 | 24.9 | |||
2003–4 | 582 | 33.4 | |||
2005–6 | 448 | 25.7 | |||
2007–8 | 69 | 3.9 | |||
Time period of the project end | |||||
2002–4 | 281 | 16.1 | |||
2005–6 | 531 | 30.5 | |||
2007–8 | 558 | 32.0 | |||
2009–10 | 372 | 21.4 | |||
Project duration [months] | 1,742 | 100.0 | 39.0 | 8.8 | 9→62 |
Overall rating of the proposal (ex ante evaluation) | 1,735 | 99.6 | 89.7 | 4.7 | 61.7→100 |
Requested grant sum [1,000 €] | 1,742 | 100.0 | 179.7 | 82.8 | 7.6→592.7 |
Project head’s sex | |||||
Man (=0) | 1,472 | 84.5 | |||
Woman (=1) | 270 | 15.5 | |||
Project head’s age | 1,739 | 99.8 | 47.1 | 9.8 | 27→87 |
Variable . | . | Per cent . | . | SD . | Range . |
---|---|---|---|---|---|
Research area | |||||
Biosciences | 399 | 22.9 | |||
Humanities | 339 | 19.5 | |||
Human medicine | 269 | 15.4 | |||
Natural sciences | 551 | 31.6 | |||
Social sciences | 105 | 6.0 | |||
Technical sciences | 79 | 4.5 | |||
Time period of the approval decision | |||||
1999–2000 | 210 | 12.1 | |||
2001–2 | 433 | 24.9 | |||
2003–4 | 582 | 33.4 | |||
2005–6 | 448 | 25.7 | |||
2007–8 | 69 | 3.9 | |||
Time period of the project end | |||||
2002–4 | 281 | 16.1 | |||
2005–6 | 531 | 30.5 | |||
2007–8 | 558 | 32.0 | |||
2009–10 | 372 | 21.4 | |||
Project duration [months] | 1,742 | 100.0 | 39.0 | 8.8 | 9→62 |
Overall rating of the proposal (ex ante evaluation) | 1,735 | 99.6 | 89.7 | 4.7 | 61.7→100 |
Requested grant sum [1,000 €] | 1,742 | 100.0 | 179.7 | 82.8 | 7.6→592.7 |
Project head’s sex | |||||
Man (=0) | 1,472 | 84.5 | |||
Woman (=1) | 270 | 15.5 | |||
Project head’s age | 1,739 | 99.8 | 47.1 | 9.8 | 27→87 |
Note : N = frequency, per cent = column per cent, M = mean, SD = standard deviation, range = minimum and maximum.
The following six research output categories were captured in quantity and number (count data) and served as the basis for the analysis: publication (peer-reviewed journal article; non-peer-reviewed journal article, monograph, anthology, mass communication, i.e. any kind of publication in mass media, e.g. newspaper article), conference contribution (invited paper, paper, poster), award, patent, career development (diploma/degree, PhD dissertation, habilitation thesis) follow-up project (FWF funded or not). It was not differentiated between different sub-categories of the mentioned research output categories. For example, hybrid, open access and standard peer-reviewed journal articles or ongoing or terminated PhD dissertations were summarized under the respective research output category. In order to avoid problems with different publication lags, the FWF treated equally manuscripts, already published, and manuscripts, accepted for publication. The ex post evaluation approach of the FWF does not distinguish between project publications written in English and written in any other language.
Because of strongly skewed distributions, the count variables were transformed in 2-point to 5-point ordinal scale variables with at most equally sized ordinal classes, to avoid sparse classes or cells in a multivariate statistical analysis. To draw up a typology, actually, binary variables might be sufficient in which it was coded whether the particular research output category (e.g. monograph) existed (= 1) for a research project or not (= 0). However, because we wanted to differentiate a qualitative dimension (types) and a quantitative dimension (amount of output), we chose an ordinal scale with a sparse number of ordinal classes that in addition allow a quantitative assessment.
The research output variables ( Table 2 ) show a large share of zeros. The most frequently produced types of publication were reviewed journal articles (an average of five per project) and conference papers (on average nine), with a large variance across the research projects. For publication of research results, monographs are used the least (0.2 monographs per project).
Data description ( N = 1,742 FWF-funded research projects)
Research output . | . | Ordinal categories | . | SD . | Max . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
. | Scale . | 0 . | 1 . | 2 . | 3 . | 4 . | . | . | . | . |
Journal article, reviewed | Number | 0 | 1–2 | 3–6 | >6 | 5.1 | 6.9 | 75 | 0.61 | |
Per cent | 23.7 | 22.8 | 26.4 | 27.1 | ||||||
Journal article, non-reviewed | Number | 0 | 1 | 2–4 | >4 | 2.8 | 5.6 | 50 | – | |
Per cent | 0.50 | 0.14 | 0.18 | 0.18 | ||||||
Contribution to anthologies | Number | 0 | 1 | >1 | 0.8 | 2.3 | 32 | 0.15 | ||
Per cent | 75.4 | 10.2 | 14.4 | |||||||
Monograph | Number | 0 | >0 | 0.2 | 0.7 | 8 | 0.15 | |||
Per cent | 89.4 | 10.6 | ||||||||
Mass communication | Number | 0 | 1 | >1 | 1.0 | 2.9 | 38 | 0.16 | ||
Per cent | 68.5 | 13.5 | 17.9 | |||||||
Award | Number | 0 | 1 | >1 | 0.5 | 1.2 | 13 | 0.28 | ||
Per cent | 74.0 | 13.5 | 12.5 | |||||||
Other output (patent, impact) | Number | 0 | 1 | >1 | 0.6 | 1.4 | 26 | 0.19 | ||
Per cent | 71.0 | 14.9 | 14.1 | |||||||
Conference paper | Number | 0 | 1 | 3–5 | 6–11 | >11 | 9.1 | 11.1 | 101 | 0.59 |
Per cent | 12.7 | 14.9 | 21.8 | 24.8 | 25.8 | |||||
Other conference contribution | Number | 0 | 1–2 | 3–6 | >6 | 4.7 | 7.5 | 98 | 0.51 | |
Per cent | 31.6 | 20.3 | 23.9 | 24.2 | ||||||
Habilitation thesis | Number | 0 | 1 | >1 | 0.6 | 0.9 | 7 | 0.12 | ||
Per cent | 60.7 | 25.8 | 13.5 | |||||||
PhD dissertation | Number | 0 | 1 | 2 | >2 | 1.1 | 1.4 | 23 | 0.30 | |
Per cent | 41.0 | 30.8 | 17.3 | 10.9 | ||||||
Diploma/degree | Number | 0 | 1 | 2 | >2 | 1.3 | 2.1 | 22 | – | |
Per cent | 53.4 | 17.2 | 10.8 | 18.6 | ||||||
Follow-up project | Number | 0 | 1 | >1 | 0.7 | 1.1 | 15 | 0.19 | ||
Per cent | 61.6 | 23.1 | 15.3 |
Research output . | . | Ordinal categories | . | SD . | Max . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
. | Scale . | 0 . | 1 . | 2 . | 3 . | 4 . | . | . | . | . |
Journal article, reviewed | Number | 0 | 1–2 | 3–6 | >6 | 5.1 | 6.9 | 75 | 0.61 | |
Per cent | 23.7 | 22.8 | 26.4 | 27.1 | ||||||
Journal article, non-reviewed | Number | 0 | 1 | 2–4 | >4 | 2.8 | 5.6 | 50 | – | |
Per cent | 0.50 | 0.14 | 0.18 | 0.18 | ||||||
Contribution to anthologies | Number | 0 | 1 | >1 | 0.8 | 2.3 | 32 | 0.15 | ||
Per cent | 75.4 | 10.2 | 14.4 | |||||||
Monograph | Number | 0 | >0 | 0.2 | 0.7 | 8 | 0.15 | |||
Per cent | 89.4 | 10.6 | ||||||||
Mass communication | Number | 0 | 1 | >1 | 1.0 | 2.9 | 38 | 0.16 | ||
Per cent | 68.5 | 13.5 | 17.9 | |||||||
Award | Number | 0 | 1 | >1 | 0.5 | 1.2 | 13 | 0.28 | ||
Per cent | 74.0 | 13.5 | 12.5 | |||||||
Other output (patent, impact) | Number | 0 | 1 | >1 | 0.6 | 1.4 | 26 | 0.19 | ||
Per cent | 71.0 | 14.9 | 14.1 | |||||||
Conference paper | Number | 0 | 1 | 3–5 | 6–11 | >11 | 9.1 | 11.1 | 101 | 0.59 |
Per cent | 12.7 | 14.9 | 21.8 | 24.8 | 25.8 | |||||
Other conference contribution | Number | 0 | 1–2 | 3–6 | >6 | 4.7 | 7.5 | 98 | 0.51 | |
Per cent | 31.6 | 20.3 | 23.9 | 24.2 | ||||||
Habilitation thesis | Number | 0 | 1 | >1 | 0.6 | 0.9 | 7 | 0.12 | ||
Per cent | 60.7 | 25.8 | 13.5 | |||||||
PhD dissertation | Number | 0 | 1 | 2 | >2 | 1.1 | 1.4 | 23 | 0.30 | |
Per cent | 41.0 | 30.8 | 17.3 | 10.9 | ||||||
Diploma/degree | Number | 0 | 1 | 2 | >2 | 1.3 | 2.1 | 22 | – | |
Per cent | 53.4 | 17.2 | 10.8 | 18.6 | ||||||
Follow-up project | Number | 0 | 1 | >1 | 0.7 | 1.1 | 15 | 0.19 | ||
Per cent | 61.6 | 23.1 | 15.3 |
Note : Per cent = row per cent, M = mean of the raw data, SD = standard deviation of the raw data, Max = maximum, R 2 indicates how well an indicator is explained by the final LC model.
In a review of the literature Gonzalez-Brambila and Velosos (2007) discuss age, sex, education, and cohort effects as empirically investigated determinants of research outputs. In our study, we included the following covariates to predict research profile type membership ( Table 1 ): time period of the approval decision, time period of the project end, project duration; overall rating of the proposal, requested grant sum; gender and age of the project head. This information was taken from an ex ante evaluation of the project proposals. In the ex ante evaluation, two to three reviewers rated each proposal on a scale from 1 to 100 (ascending from poor to excellent). The mean of the overall ratings of a proposal averaged across reviewers was 89.7 (minimum: 61.7, maximum: 100).
Latent Class Analysis (LCA) in its basic structure can be defined as a statistical procedure that extracts clusters of units (latent classes (LCs)) that are homogenous with respect to the observed nominal or ordinal scale variables ( McCutcheon 1987 ). Similar to factor analysis, LCs are extracted in such a way that the correlations between the observed variables should vanish completely within each LC (local stochastic independence). LCA is favoured towards cluster analysis due to the fact that fewer pre-decisions are required than in common cluster analysis procedures (e.g. similarity measure, aggregation algorithm). Efficient algorithms for parameter estimation (maximum likelihood) are used, and a broad range of different models (LCA, IRT models, multilevel models, and more) are offered ( Magidson and Vermunt 2004 ; Vermunt and Magidson 2005a ). In a more advanced version of LCA, MLLCA, the nested data structure is additionally considered. In our study, research projects are nested within certain scientific disciplines; LCs or project types might vary between scientific disciplines. In MLLCA, not only are projects grouped according to their output profiles but also scientific disciplines will be segmented according to their different proportions of types of output profiles. In the technical framework of MLLCA, LCs represent the types of research output profile, and latent clusters (GClass) indicate the segments of disciplines. It will be presumed that a project in a certain LC behaves the same way (same research output profile) irrespective of the latent cluster to which the project belongs.
In secondary analysis the problem frequently arises that the assumption of local stochastic independence does not fully hold. For instance, career development output categories like diploma/degree and PhD dissertation are more strongly correlated with one another than with the other research output categories, so that a LCA cannot completely clarify the association between the two career development outputs. There are three possible ways to handle this problem ( Magidson and Vermunt 2004 ): First, one or more direct effects can be added that account for the residual correlations between the observed research output variables that are responsible for the violation of the local stochastic independence assumption. Second, one or more variables that are responsible for high residual correlations can be eliminated. Third, the number of latent variables (LCs, continuous latent variables) is increased. In this study we used all three strategies. After a first model run, the residuals were inspected, and a few direct effects were included in the MLLCA model. Additionally, two variables that were responsible for high residual correlations were eliminated—non-peer-reviewed journal articles and diplomas/degrees. Last but not least a MLLCA model was tested that incorporates a continuous latent variable comparable to a factor analysis. With this C-factor not only can residual correlations among the output variables be explained but also additional quantitative differences between research projects (amount of research output) can be assessed and can be taken for a ranking of projects, respectively. If, over and above, a model fits the data with the same structure (i.e. loadings of the research output variables on the factor) for all LCs as well as or better than a model with different structures in terms of different loadings of the variables in each LC, all research projects can be compared or ranked on the same scale of the latent variable.
For statistical analysis of the data we used MLLCA as implemented in the software program Latent GOLD 4.5 ( Vermunt and Magidson 2005b ). Following Bijmolt, Paas, and Vermunt (2004) , Lukočienė, Varriale, and Vermunt (2010) and Rindskopf (2006) , in a first step we calculated a simple LCA of the research outputs to obtain types of research projects with a similar research output profile. To determine the optimal number of classes (project types, segments of disciplines), information criteria were used, such as the Bayesian information criterion (BIC) or Akaike information criterion (AIC). The lower BIC or AIC the better the model fits. These information criteria penalize models for complexity (number of parameters), making it possible to make direct comparisons among models of different numbers of parameters. Results of a simulation study conducted by Lukočienė and Vermunt (2010) for MLLCA models showed that in all simulation conditions, the more advanced criteria AIC3 ( Bozdagon 1993 ) and the BIC(k) outperformed the usual BIC to identify the true number of higher-level LCs (Lukočienė, Varriale and Vermunt 2010 ). Unlike BIC, BIC(k) uses the number of groups, here the number of disciplines, in the formula for sample size n : BIC(k) =−2 * LL – df * ln(k); AIC3 = −2 * LL −3 * df, where df denotes the degrees of freedom, LL denotes the loglikelihood. In the second step , we took the hierarchical structure of data into account, calculating an MLLCA to obtain latent clusters of scientific disciplines, or segments. In a third step we fixed the number of latent clusters of the second step and again determined the number of LCs. However, Lukočienė and Vermunt’s (2010) simulation study showed that the third step results in very small improvement of 1%. We therefore abstained from applying this step.
In the last step we included covariates in the model to explain the LC membership ( Vermunt 2010 ). However, this one-step procedure has the disadvantage that by including the covariates, the model and its parameters, respectively, could change. Therefore, a three-step procedure was suggested. First, we estimated a LC model. Second, we assigned the subjects to the LCs according to their highest posterior class membership probability. Third, the LCs were regressed on a set of covariates using a multinomial regression model. However, this procedure does not take into account the uncertainty of class membership. Bolck, Croon, and Hagenaars (2004) showed that such a modelling strategy underestimates the true relationships between LCs and covariates. Recently, Vermunt (2010) developed a procedure that takes into account the uncertainty of class membership by including the classification table that cross-tabulates modal and probabilistic class assignment ( Vermunt and Magidson (2005b) as weighting matrix into the multinomial regression model. We followed this improved three-step approach. The covariates mentioned above were included for prediction of class membership ( Table 1 ).
In the first step the nested data structure (projects are nested within scientific disciplines) was ignored, and simple LC models were explored. Table 3 shows the results of fitting the models containing one to 11 LCs with and without a continuous latent C-factor, respectively. For model comparison we used the AIC3. Out of all 22 models, Model 15 with four LCs, 107 parameters, and one C-factor shows the smallest AIC3. We therefore decided on this model. With regard to our research questions, there were four types of projects with different research output profiles (qualitative dimension). Additionally, the projects differed in their productivity, i.e. the amount of outputs, represented by the continuous latent C-factor (quantitative dimension).
Fit statistics for exploratory LC models (project types)
Note : MNR = model number, NCL = number of latent classes, LL = loglikelihood, NPAR = number of parameter, AIC3 = Akaike information criterion 3. Final model grey coloured.
Figure 1 shows the four LCs or project types with different research output profiles. The 2-point to 5-point ordinal scales were re-scaled such that the numerical values varied within the range of 0–1.0 ( Vermunt and Magidson 2005b : 117). We obtained this scaling by subtracting the lowest observed value from the class-specific mean and dividing the results by the range, where the range was nothing but the difference between highest and lowest value. The advantage of this scaling is that all variables can be depicted on the same scale as the class-specific probabilities for nominal variables. It must be noted that the LC results depicted in Fig. 1 were the results of the final MLLCA model (introduced in Section 5.2 ) and not the non-nested LC model in Table 3 . However, this does not matter, because the LC models with and without nesting do not differ.
LCs of research output profiles (* = not used in the MLLCA).
The four LCs or project types with different research output profiles can be described as follows (class sizes in per cent of the total number of projects in parentheses):
Latent Class 1 ‘ Not Book ’ (37.0%): The research output profile of this research project type is quite similar to the average profile across all projects but with fewer non-reviewed journal articles, anthologies, and monographs than the average.
Latent Class 2 ‘ Book and Non-Reviewed Journal Article ’ (35.8%): this project type uses anthologies and monographs but also non-reviewed journal articles and mass communication as primary forms of written communication. Career development—such as diploma/degree, PhD dissertation and habilitation thesis—reviewed journal articles and follow-up projects score quite below the average.
Latent Class 3 ‘ Multiple Outputs ’ (17.9%): This project type generates research outputs in multiple ways with above-average outputs as peer-reviewed journal articles, non-reviewed journal articles, anthologies, monographs, conference papers, habilitation theses, PhD dissertations, diplomas/degrees, follow-up projects, but with fewer other conference contributions.
Latent Class 4 ‘ Journal Article, Conference Contribution, and Career Development ’ (9.3%): this most productive project type focuses strongly on peer-reviewed journal articles, with many published papers in combination with conference contributions (papers or other products), career development (diploma/degree, PhD dissertation, habilitation thesis), and follow-up projects, but this type uses fewer monographs as a form of written communication.
Of all the output variables, peer-reviewed journal articles and conference contributions discriminate the best between the LCs, with a discrimination index of about 0.60 ( Table 2 , last column, R 2 ).
In a multilevel latent structure model it is presumed that there is variation among the 22 scientific disciplines in the unconditional probabilities (the probabilities belonging to each LC). In an MLLCA the 22 scientific disciplines are grouped into latent clusters or segments according to their different proportion of types of research output profiles, as obtained in Section 5.1 .
Table 4 shows the results of fitting models containing one to eight latent clusters (M 1 –M 8 ), each with four LCs and with one continuous latent C-factor, respectively. With respect to BIC(k) and AIC3, a 5-GClass model will be favoured, i.e. there are five different segments of scientific disciplines with different proportions of the project types or LCs. Additionally, using the option of ‘cluster-independent C-factor’, we tested (M 9 ) whether the same loading structure can be held in all four LCs. The BIC(k) and the AIC3 improved slightly from model M 5 to the more restricted model M 9 with 122 − 89 = 33 fewer parameters than M 5 . Therefore, the assumption of a cluster-independent C-factor held, which made it possible to compare and rank all projects on the same scale. Including direct effects, such as the association between habilitation thesis and PhD dissertation, further improved the model. Only one residual (res = 3.88) was somewhat larger than the criterion of 3.84 ( Magidson and Vermunt 2004 ). To fulfil the basic model assumption of local stochastic independence, we chose model M 10 as the final model.
Fit statistics of models for variation among scientific disciplines (GClass) with four LCs and one C-factor
MNR . | Models of disciplines . | LL . | NPAR . | BIC(k) . | AIC3 . |
---|---|---|---|---|---|
1 | 1 GClass | −17,789.4 | 106 | 35,906.4 | 35,896.8 |
2 | 2 GClass | −17,328.9 | 110 | 34,997.8 | 34,987.8 |
3 | 3 GClass | −17,211.1 | 114 | 34,774.6 | 34,764.2 |
4 | 4 GClass | −17,155.6 | 118 | 34,676.0 | 34,665.3 |
5 | 5 GClass | −17,139.7 | 122 | 34,656.4 | 34,645.3 |
6 | 6 GClass | −17,134.9 | 126 | 34,659.4 | 34,647.9 |
7 | 7 GClass | −17,133.4 | 130 | 34,668.5 | 34,656.7 |
8 | 8 GClass | −17,130.5 | 134 | 34,675.1 | 34,662.9 |
9 | 5 GClass cluster-independent C-factor | −17,188.1 | 89 | 34,651.2 | 34,643.1 |
10 | Model 9 plus four additional direct effects (follow-up—PhD dissertation, habilitation thesis—PhD dissertation, habilitation thesis—anthology, monograph—anthology) | −17,166.7 | 93 | 34,620.8 | 34,612.4 |
11 | Model 10 plus order restriction of the latent clusters | −17,351.5 | 80 | 34,950.2 | 34,943.0 |
MNR . | Models of disciplines . | LL . | NPAR . | BIC(k) . | AIC3 . |
---|---|---|---|---|---|
1 | 1 GClass | −17,789.4 | 106 | 35,906.4 | 35,896.8 |
2 | 2 GClass | −17,328.9 | 110 | 34,997.8 | 34,987.8 |
3 | 3 GClass | −17,211.1 | 114 | 34,774.6 | 34,764.2 |
4 | 4 GClass | −17,155.6 | 118 | 34,676.0 | 34,665.3 |
5 | 5 GClass | −17,139.7 | 122 | 34,656.4 | 34,645.3 |
6 | 6 GClass | −17,134.9 | 126 | 34,659.4 | 34,647.9 |
7 | 7 GClass | −17,133.4 | 130 | 34,668.5 | 34,656.7 |
8 | 8 GClass | −17,130.5 | 134 | 34,675.1 | 34,662.9 |
9 | 5 GClass cluster-independent C-factor | −17,188.1 | 89 | 34,651.2 | 34,643.1 |
10 | Model 9 plus four additional direct effects (follow-up—PhD dissertation, habilitation thesis—PhD dissertation, habilitation thesis—anthology, monograph—anthology) | −17,166.7 | 93 | 34,620.8 | 34,612.4 |
11 | Model 10 plus order restriction of the latent clusters | −17,351.5 | 80 | 34,950.2 | 34,943.0 |
Note : MNR = model number, LL = loglikelihood, NPAR = number of parameters, BIC(k) = Bayesian information criterion for k clusters, AIC3 = Akaike information criterion 3.
To assess the separation between LCs, we calculated entropy-based measures, which varied between 0 and 1.0. They show how well the observed variables were able to predict the class membership (Lukočienė, Varriale and Vermunt 2010 ). For LC, the R 2 entropy amounted to 0.78, for latent clusters R 2 entropy amounted to 0.98. The separation of both the LCs and the latent clusters is therefore very large. Another model validity index is the proportion of classification error. For each project and each LC or latent cluster a posterior probability that a project belongs to the respective class can be estimated. Out of this set of probabilities the highest one indicates the LC to which a project or discipline should be assigned (modal assignment). Overall, the modal assignments can deviate from the expected assignments according to the sum of the posterior probabilities. The classification error indicates the amount of misclassification. For model M 10 the classification error was comparatively low, with 11.0% at the level of projects and 0.7% at the level of disciplines.
Based on Fig. 1 it could be supposed that the LCs do not represent a qualitative configuration but rather a quantitative dimension, in that the individual profiles run largely parallel and differ only in the level, that is, the quantity of research output. To prove this assumption the LCs were order-restricted (model M 11 ). However, the BIC(k) as well as the AIC3 of M 11 strongly increased in comparison to all other models, with the result that the assumption of a quantitative dimension behind the LCs was not very plausible.
To illustrate the meaning of these segments of scientific disciplines, Table 5 shows the distribution of the projects among the four LCs ( Fig. 1 ) of each of the five segments of disciplines (latent clusters). The last column of numbers in Table 5 indicates the size of the LCs or types of research output profiles. The last row of numbers in Table 5 indicates the proportion of disciplines that were in each discipline segment. The latent clusters or segments of scientific disciplines can be described according to the disciplines that belong to them (cluster sizes in per cent of the total number of disciplines in parentheses):
Latent Cluster 1 ‘ Life Sciences and Medicine ’ (31.6%): biology; botany; zoology; geosciences; preclinical medicine; clinical medicine; agricultural, forestry and veterinary sciences.
Latent Cluster 2 ‘ Social Sciences / Arts and Humanities ’ (31.4%): social sciences; jurisprudence; philosophy/theology; history; linguistics and literary studies; art history; other humanities fields.
Latent Cluster 3 ‘ Formal Sciences ’ (13.9%): mathematics; computer sciences; economic sciences.
Latent Cluster 4 ‘ Technical Sciences ’ (13.5%): Other natural sciences; technical sciences; psychology.
Latent Cluster 5 ‘ Physical Sciences ’ (9.6%): physics, astronomy and mechanics; chemistry.
Relative class sizes and distribution of projects among LCs (project output types) within each latent clusters (discipline segments) for M 10 (column per cent)
Latent classes (research output profile types) . | Latent clusters (discipline segments) | LC size . | ||||
---|---|---|---|---|---|---|
. | GClass 1 . | GClass 2 . | GClass 3 . | GClass 4 . | GClass 5 . | . |
LC 1 ‘Not Book’ | 0.00 | 0.14 | 0.37 | |||
LC 2 ‘Book and Non-Reviewed Journal Article’ | 0.00 | 0.02 | 0.00 | 0.36 | ||
LC 3 ‘Multiple Outputs’ | 0.06 | 0.03 | 0.24 | 0.06 | 0.18 | |
LC 4 ‘Journal Article, Conference Contribution, Career Development’ | 0.10 | 0.00 | 0.03 | 0.04 | 0.09 | |
GClass size | 0.32 | 0.31 | 0.14 | 0.14 | 0.10 |
Latent classes (research output profile types) . | Latent clusters (discipline segments) | LC size . | ||||
---|---|---|---|---|---|---|
. | GClass 1 . | GClass 2 . | GClass 3 . | GClass 4 . | GClass 5 . | . |
LC 1 ‘Not Book’ | 0.00 | 0.14 | 0.37 | |||
LC 2 ‘Book and Non-Reviewed Journal Article’ | 0.00 | 0.02 | 0.00 | 0.36 | ||
LC 3 ‘Multiple Outputs’ | 0.06 | 0.03 | 0.24 | 0.06 | 0.18 | |
LC 4 ‘Journal Article, Conference Contribution, Career Development’ | 0.10 | 0.00 | 0.03 | 0.04 | 0.09 | |
GClass size | 0.32 | 0.31 | 0.14 | 0.14 | 0.10 |
Note : LC size = size of the latent class, GClass size = size of the latent clusters, proportions over 0.30 (except for class sizes) are in bold face .
The remaining columns in Table 5 show the distribution of projects in each discipline segment or the probability of a project showing a specific profile type given its latent cluster membership. For instance, of all projects falling into the first GClass 84% are in LC 1 (‘Not Book’), 0% are in LC 2 (‘Book and Non-Reviewed Journal Article’), 6% are in LC 3 (‘Multiple Outputs’), and 10% are in LC 4 (‘Journal Article, Conference Contribution, and Career Development’). High proportions in a cell indicate a strong association of the corresponding segment of disciplines in the column with the corresponding type of research output profile in the row. In this respect the segment ‘Life Sciences and Medicine’ (GClass 1) was strongly associated with the ‘Not Book’ project type (LC 1) (84% of projects of this segment), but 10% of this cluster fell also in the most productive type, ‘Journal Article, Conference Contribution, and Career Development’ (LC 4). In the segment ‘Social Sciences/Arts and Humanities’ (GClass 2) almost all projects (97%) are of the second ‘Book and Non-Reviewed Journal Article’ type (LC 2). Projects of the third segment ‘Formal Sciences’ are classified about 80% in the ‘Multiple Outputs’ type, 14% also in the ‘Not Book’ type. The fourth segment, ‘Technical Sciences’, is rather heterogeneous, with over 95% of the projects of this segment in the first three project types and 37% even in the ‘Book and Non-Reviewed Journal Article’ type (LC 2). The projects of the last segment, ‘Physical Sciences’, can be divided mainly into two groups: 38% in the first project type ‘Not Book’ and 56% in the most productive project type, ‘Journal Article, Conference Contribution, and Career Development’. Overall, except for ‘Humanities’, there is no one-to-one assignment of a segment of disciplines to a special type of research output profile. Disciplines show great heterogeneity in their research output profiles.
Figure 2 shows the LC proportions for each single discipline, structured according to the latent cluster (segments of disciplines). This finding also replicated the basic findings in Table 5 at the level of single disciplines. It is of interest that the ‘Book and Non-reviewed Journal Article’ type (LC 2) played an important role not only in ‘Social Sciences/Arts and Humanities’ but also in ‘Technical Sciences’.
Estimated proportions of the four LCs of projects for each scientific discipline (stacked bars plot), classified into one of five latent clusters (1–5, separated by dashed lines).
To explain the LC membership we conducted a modified multilevel multinomial regression model with the latent-class membership as categorical variable and the set of covariates as predictors ( Vermunt 2010 ). Beforehand, the continuous covariates time, age, duration, overall rating of a proposal (ex ante evaluation), and requested grant sum were z -transformed ( M = 0, S = 1) to facilitate the interpretation of the regression results independently of the units of the covariates ( Table 6 ).
Wald statistics are used to assess the statistical significance of a set of parameter estimates. Using Wald statistics, the restriction is tested that each estimate in a set of parameters associated with a given covariate equals zero ( Vermunt and Magidson 2005b ). A non-significant Wald statistic indicates that the respective covariate does not differ between the LCs. Additionally, we calculated a z -test for each single parameter. There are three covariates that explained the class membership with statistically significant Wald tests: project duration, requested grant sum, and the project head’s age. The overall rating of the proposal (ex ante evaluation), for instance, had no impact on the class membership. Research projects with a duration longer than the average of 39 months were more often in LC 4 (‘Journal Article, Conference Contribution, and Career Development’) than research projects with a shorter than average duration were. The higher the requested grant sum of a project, the less probable it was for the project to be in LC 2 (‘Book and Non-Reviewed Journal Article’), but the more probable it was for it to be in LC 4 (‘Journal Article, Conference Contribution, and Career Development’). Projects where the project head was older than the average age of 47 were more frequently in LC 2 (‘Book and Non-Reviewed Journal Article’), whereas projects where the project head was younger than 47 tended to be in LC 3 (‘Multiple Outputs’). Additionally, the percentage of projects in LC 4 (‘Journal Article, Conference Contribution, and Career Development’) decreased from project end year 2002 to project end 2010.
In sum, projects that belong to the ‘Book and Non-Reviewed Journal Article’ type (LC 2) tended to have rather low requested grant sums and project heads who were older than the average, whereas the most productive ‘Journal Article, Conference Contribution, and Career Development’ type was characterized by above-average requested grant sums and above-average project durations. Further, the percentage of this most productive type decreased over time (time of project end). The third type, ‘Multiple Outputs’, tended to have younger project heads.
Until now it was assumed that output profiles of research projects can be fully explained by the LC or types of output profiles into which the projects were classified. However, as Table 3 shows, projects differed not only with respect to LCs or latent cluster but also with respect to an additional quantitative dimension, a latent C-factor, referring to classical concepts of factor analysis. Unlike LCs, all output variables have positive loadings on this dimension—namely, with the same correlation or loading structure within each LC. Thus, the higher the value in any of the output variable, the higher the value of the C-factor is. Positive values in the C-factor represent productivity above average of the projects in this LC, and negative values indicate projects with less productivity with respect to projects in the same LC. In sum, the C-factor represents productivity differences of projects within each LC, similar to a Mixed-Rasch model in psychometrics ( Mutz, Borchers and Becker 2002 ; Mutz and Daniel 2007 ). This type of ranking can be used by the FWF (and other funding organizations) for comparative evaluation of the output of different projects within a certain time period.
According to the C-factor, the projects within each LC or project type could be ranked ( Fig. 3 ) from left (projects with the highest productivity) to right (projects with the lowest productivity). Additionally, Goldstein-adjusted confidence intervals are shown which makes it possible to interpret non-overlapping intervals of two projects as statistical significant differences at the 5% probability level ( Mutz and Daniel 2007 ). Roughly speaking, only the first and the last 100 projects in each LC actually showed statistically significant differences in their C-factor values.
Rankings of projects within LCs from left (largest amount of research output) to right (smallest amount of research output) and Goldstein-adjusted confidence intervals.
The aim of this study was to conduct a secondary analysis of final report data from the FWF (ex post evaluation) for the years 2002–10 (project end) and—using multilevel LCA—to build bottom-up a typology of research projects and, further, to classify scientific disciplines according to the different proportions of the types of research output profiles found. Referring to our four research questions, the results can be summarized as follows:
The 1,742 completed FWF-funded research projects available for a final report can be classified according to the research output profiles in the following four types with relatively high discrimination: 37% of all projects are in the ‘Not Book’ type, 35.8% in the ‘Book and Non-Reviewed Journal’ type, 17.9% in the ‘Multiple Outputs’ type, and 9.3% in the ‘Journal Article, Conference Contribution, and Career Development’ type, which is the most productive type in terms of number of journal articles and career-related activities. These project types represent primarily a qualitative configuration and not a quantitative dimension according to which projects can be ranked.
The 22 scientific disciplines can be divided into five segments of disciplines based on different proportions of the types of research output profiles: 31.6% of all projects can be classified in the segment ‘Life Science and Medicine’, 31.4% in ‘Social Sciences/Arts and Humanities’, 13.9% in ‘Formal Sciences’, 13.5% in ‘Technical Sciences’ and 9.6% in ‘Physical Sciences’, such as chemistry and physics. Only the ‘Social Sciences/Arts and Humanities’ segment is almost fully associated with one research output profile (‘Book and Non-Reviewed Journal Article’ type); all other segments show different proportions of the four research output profiles. Psychology and economic sciences are usually subsumed under humanities and social sciences. But the MLLCA showed that these two scientific disciplines do not belong to the segment ‘Social Sciences/Arts and Humanities’. Additionally, the fourth and most productive type of research output profile is highly represented (56%) in the fifth segment of disciplines, ‘Physical Sciences’, and with only 10% in ‘Life Science and Medicine’, contrary to the findings of the DFG ( Deutsche Forschungsgemeinschaft 2005 ) mentioned above in the introduction. ‘Life Sciences and Medicine’ is strongly associated (84%) with the ‘Not Book’ type. Projects of the third segment, ‘Formal Sciences’, are classified about 80% in the ‘Multiple Outputs’ type and 14% also in the ‘Not Book’ type. The fourth segment, ‘Technical Sciences’, is rather heterogeneous, with over 90% of the projects in this segment in the first three project types and 37% even in the ‘Book and Non-Reviewed Journal Article’ type. In the end, the findings of the Expert Group on Assessment of University-Based Research set up by the European Commission ( European Commission 2010 ) on the disciplines’ preferred forms of communication are too simple. To sum up, there are not only differences between scientific disciplines in the research output profiles; there is also great heterogeneity of research output profiles within disciplines and segments of disciplines, respectively.
Membership in a particular project type can essentially be explained by three covariates—project duration, requested grant sum, and the project head’s age. Projects that belong to the ‘Book and Non-Reviewed Journal Article’ type tend to be characterized by small requested grant sums and project heads who are older than the average, whereas the most productive type, ‘Journal Article, Conference Contribution, and Career Development’, tends to be characterized by high requested grant sums and longer than average project durations, but whose proportion decreases the more the date of the project termination approximates the year 2010. Reviewers’ overall rating of the proposal (ex ante evaluation) had no influence on latent-class membership.
Projects differ not only in the qualitative configuration of research outputs, their research output profiles, but also with respect to a quantitative dimension that makes productivity rankings of projects possible. The higher the output of a project in each of the research output variables, the higher its value on the quantitative (latent) dimension is. Only the first and the last 100 projects within each project type differed statistically significantly on this dimension.
However, there are also some limitations of our study that have to be discussed: first, the findings represent a specific picture of the research situation in one country, namely, Austria, in a 10-year period situation, and they may not necessarily apply in other countries. The quality of the research was not considered, such as through using international reference values for bibliographic indicators ( Opthof and Leydesdorff 2010 ; Bornmann and Mutz 2011 ) or through using discipline-specific quality criteria. Second, the study included only projects (in particular, ‘Stand-Alone Projects’) that were funded by the FWF. Research projects in Austria that were funded by other research funding organizations, that were not Stand-Alone Projects (40%) or that were funded by higher education institutions themselves could not be included. Further, research projects are mostly financed by mixed funding—that is, in part by grants from various research funding organizations and in part by matching funds from the relevant higher education institution (e.g. human resources), so that research output profiles cannot necessarily be explained by covariates of a single research funding organization. Third, the persons responsible for preparing a report (here, the project heads) always have a certain leeway to mention or not mention certain results of their research as results of the FWF-funded research projects in the final report (e.g. journal articles, career development). In social psychology terms, this phenomenon can be subsumed under the concept of ‘social desirability’ ( Nederhof 1985 ). Social desirability is a psychological tendency to respond in a manner that conforms to consensual standards and general expectancies in a culture. The findings of this study could thus also in part reflect different report policies in the different scientific disciplines.
Despite these limitations, we draw the following conclusions from the results:
Concept of ‘ research output ’ : If the aim is to include all disciplines in the ex post research evaluation, it is necessary to define the term ‘research output’ more broadly, as do the RCUK and the FWF, and to include—in addition to journal articles—also other output categories, such as monographs, anthologies, conference contributions, and patents, in order to treat all disciplines fairly with regard to research output.
Arts and Humanities : As has been repeatedly demanded, the arts and humanities really should be treated as an independent and relatively uniform area ( Nederhof et al. 1989 ; Nederhof 2006 ). Instead of counting only journal articles and their citations, however, it is important to include also monographs and anthologies ( Kousha and Thelwell 2009 ). Psychology and economic sciences do not belong to the segment ‘Social Sciences/Arts and Humanities’. Therefore, it is rather problematic to subsume psychology, economic sciences, social sciences, sociology, and humanities in one unique concept, ‘Social Sciences and Humanities’, as is often the case ( Archambault et al. 2006 ; Nederhof 2006 ).
Hierarchy of the sciences : A most familiar and widespread belief is that scientific disciplines can be classified as ‘hard’ sciences and ‘soft’ sciences, with physics at the top of the hierarchy, social sciences at the bottom and biology somewhere in between ( Smith et al. 2000 ). The strategy followed here made it possible to work out, bottom-up from the research outputs of funded research projects, an empirically based typology of scientific disciplines that at its heart is not hierarchically structured. The typology found reflects much more strongly the real structure of science than the top-down classification systems of sciences allow. However, the identified research output profiles do not unambiguously indicate the segment of the discipline. For instance, almost all projects in the segment ‘Social Sciences/Arts and Humanities’ are of the ‘Book and Non-Reviewed Journal Article’ type, but not all projects of the ‘Book and Non-Reviewed Journal Article’ type are in the segment ‘Social sciences/Arts and Humanities’; there is also a high proportion of ‘Book and Non-Reviewed Journal Article’ type projects in the segment ‘Technical Sciences’.
Research output profiles: Using MLLCA, research projects are not examined with regard to few arbitrarily selected project outputs; instead, the profile or combination of multiple research outputs is analysed. This should receive more attention also in ex post research evaluations of projects.
Ranking of projects : In addition, with MLLCA a qualitative dimension of different types of projects and segments of disciplines can be distinguished from a quantitative dimension that captures research productivity. In this way, projects and possibly also scientific disciplines can be ranked according to their productivity.
Selected model parameters of the regression from LCs on covariates
Covariate . | Latent classes | Overall test Wald . | |||||||
---|---|---|---|---|---|---|---|---|---|
. | LC 1 | LC 2 | LC 3 | LC 4 | . | ||||
. | Not Book | Book and Non-Reviewed Journal Article | Multiple Outputs | Journal Article, Conference Contribution, Career Development | . | ||||
. | Par . | SE . | Par . | SE . | Par . | SE . | Par . | SE . | . |
Time period of the approval decision | −0.11 | 0.63 | −0.85 | 1.09 | −1.05 | 0.91 | 2.01 | 1.05 | 3.73 |
Time period of the project end | 0.40 | 0.63 | 1.01 | 1.05 | 0.81 | 0.90 | −2.22 | 1.09 | 4.26 |
Project duration | −0.11 | 0.29 | −0.93 | 0.47 | −0.52 | 0.40 | 1.56 | 0.51 | 9.62** |
Overall rating of the proposal | −0.19 | 0.15 | −0.16 | 0.25 | −0.04 | 0.21 | 0.40 | 0.26 | 3.53 |
Requested grant sum | −0.28 | 0.19 | −1.17 | 0.37 | 0.45 | 0.26 | 1.00 | 0.28 | 23.90** |
Project head’s sex | 0.51 | 0.61 | −0.10 | 0.97 | −0.72 | 1.12 | 0.30 | 0.76 | 0.77 |
Project head’s age | −0.25 | 0.13 | 0.72 | 0.23 | −0.49 | 0.22 | 0.02 | 0.21 | 13.59** |
Covariate . | Latent classes | Overall test Wald . | |||||||
---|---|---|---|---|---|---|---|---|---|
. | LC 1 | LC 2 | LC 3 | LC 4 | . | ||||
. | Not Book | Book and Non-Reviewed Journal Article | Multiple Outputs | Journal Article, Conference Contribution, Career Development | . | ||||
. | Par . | SE . | Par . | SE . | Par . | SE . | Par . | SE . | . |
Time period of the approval decision | −0.11 | 0.63 | −0.85 | 1.09 | −1.05 | 0.91 | 2.01 | 1.05 | 3.73 |
Time period of the project end | 0.40 | 0.63 | 1.01 | 1.05 | 0.81 | 0.90 | −2.22 | 1.09 | 4.26 |
Project duration | −0.11 | 0.29 | −0.93 | 0.47 | −0.52 | 0.40 | 1.56 | 0.51 | 9.62** |
Overall rating of the proposal | −0.19 | 0.15 | −0.16 | 0.25 | −0.04 | 0.21 | 0.40 | 0.26 | 3.53 |
Requested grant sum | −0.28 | 0.19 | −1.17 | 0.37 | 0.45 | 0.26 | 1.00 | 0.28 | 23.90** |
Project head’s sex | 0.51 | 0.61 | −0.10 | 0.97 | −0.72 | 1.12 | 0.30 | 0.76 | 0.77 |
Project head’s age | −0.25 | 0.13 | 0.72 | 0.23 | −0.49 | 0.22 | 0.02 | 0.21 | 13.59** |
Note : LC = latent class, Par = parameter estimate, SE = standard error, Wald = Wald test, df = degrees of freedom.
*p < 0.05 ( z -test) **p < 0.05 (Wald test, df = 3).
Google Scholar
Google Preview
Month: | Total Views: |
---|---|
December 2016 | 5 |
January 2017 | 4 |
February 2017 | 4 |
March 2017 | 11 |
April 2017 | 6 |
May 2017 | 2 |
June 2017 | 2 |
July 2017 | 7 |
August 2017 | 5 |
September 2017 | 6 |
October 2017 | 13 |
November 2017 | 14 |
December 2017 | 23 |
January 2018 | 28 |
February 2018 | 15 |
March 2018 | 25 |
April 2018 | 47 |
May 2018 | 27 |
June 2018 | 19 |
July 2018 | 26 |
August 2018 | 34 |
September 2018 | 15 |
October 2018 | 25 |
November 2018 | 31 |
December 2018 | 15 |
January 2019 | 20 |
February 2019 | 53 |
March 2019 | 12 |
April 2019 | 34 |
May 2019 | 35 |
June 2019 | 20 |
July 2019 | 29 |
August 2019 | 26 |
September 2019 | 14 |
October 2019 | 21 |
November 2019 | 14 |
December 2019 | 14 |
January 2020 | 12 |
February 2020 | 14 |
March 2020 | 20 |
April 2020 | 13 |
May 2020 | 18 |
June 2020 | 22 |
July 2020 | 16 |
August 2020 | 41 |
September 2020 | 20 |
October 2020 | 24 |
November 2020 | 9 |
December 2020 | 19 |
January 2021 | 20 |
February 2021 | 19 |
March 2021 | 25 |
April 2021 | 28 |
May 2021 | 20 |
June 2021 | 26 |
July 2021 | 16 |
August 2021 | 16 |
September 2021 | 15 |
October 2021 | 27 |
November 2021 | 18 |
December 2021 | 15 |
January 2022 | 18 |
February 2022 | 10 |
March 2022 | 36 |
April 2022 | 18 |
May 2022 | 27 |
June 2022 | 39 |
July 2022 | 47 |
August 2022 | 24 |
September 2022 | 49 |
October 2022 | 31 |
November 2022 | 48 |
December 2022 | 24 |
January 2023 | 37 |
February 2023 | 14 |
March 2023 | 33 |
April 2023 | 21 |
May 2023 | 30 |
June 2023 | 36 |
July 2023 | 18 |
August 2023 | 41 |
September 2023 | 27 |
October 2023 | 19 |
November 2023 | 24 |
December 2023 | 41 |
January 2024 | 33 |
February 2024 | 49 |
March 2024 | 21 |
April 2024 | 25 |
May 2024 | 29 |
June 2024 | 33 |
July 2024 | 43 |
August 2024 | 6 |
Citing articles via.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide
Sign In or Create an Account
This PDF is available to Subscribers Only
For full access to this pdf, sign in to an existing account, or purchase an annual subscription.
What they are & how to write them (with examples)
By: Derek Jansen (MBA) | Expert Reviewed By: Eunice Rautenbach (DTech) | March 2023
If you’re new to academic research, you’re bound to encounter the concept of a “ research problem ” or “ problem statement ” fairly early in your learning journey. Having a good research problem is essential, as it provides a foundation for developing high-quality research, from relatively small research papers to a full-length PhD dissertations and theses.
In this post, we’ll unpack what a research problem is and how it’s related to a problem statement . We’ll also share some examples and provide a step-by-step process you can follow to identify and evaluate study-worthy research problems for your own project.
What is a research problem.
A research problem is, at the simplest level, the core issue that a study will try to solve or (at least) examine. In other words, it’s an explicit declaration about the problem that your dissertation, thesis or research paper will address. More technically, it identifies the research gap that the study will attempt to fill (more on that later).
Let’s look at an example to make the research problem a little more tangible.
To justify a hypothetical study, you might argue that there’s currently a lack of research regarding the challenges experienced by first-generation college students when writing their dissertations [ PROBLEM ] . As a result, these students struggle to successfully complete their dissertations, leading to higher-than-average dropout rates [ CONSEQUENCE ]. Therefore, your study will aim to address this lack of research – i.e., this research problem [ SOLUTION ].
A research problem can be theoretical in nature, focusing on an area of academic research that is lacking in some way. Alternatively, a research problem can be more applied in nature, focused on finding a practical solution to an established problem within an industry or an organisation. In other words, theoretical research problems are motivated by the desire to grow the overall body of knowledge , while applied research problems are motivated by the need to find practical solutions to current real-world problems (such as the one in the example above).
As you can probably see, the research problem acts as the driving force behind any study , as it directly shapes the research aims, objectives and research questions , as well as the research approach. Therefore, it’s really important to develop a very clearly articulated research problem before you even start your research proposal . A vague research problem will lead to unfocused, potentially conflicting research aims, objectives and research questions .
As the name suggests, a problem statement (within a research context, at least) is an explicit statement that clearly and concisely articulates the specific research problem your study will address. While your research problem can span over multiple paragraphs, your problem statement should be brief , ideally no longer than one paragraph . Importantly, it must clearly state what the problem is (whether theoretical or practical in nature) and how the study will address it.
Here’s an example of a statement of the problem in a research context:
Rural communities across Ghana lack access to clean water, leading to high rates of waterborne illnesses and infant mortality. Despite this, there is little research investigating the effectiveness of community-led water supply projects within the Ghanaian context. Therefore, this study aims to investigate the effectiveness of such projects in improving access to clean water and reducing rates of waterborne illnesses in these communities.
As you can see, this problem statement clearly and concisely identifies the issue that needs to be addressed (i.e., a lack of research regarding the effectiveness of community-led water supply projects) and the research question that the study aims to answer (i.e., are community-led water supply projects effective in reducing waterborne illnesses?), all within one short paragraph.
Wherever there is a lack of well-established and agreed-upon academic literature , there is an opportunity for research problems to arise, since there is a paucity of (credible) knowledge. In other words, research problems are derived from research gaps . These gaps can arise from various sources, including the emergence of new frontiers or new contexts, as well as disagreements within the existing research.
Let’s look at each of these scenarios:
New frontiers – new technologies, discoveries or breakthroughs can open up entirely new frontiers where there is very little existing research, thereby creating fresh research gaps. For example, as generative AI technology became accessible to the general public in 2023, the full implications and knock-on effects of this were (or perhaps, still are) largely unknown and therefore present multiple avenues for researchers to explore.
New contexts – very often, existing research tends to be concentrated on specific contexts and geographies. Therefore, even within well-studied fields, there is often a lack of research within niche contexts. For example, just because a study finds certain results within a western context doesn’t mean that it would necessarily find the same within an eastern context. If there’s reason to believe that results may vary across these geographies, a potential research gap emerges.
Disagreements – within many areas of existing research, there are (quite naturally) conflicting views between researchers, where each side presents strong points that pull in opposing directions. In such cases, it’s still somewhat uncertain as to which viewpoint (if any) is more accurate. As a result, there is room for further research in an attempt to “settle” the debate.
Of course, many other potential scenarios can give rise to research gaps, and consequently, research problems, but these common ones are a useful starting point. If you’re interested in research gaps, you can learn more here .
Given that research problems flow from research gaps , finding a strong research problem for your research project means that you’ll need to first identify a clear research gap. Below, we’ll present a four-step process to help you find and evaluate potential research problems.
If you’ve read our other articles about finding a research topic , you’ll find the process below very familiar as the research problem is the foundation of any study . In other words, finding a research problem is much the same as finding a research topic.
Step 1 – Identify your area of interest
Naturally, the starting point is to first identify a general area of interest . Chances are you already have something in mind, but if not, have a look at past dissertations and theses within your institution to get some inspiration. These present a goldmine of information as they’ll not only give you ideas for your own research, but they’ll also help you see exactly what the norms and expectations are for these types of projects.
At this stage, you don’t need to get super specific. The objective is simply to identify a couple of potential research areas that interest you. For example, if you’re undertaking research as part of a business degree, you may be interested in social media marketing strategies for small businesses, leadership strategies for multinational companies, etc.
Depending on the type of project you’re undertaking, there may also be restrictions or requirements regarding what topic areas you’re allowed to investigate, what type of methodology you can utilise, etc. So, be sure to first familiarise yourself with your institution’s specific requirements and keep these front of mind as you explore potential research ideas.
Step 2 – Review the literature and develop a shortlist
Once you’ve decided on an area that interests you, it’s time to sink your teeth into the literature . In other words, you’ll need to familiarise yourself with the existing research regarding your interest area. Google Scholar is a good starting point for this, as you can simply enter a few keywords and quickly get a feel for what’s out there. Keep an eye out for recent literature reviews and systematic review-type journal articles, as these will provide a good overview of the current state of research.
At this stage, you don’t need to read every journal article from start to finish . A good strategy is to pay attention to the abstract, intro and conclusion , as together these provide a snapshot of the key takeaways. As you work your way through the literature, keep an eye out for what’s missing – in other words, what questions does the current research not answer adequately (or at all)? Importantly, pay attention to the section titled “ further research is needed ”, typically found towards the very end of each journal article. This section will specifically outline potential research gaps that you can explore, based on the current state of knowledge (provided the article you’re looking at is recent).
Take the time to engage with the literature and develop a big-picture understanding of the current state of knowledge. Reviewing the literature takes time and is an iterative process , but it’s an essential part of the research process, so don’t cut corners at this stage.
As you work through the review process, take note of any potential research gaps that are of interest to you. From there, develop a shortlist of potential research gaps (and resultant research problems) – ideally 3 – 5 options that interest you.
Step 3 – Evaluate your potential options
Once you’ve developed your shortlist, you’ll need to evaluate your options to identify a winner. There are many potential evaluation criteria that you can use, but we’ll outline three common ones here: value, practicality and personal appeal.
Value – a good research problem needs to create value when successfully addressed. Ask yourself:
Practicality – a good research problem needs to be manageable in light of your resources. Ask yourself:
Personal appeal – a research project is a commitment, so the research problem that you choose needs to be genuinely attractive and interesting to you. Ask yourself:
Depending on how many potential options you have, you may want to consider creating a spreadsheet where you numerically rate each of the options in terms of these criteria. Remember to also include any criteria specified by your institution . From there, tally up the numbers and pick a winner.
Step 4 – Craft your problem statement
Once you’ve selected your research problem, the final step is to craft a problem statement. Remember, your problem statement needs to be a concise outline of what the core issue is and how your study will address it. Aim to fit this within one paragraph – don’t waffle on. Have a look at the problem statement example we mentioned earlier if you need some inspiration.
We’ve covered a lot of ground. Let’s do a quick recap of the key takeaways:
This post was based on one of our popular Research Bootcamps . If you're working on a research project, you'll definitely want to check this out ...
I APPRECIATE YOUR CONCISE AND MIND-CAPTIVATING INSIGHTS ON THE STATEMENT OF PROBLEMS. PLEASE I STILL NEED SOME SAMPLES RELATED TO SUICIDES.
Very pleased and appreciate clear information.
Your videos and information have been a life saver for me throughout my dissertation journey. I wish I’d discovered them sooner. Thank you!
Very interesting. Thank you. Please I need a PhD topic in climate change in relation to health.
Your email address will not be published. Required fields are marked *
Save my name, email, and website in this browser for the next time I comment.
Run a free plagiarism check in 10 minutes, generate accurate citations for free.
Published on October 12, 2022 by Shona McCombes and Tegan George. Revised on November 21, 2023.
A research proposal describes what you will investigate, why it’s important, and how you will conduct your research.
The format of a research proposal varies between fields, but most proposals will contain at least these elements:
Literature review.
While the sections may vary, the overall objective is always the same. A research proposal serves as a blueprint and guide for your research plan, helping you get organized and feel confident in the path forward you choose to take.
Research proposal purpose, research proposal examples, research design and methods, contribution to knowledge, research schedule, other interesting articles, frequently asked questions about research proposals.
Academics often have to write research proposals to get funding for their projects. As a student, you might have to write a research proposal as part of a grad school application , or prior to starting your thesis or dissertation .
In addition to helping you figure out what your research can look like, a proposal can also serve to demonstrate why your project is worth pursuing to a funder, educational institution, or supervisor.
Show your reader why your project is interesting, original, and important. | |
Demonstrate your comfort and familiarity with your field. Show that you understand the current state of research on your topic. | |
Make a case for your . Demonstrate that you have carefully thought about the data, tools, and procedures necessary to conduct your research. | |
Confirm that your project is feasible within the timeline of your program or funding deadline. |
The length of a research proposal can vary quite a bit. A bachelor’s or master’s thesis proposal can be just a few pages, while proposals for PhD dissertations or research funding are usually much longer and more detailed. Your supervisor can help you determine the best length for your work.
One trick to get started is to think of your proposal’s structure as a shorter version of your thesis or dissertation , only without the results , conclusion and discussion sections.
Download our research proposal template
Professional editors proofread and edit your paper by focusing on:
See an example
Writing a research proposal can be quite challenging, but a good starting point could be to look at some examples. We’ve included a few for you below.
Like your dissertation or thesis, the proposal will usually have a title page that includes:
The first part of your proposal is the initial pitch for your project. Make sure it succinctly explains what you want to do and why.
Your introduction should:
To guide your introduction , include information about:
As you get started, it’s important to demonstrate that you’re familiar with the most important research on your topic. A strong literature review shows your reader that your project has a solid foundation in existing knowledge or theory. It also shows that you’re not simply repeating what other people have already done or said, but rather using existing research as a jumping-off point for your own.
In this section, share exactly how your project will contribute to ongoing conversations in the field by:
Following the literature review, restate your main objectives . This brings the focus back to your own project. Next, your research design or methodology section will describe your overall approach, and the practical steps you will take to answer your research questions.
? or ? , , or research design? | |
, )? ? | |
, , , )? | |
? |
To finish your proposal on a strong note, explore the potential implications of your research for your field. Emphasize again what you aim to contribute and why it matters.
For example, your results might have implications for:
Last but not least, your research proposal must include correct citations for every source you have used, compiled in a reference list . To create citations quickly and easily, you can use our free APA citation generator .
Some institutions or funders require a detailed timeline of the project, asking you to forecast what you will do at each stage and how long it may take. While not always required, be sure to check the requirements of your project.
Here’s an example schedule to help you get started. You can also download a template at the button below.
Download our research schedule template
Research phase | Objectives | Deadline |
---|---|---|
1. Background research and literature review | 20th January | |
2. Research design planning | and data analysis methods | 13th February |
3. Data collection and preparation | with selected participants and code interviews | 24th March |
4. Data analysis | of interview transcripts | 22nd April |
5. Writing | 17th June | |
6. Revision | final work | 28th July |
If you are applying for research funding, chances are you will have to include a detailed budget. This shows your estimates of how much each part of your project will cost.
Make sure to check what type of costs the funding body will agree to cover. For each item, include:
To determine your budget, think about:
If you want to know more about the research process , methodology , research bias , or statistics , make sure to check out some of our other articles with explanations and examples.
Methodology
Statistics
Research bias
Once you’ve decided on your research objectives , you need to explain them in your paper, at the end of your problem statement .
Keep your research objectives clear and concise, and use appropriate verbs to accurately convey the work that you will carry out for each one.
I will compare …
A research aim is a broad statement indicating the general purpose of your research project. It should appear in your introduction at the end of your problem statement , before your research objectives.
Research objectives are more specific than your research aim. They indicate the specific ways you’ll address the overarching aim.
A PhD, which is short for philosophiae doctor (doctor of philosophy in Latin), is the highest university degree that can be obtained. In a PhD, students spend 3–5 years writing a dissertation , which aims to make a significant, original contribution to current knowledge.
A PhD is intended to prepare students for a career as a researcher, whether that be in academia, the public sector, or the private sector.
A master’s is a 1- or 2-year graduate degree that can prepare you for a variety of careers.
All master’s involve graduate-level coursework. Some are research-intensive and intend to prepare students for further study in a PhD; these usually require their students to write a master’s thesis . Others focus on professional training for a specific career.
Critical thinking refers to the ability to evaluate information and to be aware of biases or assumptions, including your own.
Like information literacy , it involves evaluating arguments, identifying and solving problems in an objective and systematic way, and clearly communicating your ideas.
The best way to remember the difference between a research plan and a research proposal is that they have fundamentally different audiences. A research plan helps you, the researcher, organize your thoughts. On the other hand, a dissertation proposal or research proposal aims to convince others (e.g., a supervisor, a funding body, or a dissertation committee) that your research topic is relevant and worthy of being conducted.
If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.
McCombes, S. & George, T. (2023, November 21). How to Write a Research Proposal | Examples & Templates. Scribbr. Retrieved August 12, 2024, from https://www.scribbr.com/research-process/research-proposal/
Other students also liked, how to write a problem statement | guide & examples, writing strong research questions | criteria & examples, how to write a literature review | guide, examples, & templates, "i thought ai proofreading was useless but..".
I've been using Scribbr for years now and I know it's a service that won't disappoint. It does a good job spotting mistakes”
4-minute read
Are you planning a research project? If so, you’ll need a research paradigm. But what exactly is a research paradigm, and why is it important? This blog post will cover the following:
● The definition of a research paradigm
● Why research paradigms are important
● Common examples of research paradigms
● Merging research paradigms
● Expert editing and proofreading
Read on to find out more or learn about research paradigms in the video below!
A research paradigm is a philosophical framework that your research is based on. It offers a pattern of beliefs and understandings from which the theories and practices of your research project operate.
A research paradigm consists of ontology, epistemology, and research methodology .
● Ontology answers the question: “What is reality?” That is, does a single reality exist within your research? An example of an ontological question would be: “Does God exist?” There are two possible realities (or ontologies) in response to this question: “Yes, God exists,” or “No, God does not exist.”
● Epistemology is the study of knowledge. It answers the question: “How is it possible to know reality?” Epistemology incorporates the validity, parameters, and methods of acquiring knowledge. An example of an epistemological question would be: “How is it possible to know whether God exists or not?”
● Research Methodology answers the question: “How do we go about discovering the answer or reality?” This includes the process of data collection and analysis. Research methodology should outline how you conduct your research and demonstrate that the findings are valid.
Subscribe to our newsletter and get writing tips from our editors straight to your inbox.
Together, ontology and epistemology comprise research philosophy.
Research philosophy combined with research methodology comprises a research paradigm.
Research paradigms are important because they form the philosophical basis of a research project. Research paradigms influence how different schools of learning (such as the sciences versus the humanities) undertake their research. Once a research philosophy has been determined, an appropriate methodology can be chosen.
Furthermore, a knowledge of the philosophical foundation of your research will increase its quality and improve your performance in any analysis you may have to undergo!
1. Positivism
Positivists believe that there’s a single reality that’s possible to measure and understand. Because of this, they’re most likely to use quantitative methods in their research. Typically, positivists propose a hypothesis that can be proved or disproved using statistical data analysis. Positivism tends to investigate the existence of a relationship between two variables rather than the reason behind it.
2. Constructivism
Constructivists believe that there’s no single reality or truth, but rather multiple realities. They devote themselves to understanding and interpreting the meaning attached to an action. For this reason, constructivists tend to use qualitative research methods , such as interviews or case studies, which focus on providing different perspectives. Constructivism aims to provide the answer to “why.” For example, asking “Why do 25% of the employees of an organization regularly arrive late to work?” rather than merely establishing the relationship between two variables (e.g., time of arrival at work and availability of nearby parking).
3. Pragmatists
Pragmatists believe that reality is continually interpreted and renegotiated against the backdrop of new and unpredictable situations. Because of this, the philosophy they apply in research depends on the research question itself. Pragmatists often combine positivist and constructivist principles in the same research project, using both qualitative and quantitative methods to investigate different components of a research problem. They believe that the optimal research methods are those that most successfully answer the research question.
While most social science research operates from either a positivist (experimental) or constructivist paradigm, it’s possible to combine both, as the field of psychology often does. Quantitative and qualitative methodology are frequently used together in psychology, illustrating the subject’s footing in multiple research paradigms (positivist and constructivist).
If you’re writing a research proposal or paper , you’ll want to ensure that your writing is error-free, fluent, and precise. Although re-reading your own work is valuable, it can be very helpful to get another opinion on your writing. We offer a free trial of proofreading and editing services when you submit your first document. Find out more today!
Post A New Comment
5-minute read
Promoting a brand means sharing valuable insights to connect more deeply with your audience, and...
6-minute read
If you’re seeking funding to support your charitable endeavors as a nonprofit organization, you’ll need...
9-minute read
Is your content getting noticed? Capturing and maintaining an audience’s attention is a challenge when...
8-minute read
Are you looking to enhance engagement and captivate your audience through your professional documents? Interactive...
7-minute read
Voice search optimization is rapidly shaping the digital landscape, requiring content professionals to adapt their...
Are you a creative freelancer looking to make a lasting impression on potential clients or...
Division of Extension
Home » Enhancing Program Performance with Logic Models » Section 2: More about Outcomes » 2.4: Examples of Outputs vs. Outcomes
Try not to confuse outcomes with outputs. Outputs are the activities we do or accomplish that help achieve outcomes. Outcomes are the results of those activities for individuals, families, groups, or communities. Look at the following examples.
Outputs – Activities | Outcomes |
---|---|
The program trains and empowers community volunteers. | Community volunteers have knowledge and skill to work effectively with at-risk youth. |
Program staff teach financial management skills to low-income families. | Low-income families are better able to manage their resources. |
The camp experience provides leadership development opportunities for 4-H youth. | Campers, aged 12-15 years of age, learn new leadership and communication skills while at camp. |
An annual conference disseminates the latest forage research. | Forage producers in Pasture County know current research information and use it to make informed decisions. |
Another way to look at the difference between outputs and outcomes (Hatry, 1999):
Recruiting and training staff and volunteers.
In most cases, recruitment and training refer to internal program functions intended to support or improve program activities. The number of staff and/or volunteers recruited, the number trained, the resources committed to their development, etc. indicate the volume of these internal functions. These aspects help our programs accomplish outcomes; they are not outcomes. They do not represent benefits or changes for program participants or beneficiaries.
If, however, the program is addressing a situation of low volunteer involvement in community affairs and the purpose of the program is to increase volunteering among community residents as a part of a larger community development initiative, then increased numbers of residents volunteering in community life would be an outcome.
This information relates to “participation” or “reach” in our logic model that are part of Outputs. It indicates the volume or extent to which we reached the target audience. It does not indicate whether the participants or clients benefited or are doing anything differently as a result of the program, so it is not an outcome.
If, however, the purpose of the program is to increase use of a service by an underserved group, then numbers using the service would be an outcome. Notice, the outcome is not numbers attending or served; the outcome is expressed as use that indicates behavioral change.
These items refer to activities we undertake and accomplish. They may be classified as “what we do”. These are Outputs. They may be essential aspects that are necessary and make it possible for a group or community to change. But, they do not represent benefits or changes in participants and so are not outcomes.
For our purposes in education and outreach programming, client satisfaction may be necessary but is not sufficient. A participant may be satisfied with various aspects of the program (professionalism of staff, location, facility, timeliness, responsiveness of service, etc) but this does not mean that the person learned, benefited or his/her condition improved. If a participant is pleased and satisfied with the program, it may mean that s/he will fully participate and complete a program. As such, satisfaction can be an important step along the way to outcomes. It, however, is generally not an outcome.
In some cases, we may have to settle for participant satisfaction. In programs where individuals are extremely mobile or it is difficult to track people beyond the immediate program service, satisfaction measures may be the best we can do.
PREVIOUS PAGE
We teach, learn, lead and serve, connecting people with the University of Wisconsin, and engaging with them in transforming lives and communities.
Explore Extension »
Connect with your County Extension Office »
Find an Extension employee in our staff directory »
Get the latest news and updates on Extension's work around the state
Feedback, questions or accessibility issues: [email protected] | © 2024 The Board of Regents of the University of Wisconsin System Privacy Policy | Non-Discrimination Policy & How to File a Complaint | Disability Accommodation Requests
An EEO/AA employer, University of Wisconsin-Madison Division of Extension provides equal opportunities in employment and programming, including Title VI, Title IX, the Americans with Disabilities Act (ADA) and Section 504 of the Rehabilitation Act requirements.
Table of Contents
Before conducting a study, a research proposal should be created that outlines researchers’ plans and methodology and is submitted to the concerned evaluating organization or person. Creating a research proposal is an important step to ensure that researchers are on track and are moving forward as intended. A research proposal can be defined as a detailed plan or blueprint for the proposed research that you intend to undertake. It provides readers with a snapshot of your project by describing what you will investigate, why it is needed, and how you will conduct the research.
Your research proposal should aim to explain to the readers why your research is relevant and original, that you understand the context and current scenario in the field, have the appropriate resources to conduct the research, and that the research is feasible given the usual constraints.
This article will describe in detail the purpose and typical structure of a research proposal , along with examples and templates to help you ace this step in your research journey.
A research proposal¹ ,² can be defined as a formal report that describes your proposed research, its objectives, methodology, implications, and other important details. Research proposals are the framework of your research and are used to obtain approvals or grants to conduct the study from various committees or organizations. Consequently, research proposals should convince readers of your study’s credibility, accuracy, achievability, practicality, and reproducibility.
With research proposals , researchers usually aim to persuade the readers, funding agencies, educational institutions, and supervisors to approve the proposal. To achieve this, the report should be well structured with the objectives written in clear, understandable language devoid of jargon. A well-organized research proposal conveys to the readers or evaluators that the writer has thought out the research plan meticulously and has the resources to ensure timely completion.
A research proposal is a sales pitch and therefore should be detailed enough to convince your readers, who could be supervisors, ethics committees, universities, etc., that what you’re proposing has merit and is feasible . Research proposals can help students discuss their dissertation with their faculty or fulfill course requirements and also help researchers obtain funding. A well-structured proposal instills confidence among readers about your ability to conduct and complete the study as proposed.
Research proposals can be written for several reasons:³
Research proposals should aim to answer the three basic questions—what, why, and how.
The What question should be answered by describing the specific subject being researched. It should typically include the objectives, the cohort details, and the location or setting.
The Why question should be answered by describing the existing scenario of the subject, listing unanswered questions, identifying gaps in the existing research, and describing how your study can address these gaps, along with the implications and significance.
The How question should be answered by describing the proposed research methodology, data analysis tools expected to be used, and other details to describe your proposed methodology.
Here is a research proposal sample template (with examples) from the University of Rochester Medical Center. 4 The sections in all research proposals are essentially the same although different terminology and other specific sections may be used depending on the subject.
If you want to know how to make a research proposal impactful, include the following components:¹
1. Introduction
This section provides a background of the study, including the research topic, what is already known about it and the gaps, and the significance of the proposed research.
2. Literature review
This section contains descriptions of all the previous relevant studies pertaining to the research topic. Every study cited should be described in a few sentences, starting with the general studies to the more specific ones. This section builds on the understanding gained by readers in the Introduction section and supports it by citing relevant prior literature, indicating to readers that you have thoroughly researched your subject.
3. Objectives
Once the background and gaps in the research topic have been established, authors must now state the aims of the research clearly. Hypotheses should be mentioned here. This section further helps readers understand what your study’s specific goals are.
4. Research design and methodology
Here, authors should clearly describe the methods they intend to use to achieve their proposed objectives. Important components of this section include the population and sample size, data collection and analysis methods and duration, statistical analysis software, measures to avoid bias (randomization, blinding), etc.
5. Ethical considerations
This refers to the protection of participants’ rights, such as the right to privacy, right to confidentiality, etc. Researchers need to obtain informed consent and institutional review approval by the required authorities and mention this clearly for transparency.
6. Budget/funding
Researchers should prepare their budget and include all expected expenditures. An additional allowance for contingencies such as delays should also be factored in.
7. Appendices
This section typically includes information that supports the research proposal and may include informed consent forms, questionnaires, participant information, measurement tools, etc.
8. Citations
Writing a research proposal begins much before the actual task of writing. Planning the research proposal structure and content is an important stage, which if done efficiently, can help you seamlessly transition into the writing stage. 3,5
Key Takeaways
Here’s a summary of the main points about research proposals discussed in the previous sections:
Q1. How is a research proposal evaluated?
A1. In general, most evaluators, including universities, broadly use the following criteria to evaluate research proposals . 6
Q2. What is the difference between the Introduction and Literature Review sections in a research proposal ?
A2. The Introduction or Background section in a research proposal sets the context of the study by describing the current scenario of the subject and identifying the gaps and need for the research. A Literature Review, on the other hand, provides references to all prior relevant literature to help corroborate the gaps identified and the research need.
Q3. How long should a research proposal be?
A3. Research proposal lengths vary with the evaluating authority like universities or committees and also the subject. Here’s a table that lists the typical research proposal lengths for a few universities.
Arts programs | 1,000-1,500 | |
University of Birmingham | Law School programs | 2,500 |
PhD | 2,500 | |
2,000 | ||
Research degrees | 2,000-3,500 |
Q4. What are the common mistakes to avoid in a research proposal ?
A4. Here are a few common mistakes that you must avoid while writing a research proposal . 7
Thus, a research proposal is an essential document that can help you promote your research and secure funds and grants for conducting your research. Consequently, it should be well written in clear language and include all essential details to convince the evaluators of your ability to conduct the research as proposed.
This article has described all the important components of a research proposal and has also provided tips to improve your writing style. We hope all these tips will help you write a well-structured research proposal to ensure receipt of grants or any other purpose.
References
Paperpal is a comprehensive AI writing toolkit that helps students and researchers achieve 2x the writing in half the time. It leverages 21+ years of STM experience and insights from millions of research articles to provide in-depth academic writing, language editing, and submission readiness support to help you write better, faster.
Get accurate academic translations, rewriting support, grammar checks, vocabulary suggestions, and generative AI assistance that delivers human precision at machine speed. Try for free or upgrade to Paperpal Prime starting at US$19 a month to access premium features, including consistency, plagiarism, and 30+ submission readiness checks to help you succeed.
Experience the future of academic writing – Sign up to Paperpal and start writing for free!
How to write a phd research proposal.
The future of academia: how ai tools are changing the way we do research, you may also like, dissertation printing and binding | types & comparison , what is a dissertation preface definition and examples , how to write your research paper in apa..., how to choose a dissertation topic, how to write an academic paragraph (step-by-step guide), maintaining academic integrity with paperpal’s generative ai writing..., research funding basics: what should a grant proposal..., how to write an abstract in research papers..., how to write dissertation acknowledgements.
From peer-reviewed papers to book chapters, monographs and conference proceedings
There are three ways to add your research outputs to Pure. This section refers to creating records from the pre-built templates that are available in Pure.
You can also import your research outputs from an online source or using a BibTeX or RIS file .
There are a variety of research outputs and Pure has 47 sub-type templates that can be used. The most common types are listed in the table at the bottom of this section. Please refer to the guide for each type when adding research output records.
A short video has been created demonstrating how to add a publication using the Contribution to Journal > Article template. The principles in this video apply to all the other research output templates. Click on the ≡ icon in the top left hand corner of the video to skip to different sections.
There are a variety of research outputs and Pure has 47 sub-type templates that can be used. These are listed in the table below. Please refer to the guide for each type when adding research output records.
Research output type | Research output sub-type |
---|---|
Contribution to journal
| Article, Letter, comment/debat, Book/Film/Article review, Literature review, Editorial, Special issue, Meeting abstract, Review article, Short survey |
Chapter in Book/Report/Conference proceeding
| Chapter (peer-reviewed), Chapter, Entry for encyclopedia/dictionary, Conference contribution, Foreword/postscript, Other chapter contribution |
Book/Report
| Book, Anthology, Scholarly edition, Commissioned report, Other report |
Contribution to specialist publication | Article, Featured article, Book/Film/Article review, Editorial, Letter, Special issue |
Working paper | Working paper, Discussion paper |
Contribution to conference | Paper, Poster, Abstract, Other |
Non-textual form | Software, Data set/Database, Digital or Visual Products, Web publication/site, Artefact, Exhibition, Performance, Composition, Design, Devices and Products |
Thesis
| Doctoral Thesis, Master's Thesis |
Patent | Patent |
Other contribution | Other contribution |
If you are not sure which template to use, please ask your local Pure contact for guidance. They will also be able to advise you on Open Access and/or REF-related requirements.
Please note that the research output records that you add will be validated by your College or School research administrator. Only validated research output records will be displayed on Edinburgh Research Explorer .
The external persons affiliations on research output records in Pure populate the research network map on the Pure Portal. If there are no external persons affiliations on a research output record, the network map on profile pages will not include that research output record and may appear empty.
Please follow the steps below to add external persons affiliations to the research output records.
It is also possible that there is already a research output record in Pure for your research output. You can ask to be added to this existing record by claiming the record. Please use the guide below to claim content.
Chris Drew (PhD)
Dr. Chris Drew is the founder of the Helpful Professor. He holds a PhD in education and has published over 20 articles in scholarly journals. He is the former editor of the Journal of Learning Development in Higher Education. [Image Descriptor: Photo of Chris]
Learn about our Editorial Process
Research objectives refer to the definitive statements made by researchers at the beginning of a research project detailing exactly what a research project aims to achieve.
These objectives are explicit goals clearly and concisely projected by the researcher to present a clear intention or course of action for his or her qualitative or quantitative study.
Research objectives are typically nested under one overarching research aim. The objectives are the steps you’ll need to take in order to achieve the aim (see the examples below, for example, which demonstrate an aim followed by 3 objectives, which is what I recommend to my research students).
Research aim and research objectives are fundamental constituents of any study, fitting together like two pieces of the same puzzle.
The ‘research aim’ describes the overarching goal or purpose of the study (Kumar, 2019). This is usually a broad, high-level purpose statement, summing up the central question that the research intends to answer.
Example of an Overarching Research Aim:
“The aim of this study is to explore the impact of climate change on crop productivity.”
Comparatively, ‘research objectives’ are concrete goals that underpin the research aim, providing stepwise actions to achieve the aim.
Objectives break the primary aim into manageable, focused pieces, and are usually characterized as being more specific, measurable, achievable, relevant, and time-bound (SMART).
Examples of Specific Research Objectives:
1. “To examine the effects of rising temperatures on the yield of rice crops during the upcoming growth season.” 2. “To assess changes in rainfall patterns in major agricultural regions over the first decade of the twenty-first century (2000-2010).” 3. “To analyze the impact of changing weather patterns on crop diseases within the same timeframe.”
The distinction between these two terms, though subtle, is significant for successfully conducting a study. The research aim provides the study with direction, while the research objectives set the path to achieving this aim, thereby ensuring the study’s efficiency and effectiveness.
I usually recommend to my students that they use the SMART framework to create their research objectives.
SMART is an acronym standing for Specific, Measurable, Achievable, Relevant, and Time-bound. It provides a clear method of defining solid research objectives and helps students know where to start in writing their objectives (Locke & Latham, 2013).
Each element of this acronym adds a distinct dimension to the framework, aiding in the creation of comprehensive, well-delineated objectives.
Here is each step:
You’re not expected to fit every single element of the SMART framework in one objective, but across your objectives, try to touch on each of the five components.
1. Field: Psychology
Aim: To explore the impact of sleep deprivation on cognitive performance in college students.
2. Field: Environmental Science
Aim: To understand the effects of urban green spaces on human well-being in a metropolitan city.
3. Field: Technology
Aim: To investigate the influence of using social media on productivity in the workplace.
4. Field: Education
Aim: To examine the effectiveness of online vs traditional face-to-face learning on student engagement and achievement.
5. Field: Health
Aim: To determine the impact of a Mediterranean diet on cardiac health among adults over 50.
6. Field: Environmental Science
Aim: To analyze the impact of urban farming on community sustainability.
7. Field: Sociology
Aim: To investigate the influence of home offices on work-life balance during remote work.
8. Field: Economics
Aim: To evaluate the effects of minimum wage increases on small businesses.
9. Field: Education
Aim: To explore the role of extracurricular activities in promoting soft skills among high school students.
10. Field: Technology
Aim: To assess the impact of virtual reality (VR) technology on the tourism industry.
11. Field: Biochemistry
Aim: To examine the role of antioxidants in preventing cellular damage.
12. Field: Linguistics
Aim: To determine the influence of early exposure to multiple languages on cognitive development in children.
13. Field: Art History
Aim: To explore the impact of the Renaissance period on modern-day art trends.
14. Field: Cybersecurity
Aim: To assess the effectiveness of two-factor authentication (2FA) in preventing unauthorized system access.
15. Field: Cultural Studies
Aim: To analyze the role of music in cultural identity formation among ethnic minorities.
16. Field: Astronomy
Aim: To explore the impact of solar activity on satellite communication.
17. Field: Literature
Aim: To examine narrative techniques in contemporary graphic novels.
18. Field: Renewable Energy
Aim: To investigate the feasibility of solar energy as a primary renewable resource within urban areas.
19. Field: Sports Science
Aim: To evaluate the role of pre-game rituals in athlete performance.
20. Field: Ecology
Aim: To investigate the effects of urban noise pollution on bird populations.
21. Field: Food Science
Aim: To examine the influence of cooking methods on the nutritional value of vegetables.
The importance of research objectives cannot be overstated. In essence, these guideposts articulate what the researcher aims to discover, understand, or examine (Kothari, 2014).
When drafting research objectives, it’s essential to make them simple and comprehensible, specific to the point of being quantifiable where possible, achievable in a practical sense, relevant to the chosen research question, and time-constrained to ensure efficient progress (Kumar, 2019).
Remember that a good research objective is integral to the success of your project, offering a clear path forward for setting out a research design , and serving as the bedrock of your study plan. Each objective must distinctly address a different dimension of your research question or problem (Kothari, 2014). Always bear in mind that the ultimate purpose of your research objectives is to succinctly encapsulate your aims in the clearest way possible, facilitating a coherent, comprehensive and rational approach to your planned study, and furnishing a scientific roadmap for your journey into the depths of knowledge and research (Kumar, 2019).
Kothari, C.R (2014). Research Methodology: Methods and Techniques . New Delhi: New Age International.
Kumar, R. (2019). Research Methodology: A Step-by-Step Guide for Beginners .New York: SAGE Publications.
Doran, G. T. (1981). There’s a S.M.A.R.T. way to write management’s goals and objectives. Management review, 70 (11), 35-36.
Locke, E. A., & Latham, G. P. (2013). New Developments in Goal Setting and Task Performance . New York: Routledge.
Your email address will not be published. Required fields are marked *
Published: August 08, 2024
One of the most underrated skills you can have as a marketer is marketing research — which is great news for this unapologetic cyber sleuth.
From brand design and product development to buyer personas and competitive analysis, I’ve researched a number of initiatives in my decade-long marketing career.
And let me tell you: having the right marketing research methods in your toolbox is a must.
Market research is the secret to crafting a strategy that will truly help you accomplish your goals. The good news is there is no shortage of options.
Thanks to the Internet, we have more marketing research (or market research) methods at our fingertips than ever, but they’re not all created equal. Let’s quickly go over how to choose the right one.
5 Research and Planning Templates + a Free Guide on How to Use Them in Your Market Research
All fields are required.
Click this link to access this resource at any time.
What are you researching? Do you need to understand your audience better? How about your competition? Or maybe you want to know more about your customer’s feelings about a specific product.
Before starting your research, take some time to identify precisely what you’re looking for. This could be a goal you want to reach, a problem you need to solve, or a question you need to answer.
For example, an objective may be as foundational as understanding your ideal customer better to create new buyer personas for your marketing agency (pause for flashbacks to my former life).
Or if you’re an organic sode company, it could be trying to learn what flavors people are craving.
Next, determine what data type will best answer the problems or questions you identified. There are primarily two types: qualitative and quantitative. (Sound familiar, right?)
Understanding the differences between qualitative and quantitative data will help you pinpoint which research methods will yield the desired results.
For instance, thinking of our earlier examples, qualitative data would usually be best suited for buyer personas, while quantitative data is more useful for the soda flavors.
However, truth be told, the two really work together.
Qualitative conclusions are usually drawn from quantitative, numerical data. So, you’ll likely need both to get the complete picture of your subject.
For example, if your quantitative data says 70% of people are Team Black and only 30% are Team Green — Shout out to my fellow House of the Dragon fans — your qualitative data will say people support Black more than Green.
(As they should.)
You’ll also want to understand the difference between primary and secondary research.
Primary research involves collecting new, original data directly from the source (say, your target market). In other words, it’s information gathered first-hand that wasn’t found elsewhere.
Some examples include conducting experiments, surveys, interviews, observations, or focus groups.
Meanwhile, secondary research is the analysis and interpretation of existing data collected from others. Think of this like what we used to do for school projects: We would read a book, scour the internet, or pull insights from others to work from.
So, which is better?
Personally, I say any research is good research, but if you have the time and resources, primary research is hard to top. With it, you don’t have to worry about your source's credibility or how relevant it is to your specific objective.
You are in full control and best equipped to get the reliable information you need.
Once you know your objective and what kind of data you want, you’re ready to select your marketing research method.
For instance, let’s say you’re a restaurant trying to see how attendees felt about the Speed Dating event you hosted last week.
You shouldn’t run a field experiment or download a third-party report on speed dating events; those would be useless to you. You need to conduct a survey that allows you to ask pointed questions about the event.
This would yield both qualitative and quantitative data you can use to improve and bring together more love birds next time around.
Now that you know what you’re looking for in a marketing research method, let’s dive into the best options.
Note: According to HubSpot’s 2024 State of Marketing report, understanding customers and their needs is one of the biggest challenges facing marketers today. The options we discuss are great consumer research methodologies , but they can also be used for other areas.
1. interviews.
Interviews are a form of primary research where you ask people specific questions about a topic or theme. They typically deliver qualitative information.
I’ve conducted many interviews for marketing purposes, but I’ve also done many for journalistic purposes, like this profile on comedian Zarna Garg . There’s no better way to gather candid, open-ended insights in my book, but that doesn’t mean they’re a cure-all.
What I like: Real-time conversations allow you to ask different questions if you’re not getting the information you need. They also push interviewees to respond quickly, which can result in more authentic answers.
What I dislike: They can be time-consuming and harder to measure (read: get quantitative data) unless you ask pointed yes or no questions.
Best for: Creating buyer personas or getting feedback on customer experience, a product, or content.
Focus groups are similar to conducting interviews but on a larger scale.
In marketing and business, this typically means getting a small group together in a room (or Zoom), asking them questions about various topics you are researching. You record and/or observe their responses to then take action.
They are ideal for collecting long-form, open-ended feedback, and subjective opinions.
One well-known focus group you may remember was run by Domino’s Pizza in 2009 .
After poor ratings and dropping over $100 million in revenue, the brand conducted focus groups with real customers to learn where they could have done better.
It was met with comments like “worst excuse for pizza I’ve ever had” and “the crust tastes like cardboard.” But rather than running from the tough love, it took the hit and completely overhauled its recipes.
The team admitted their missteps and returned to the market with better food and a campaign detailing their “Pizza Turn Around.”
The result? The brand won a ton of praise for its willingness to take feedback, efforts to do right by its consumers, and clever campaign. But, most importantly, revenue for Domino’s rose by 14.3% over the previous year.
The brand continues to conduct focus groups and share real footage from them in its promotion:
What I like: Similar to interviewing, you can dig deeper and pivot as needed due to the real-time nature. They’re personal and detailed.
What I dislike: Once again, they can be time-consuming and make it difficult to get quantitative data. There is also a chance some participants may overshadow others.
Best for: Product research or development
Pro tip: Need help planning your focus group? Our free Market Research Kit includes a handy template to start organizing your thoughts in addition to a SWOT Analysis Template, Survey Template, Focus Group Template, Presentation Template, Five Forces Industry Analysis Template, and an instructional guide for all of them. Download yours here now.
Surveys are a form of primary research where individuals are asked a collection of questions. It can take many different forms.
They could be in person, over the phone or video call, by email, via an online form, or even on social media. Questions can be also open-ended or closed to deliver qualitative or quantitative information.
A great example of a close-ended survey is HubSpot’s annual State of Marketing .
In the State of Marketing, HubSpot asks marketing professionals from around the world a series of multiple-choice questions to gather data on the state of the marketing industry and to identify trends.
The survey covers various topics related to marketing strategies, tactics, tools, and challenges that marketers face. It aims to provide benchmarks to help you make informed decisions about your marketing.
It also helps us understand where our customers’ heads are so we can better evolve our products to meet their needs.
Apple is no stranger to surveys, either.
In 2011, the tech giant launched Apple Customer Pulse , which it described as “an online community of Apple product users who provide input on a variety of subjects and issues concerning Apple.”
"For example, we did a large voluntary survey of email subscribers and top readers a few years back."
While these readers gave us a long list of topics, formats, or content types they wanted to see, they sometimes engaged more with content types they didn’t select or favor as much on the surveys when we ran follow-up ‘in the wild’ tests, like A/B testing.”
Pepsi saw similar results when it ran its iconic field experiment, “The Pepsi Challenge” for the first time in 1975.
The beverage brand set up tables at malls, beaches, and other public locations and ran a blindfolded taste test. Shoppers were given two cups of soda, one containing Pepsi, the other Coca-Cola (Pepsi’s biggest competitor). They were then asked to taste both and report which they preferred.
People overwhelmingly preferred Pepsi, and the brand has repeated the experiment multiple times over the years to the same results.
What I like: It yields qualitative and quantitative data and can make for engaging marketing content, especially in the digital age.
What I dislike: It can be very time-consuming. And, if you’re not careful, there is a high risk for scientific error.
Best for: Product testing and competitive analysis
Pro tip: " Don’t make critical business decisions off of just one data set," advises Pamela Bump. "Use the survey, competitive intelligence, external data, or even a focus group to give you one layer of ideas or a short-list for improvements or solutions to test. Then gather your own fresh data to test in an experiment or trial and better refine your data-backed strategy."
8. public domain or third-party research.
While original data is always a plus, there are plenty of external resources you can access online and even at a library when you’re limited on time or resources.
Some reputable resources you can use include:
It’s also smart to turn to reputable organizations that are specific to your industry or field. For instance, if you’re a gardening or landscaping company, you may want to pull statistics from the Environmental Protection Agency (EPA).
If you’re a digital marketing agency, you could look to Google Research or HubSpot Research . (Hey, I know them!)
What I like: You can save time on gathering data and spend more time on analyzing. You can also rest assured the data is from a source you trust.
What I dislike: You may not find data specific to your needs.
Best for: Companies under a time or resource crunch, adding factual support to content
Pro tip: Fellow HubSpotter Iskiev suggests using third-party data to inspire your original research. “Sometimes, I use public third-party data for ideas and inspiration. Once I have written my survey and gotten all my ideas out, I read similar reports from other sources and usually end up with useful additions for my own research.”
If the data you need isn’t available publicly and you can’t do your own market research, you can also buy some. There are many reputable analytics companies that offer subscriptions to access their data. Statista is one of my favorites, but there’s also Euromonitor , Mintel , and BCC Research .
What I like: Same as public domain research
What I dislike: You may not find data specific to your needs. It also adds to your expenses.
Best for: Companies under a time or resource crunch or adding factual support to content
You’re not going to like my answer, but “it depends.” The best marketing research method for you will depend on your objective and data needs, but also your budget and timeline.
My advice? Aim for a mix of quantitative and qualitative data. If you can do your own original research, awesome. But if not, don’t beat yourself up. Lean into free or low-cost tools . You could do primary research for qualitative data, then tap public sources for quantitative data. Or perhaps the reverse is best for you.
Whatever your marketing research method mix, take the time to think it through and ensure you’re left with information that will truly help you achieve your goals.
Related articles.
Free Guide & Templates to Help Your Market Research
Marketing software that helps you drive revenue, save time and resources, and measure and optimize your investments — all on one easy-to-use platform
You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.
All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .
Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.
Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.
Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.
Original Submission Date Received: .
Find support for a specific problem in the support section of our website.
Please let us know what you think of our products and services.
Visit our dedicated information section to learn more about MDPI.
Research on evaluation of city–industry integration in industrial parks, 1. introduction, 2. evaluation indicator system of city–industry integration in industrial parks, 2.1. the original meaning of industry-city integration, 2.2. indicators of land–industry integration in industrial park, 2.3. indicators of residence–industry integration in industrial park, 3. data collection and processing of city–industry integration in industrial parks, 3.1. data collection methods and procedures, 3.1.1. select the evaluation objects of city–industry integration in etdzs, 3.1.2. determine the scope of the sample etdzs on the map, 3.1.3. pick up the polygon vertex coordinates of factories, green space water area and unbuilt area in etdzs and calculate their area, 3.1.4. obtain the data of enterprises in etdzs, 3.1.5. get data on the land area of residential areas in etdzs, 3.1.6. get air quality index (aqi) data, 3.1.7. obtain the data of rail transit stations, 3.2. calculation process and data characteristics of the complex index—matching degree between residence and environment, 3.2.1. supportive residential area score in industrial parks, 3.2.2. the inverse matching relationship between the air quality composite index and supportive residential area, 3.2.3. matching degree score between residence and environment in industrial parks, 4. evaluation process and results of city–industry integration in industrial parks, 4.1. standardization of basic indicators, 4.2. determine the weight of indicators based on the analytic hierarchy process of expert scores, 4.2.1. modelling the hierarchy, 4.2.2. constructing the comparison discriminant matrix, 4.2.3. hierarchical single sorting with individual expert weights for indicators, 4.2.4. maximum eigenvalues of judgement matrices and consistency tests, 4.2.5. determine the average weight of experts for the indicator, 4.3. evaluation results of city–industry integration in sample industrial parks, 4.4. evaluation verification based on entropy weight method.
5.1. the enlightenment for practice from the benchmark industrial park of city–industry integration: from the chengdu model to the beijing model, 5.2. academic contributions, 6. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, acknowledgments, conflicts of interest.
Name of ETDZ | Vertex Number | Longitude | Latitude |
---|---|---|---|
Beijing ETDZ | 2\1 | 116.593694686889 | 39.7775400221545 |
Beijing ETDZ | 2\2 | 116.566790061568 | 39.8058337127532 |
Beijing ETDZ | 2\3 | 116.563310623168 | 39.8012829059164 |
Beijing ETDZ | 2\4 | 116.552753448486 | 39.7962052557995 |
Beijing ETDZ | 2\5 | 116.532583236694 | 39.8201394667626 |
Beijing ETDZ | 2\6 | 116.512670516967 | 39.8113711500501 |
Beijing ETDZ | 2\7 | 116.502628326416 | 39.8184254477015 |
Beijing ETDZ | 2\8 | 116.476020812988 | 39.8052392565111 |
Beijing ETDZ | 2\9 | 116.478853225708 | 39.8024698349971 |
Beijing ETDZ | 2\10 | 116.470527648925 | 39.7974582172923 |
Beijing ETDZ | 2\12 | 116.492757797241 | 39.7848616184177 |
Beijing ETDZ | 2\13 | 116.483144760131 | 39.7802444862761 |
Beijing ETDZ | 2\14 | 116.495332717895 | 39.7737140133473 |
Beijing ETDZ | 2\15 | 116.493530273437 | 39.7645439197561 |
Beijing ETDZ | 2\16 | 116.481084823608 | 39.7655995323052 |
Beijing ETDZ | 2\17 | 116.486663818359 | 39.7322739560859 |
Beijing ETDZ | 2\26 | 116.503314971923 | 39.7116766334180 |
Beijing ETDZ | 2\27 | 116.532339864756 | 39.7147385924770 |
Beijing ETDZ | 2\29 | 116.530622866242 | 39.7422610182087 |
Beijing ETDZ | 2\30 | 116.547775268554 | 39.7675787622142 |
Beijing ETDZ | 2\32 | 116.575155258178 | 39.7733182071942 |
Digital Scale | Implication |
---|---|
1 | Equally important |
3 | One factor is slightly more important than the other |
5 | One factor is significantly more important than the other |
7 | One factor is more strongly important than the other |
9 | One factor is extremely more important than the other |
2, 4, 6, 8 | The median of the two adjacent judgments above |
First-Level Indicator (Intermediate Layer Element) | Second-Level Indicator (Intermediate Layer Element) | Three-Level Indicator (Factor Layer) | Index Calculation Formula |
---|---|---|---|
Intensive degree of production function areas (Land–industry integration) | Industrial land efficiency | Investment intensity | 0.5 × Registered capital of unit industrial land + 0.5 × paid-in capital of unit industrial land |
Employment density | Number of people paying social security in industrial enterprises/Industrial land area | ||
Density of invention patents on industrial land | Authorized patents of inventions for industrial enterprises /industrial land area | ||
Industrial output intensity | Industrial added value/industrial land area | ||
Service industry land efficiency | Service output intensity | Added value of service sector/land use of service sector(built-up area—factory area—green space and water area—residential area) | |
Density of invention patents on services land | Authorized patents of inventions for services enterprises/service sector land area | ||
Service employment density | Number of people paying social security in services enterprises/services land area | ||
Integration degree of production functional area and life service functional area (Residence–industry integration) | Match degree between residence and environment | - | match degree between residence and environment calculation formula: 1 − |z + z − 1|. z refers to industrial park residential area support, positive indicator. z refers to the composite air quality index of the industrial park, inverse indicator. |
Supportive Rail transit facilities | - | Standardization of rail traffic numbers |
Click here to enlarge figure
First-Level Indicator (Criterion Layer B) | Second-Level Indicator (Sub-Criterion Layer C) | Three-Level Indicator (Elements Layer D) | Index Calculation Formula | |
---|---|---|---|---|
Land–industry integration (Coordination degree and balance between internal industries and carriers of production function zones and ser-vice function zones, B1) | Industrial land efficiency (C11) | Investment intensity (D111) | 0.5 × Registered capital of unit industrial land + 0.5 × paid-in capital of unit industrial land | |
Employment density (D112) | Number of people paying social security in industrial enterprises/Industrial land area (factory area) | |||
Density of invention patents on industrial land (D113) | Authorized patents of inventions for industrial enterprises/industrial land area | |||
Industrial output intensity (D114) | Industrial added value/industrial land area | |||
Service industry land efficiency (C12) | Output intensity of the service sector (D121) | Added value of service sector/land use of service sector, where land area of service sector = built-up area—factory area—green space and water area—residential area | ||
Density of invention patents on services land (D122) | Authorized patents of inventions for services enterprises/service sector land area | |||
Services employment density (D123) | Service employment density = Number of people paying social security in service sector enterprises/service sector land area | |||
Residence–industry integration (Coordination and integration of production functional areas and residential service functional areas, B2) | Matching degree be-tween residence and environment (C21) | Residential area supporting scale (z ) | The standardized value of per capita residential area × 0.5 + the standardized value of the proportion of residential area to built-up area × 0.5 | match degree be-tween residence and environment calculation formula: |z + z − 1| |
Air Quality Composite Index (z ) | AQI standardized value of industrial park× 0.5+ standardized value of (AQI of industrial park ÷ AQI of the mother city of industrial park) × 0.5 | |||
Rail transit supporting facilities (C22) | The range standardization of “number of rail transit stations/built-up area of ETDZs” |
ETDZs | Industrial Investment Intensity | Employment Density in Industrial Area | Patent Density of Industrial Inventions | Industrial Output Intensity | Service Output Intensity | Service Employment Density | Patent Density of Inventions in the Service Sector | Air Quality Composite Index | Match Degree between Residence and Environment | Rail Stations per Unit Area |
---|---|---|---|---|---|---|---|---|---|---|
Tianjin | 0.147 | 0.329 | 0.166 | 0.615 | 0.743 | 0.249 | 0.099 | 0.519 | 0.287 | 0.44 |
Beijing | 1.000 | 1.000 | 1.000 | 1.000 | 0.912 | 1.000 | 1.000 | 0.168 | 0.337 | 1.00 |
Nantong | 0.056 | 0.052 | 0.090 | 0.171 | 0.084 | 0.025 | 0.007 | 0.391 | 0.384 | 0.00 |
Kunshan | 0.024 | 0.059 | 0.132 | 0.554 | 0.405 | 0.056 | 0.043 | 0.493 | 0.679 | 0.000 |
Ningguo | 0.030 | 0.127 | 0.053 | 0.088 | 0.535 | 0.463 | 0.017 | 0.141 | 0.511 | 0.000 |
Daya Bay | 0.024 | 0.015 | 0.004 | 0.162 | 0.069 | 0.004 | 0.000 | 0.328 | 0.742 | 0.000 |
Kunming | 0.055 | 0.064 | 0.034 | 0.182 | 0.076 | 0.062 | 0.029 | 0.370 | 1.000 | 0.180 |
Ningbo Daxie | 0.066 | 0.064 | 0.128 | 0.300 | 0.167 | 0.034 | 0.003 | 0.417 | 0.001 | 0.000 |
Chengdu (Damian) | 0.127 | 0.242 | 0.098 | 0.612 | 0.117 | 0.059 | 0.009 | 0.445 | 0.530 | 0.987 |
Rugao | 0.095 | 0.174 | 0.136 | 0.866 | 0.331 | 0.011 | 0.008 | 0.644 | 0.299 | 0.000 |
Quanzhou | 0.154 | 0.474 | 0.222 | 0.720 | 0.542 | 0.133 | 0.096 | 0.505 | 0.385 | 0.000 |
Jiashan | 0.056 | 0.166 | 0.094 | 0.446 | 1.000 | 0.107 | 0.021 | 0.523 | 0.401 | 0.000 |
Zouping | 0.212 | 0.158 | 0.014 | 0.207 | 0.088 | 0.009 | 0.001 | 0.679 | 0.128 | 0.000 |
Lianyungang | 0.001 | 0.060 | 0.099 | 0.102 | 0.029 | 0.037 | 0.002 | 0.558 | 0.421 | 0.000 |
Hanzhong | 0.019 | 0.030 | 0.005 | 0.580 | 0.229 | 0.033 | 0.000 | 0.677 | 0.574 | 0.000 |
Korla | 0.053 | 0.145 | 0.000 | 0.629 | 0.000 | 0.003 | 0.000 | 0.839 | 0.058 | 0.000 |
Zhangjiagang | 0.173 | 0.465 | 0.247 | 0.358 | 0.138 | 0.042 | 0.018 | 0.482 | 0.692 | 0.000 |
Linyi | 0.033 | 0.047 | 0.028 | 0.061 | 0.008 | 0.000 | 0.001 | 0.625 | 0.693 | 0.000 |
Longyan | 0.000 | 0.000 | 0.076 | 0.000 | 0.213 | 0.120 | 0.045 | 0.372 | 0.256 | 0.000 |
Hai’an | 0.053 | 0.097 | 0.191 | 0.346 | 0.083 | 0.013 | 0.008 | 0.644 | 0.366 | 0.423 |
Wuhan | 0.128 | 0.181 | 0.135 | 0.474 | 0.056 | 0.013 | 0.009 | 0.460 | 0.356 | 0.436 |
Zhengzhou | 0.040 | 0.080 | 0.036 | 0.355 | 0.132 | 0.046 | 0.005 | 0.692 | 0.426 | 0.236 |
Changchun | 0.060 | 0.126 | 0.010 | 0.522 | 0.304 | 0.009 | 0.002 | 0.417 | 0.870 | 0.539 |
First-Level Indicator (Guideline Layer B) | Weight (w) | Second-Level Indicator (Sub-Guideline Layer C) | Weight (w) | Three-Level Indicator or Definitions (Element Layer D) | Weight (w) |
---|---|---|---|---|---|
Land–industry integration (Coordination degree and balance between internal industries and carriers of production function zones and service function zones, B1) | 0.417 | Industrial land efficiency (C11) | 0.597 | Industrial investment intensity D111 | 0.211 |
Industrial employment density D112 | 0.183 | ||||
density of invention patents on industrial land D113 | 0.098 | ||||
Industrial output intensity D114 | 0.508 | ||||
Service industry land efficiency (C12) | 0.403 | Service industry output intensity D121 | 0.512 | ||
density of invention patents on services land D122 | 0.178 | ||||
Services employment density D1232 | 0.31 | ||||
Residence–industry integration (Coordination and integration of production functional areas and residential service functional areas, B2) | 0.583 | Matching degree be-tween residence and environment (C21) | 0.556 | The degree of negative correlation between residential area size and AQI | |
Rail transit supporting facilities (C22) | 0.444 | Rail transit station per unit area |
ETDZs | Land–Industry Integration Weighted Score with Ranking | Industrial Land Efficiency Weighted Score and Ranking | Weighted Score and Ranking of Service Sector Land Use Efficiency | |||
---|---|---|---|---|---|---|
Beijing | 0.982 | 1 | 1.000 | 1 | 0.955 | 1 |
Tianjin | 0.442 | 2 | 0.420 | 4 | 0.475 | 3 |
Quanzhou | 0.438 | 3 | 0.507 | 2 | 0.336 | 5 |
Jiashan | 0.387 | 4 | 0.278 | 12 | 0.548 | 2 |
Rugao | 0.372 | 5 | 0.505 | 3 | 0.174 | 7 |
Kunshan | 0.279 | 6 | 0.310 | 9 | 0.232 | 6 |
Chengdu | 0.266 | 7 | 0.392 | 5 | 0.080 | 14 |
Changchun | 0.244 | 8 | 0.302 | 11 | 0.159 | 8 |
Hanzhong | 0.233 | 9 | 0.304 | 10 | 0.128 | 10 |
Ningguo | 0.217 | 10 | 0.079 | 20 | 0.420 | 4 |
Korla | 0.214 | 11 | 0.357 | 6 | 0.001 | 23 |
Zhangjiagang | 0.231 | 12 | 0.328 | 7 | 0.087 | 12 |
Wuhan | 0.201 | 13 | 0.314 | 8 | 0.034 | 20 |
Zhengzhou | 0.157 | 14 | 0.207 | 14 | 0.083 | 13 |
Ningbo Daxie | 0.153 | 15 | 0.190 | 15 | 0.096 | 11 |
Hai’an | 0.153 | 16 | 0.223 | 13 | 0.048 | 17 |
Zouping | 0.127 | 17 | 0.180 | 16 | 0.048 | 18 |
Kunming | 0.097 | 18 | 0.119 | 17 | 0.063 | 15 |
Nantong | 0.090 | 19 | 0.117 | 18 | 0.052 | 16 |
Daya Bay | 0.069 | 20 | 0.090 | 19 | 0.036 | 19 |
Longyan | 0.067 | 21 | 0.007 | 23 | 0.154 | 9 |
Lianyungang | 0.054 | 22 | 0.073 | 21 | 0.027 | 21 |
Linyi | 0.031 | 23 | 0.049 | 22 | 0.005 | 22 |
ETDZs | Weighted Score and Ranking for Residence–Industry Integration | Match Degree between Residence and Environment and Ranking | Standardized Scores and Rankings of Rail Transit Stations per Unit Area | |||
---|---|---|---|---|---|---|
Chengdu | 0.978 | 1 | 0.971 | 3 | 0.987 | 2 |
Wuhan | 0.579 | 2 | 0.694 | 12 | 0.436 | 5 |
Tianjin | 0.570 | 3 | 0.677 | 14 | 0.437 | 4 |
Hai’an | 0.556 | 4 | 1.000 | 1 | 0 | 8 |
Zhengzhou | 0.554 | 5 | 0.809 | 9 | 0.236 | 6 |
Lianyungang | 0.543 | 6 | 0.978 | 2 | 0 | 8 |
Beijing | 0.529 | 7 | 0.151 | 22 | 1.000 | 1 |
Changchun | 0.525 | 8 | 0.514 | 17 | 0.539 | 3 |
Rugao | 0.509 | 9 | 0.916 | 4 | 0 | 8 |
Daya Bay | 0.497 | 10 | 0.894 | 5 | 0 | 8 |
Jiashan | 0.491 | 11 | 0.884 | 6 | 0 | 8 |
Korla | 0.465 | 12 | 0.837 | 7 | 0 | 8 |
Quanzhou | 0.458 | 13 | 0.824 | 8 | 0 | 8 |
Kunshan | 0.397 | 14 | 0.715 | 10 | 0 | 8 |
Zhangjiagang | 0.396 | 15 | 0.712 | 11 | 0 | 8 |
Zouping | 0.377 | 16 | 0.678 | 13 | 0 | 8 |
Nantong | 0.346 | 17 | 0.622 | 15 | 0 | 8 |
Hanzhong | 0.320 | 18 | 0.577 | 16 | 0 | 8 |
Kunming | 0.285 | 19 | 0.370 | 20 | 0.180 | 7 |
Linyi | 0.255 | 20 | 0.459 | 18 | 0 | 8 |
Ningguo | 0.226 | 21 | 0.408 | 19 | 0 | 8 |
Longyan | 0.204 | 22 | 0.366 | 21 | 0 | 8 |
Ningbo Daxie | 0 | 23 | 0 | 23 | 0 | 8 |
First-Level Indicator | Entropy Weight (w) | Second-Level Indicator | Entropy Weight (w) | Three-Level Indicator or Interpretations | Entropy Weight (w) |
---|---|---|---|---|---|
Land–industry integration | 0.577 | Industrial land efficiency | 0.373 | Industrial investment intensity, | 0.313 |
Industrial employment density | 0.248 | ||||
Density of invention patents on industrial land | 0.329 | ||||
Industrial output intensity | 0.111 | ||||
Service industry land efficiency | 0.627 | Service industry output intensity | 0.142 | ||
Density of invention patents on industrial land | 0.549 | ||||
Services employment density | 0.309 | ||||
Residence–industry integration | 0.423 | Matching degree be-tween residence and environment | 0.1 | The degree of negative correlation between residential area size and AQI | |
Rail transit supporting facilities | 0.9 | Rail transit station per unit area |
Industrial Technology Level | Green Manufacturing Maturity | Negative Externality | Service Sector Development | Residential Area Ratio | Residential Land Area per Capita | Density of Rail Transit Stations | |
---|---|---|---|---|---|---|---|
Beijing | extremely high | extremely high | extremely low | high | medium to high | medium to low | high |
Chengdu | medium | medium to high | medium to low | medium | medium to high | medium to high | high |
The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
Xu, M.; Luo, Y.; Li, D. Research on Evaluation of City–Industry Integration in Industrial Parks. Sustainability 2024 , 16 , 6906. https://doi.org/10.3390/su16166906
Xu M, Luo Y, Li D. Research on Evaluation of City–Industry Integration in Industrial Parks. Sustainability . 2024; 16(16):6906. https://doi.org/10.3390/su16166906
Xu, Mingqiang, Yaoyao Luo, and Dingyao Li. 2024. "Research on Evaluation of City–Industry Integration in Industrial Parks" Sustainability 16, no. 16: 6906. https://doi.org/10.3390/su16166906
Article access statistics, further information, mdpi initiatives, follow mdpi.
Subscribe to receive issue release notifications and newsletters from MDPI journals
Use discount code STYLEBLOG15 for 15% off APA Style print products with free shipping in the United States.
We, the APA Style team, are not robots. We can all pass a CAPTCHA test , and we know our roles in a Turing test . And, like so many nonrobot human beings this year, we’ve spent a fair amount of time reading, learning, and thinking about issues related to large language models, artificial intelligence (AI), AI-generated text, and specifically ChatGPT . We’ve also been gathering opinions and feedback about the use and citation of ChatGPT. Thank you to everyone who has contributed and shared ideas, opinions, research, and feedback.
In this post, I discuss situations where students and researchers use ChatGPT to create text and to facilitate their research, not to write the full text of their paper or manuscript. We know instructors have differing opinions about how or even whether students should use ChatGPT, and we’ll be continuing to collect feedback about instructor and student questions. As always, defer to instructor guidelines when writing student papers. For more about guidelines and policies about student and author use of ChatGPT, see the last section of this post.
If you’ve used ChatGPT or other AI tools in your research, describe how you used the tool in your Method section or in a comparable section of your paper. For literature reviews or other types of essays or response or reaction papers, you might describe how you used the tool in your introduction. In your text, provide the prompt you used and then any portion of the relevant text that was generated in response.
Unfortunately, the results of a ChatGPT “chat” are not retrievable by other readers, and although nonretrievable data or quotations in APA Style papers are usually cited as personal communications , with ChatGPT-generated text there is no person communicating. Quoting ChatGPT’s text from a chat session is therefore more like sharing an algorithm’s output; thus, credit the author of the algorithm with a reference list entry and the corresponding in-text citation.
When prompted with “Is the left brain right brain divide real or a metaphor?” the ChatGPT-generated text indicated that although the two brain hemispheres are somewhat specialized, “the notation that people can be characterized as ‘left-brained’ or ‘right-brained’ is considered to be an oversimplification and a popular myth” (OpenAI, 2023).
OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat
You may also put the full text of long responses from ChatGPT in an appendix of your paper or in online supplemental materials, so readers have access to the exact text that was generated. It is particularly important to document the exact text created because ChatGPT will generate a unique response in each chat session, even if given the same prompt. If you create appendices or supplemental materials, remember that each should be called out at least once in the body of your APA Style paper.
When given a follow-up prompt of “What is a more accurate representation?” the ChatGPT-generated text indicated that “different brain regions work together to support various cognitive processes” and “the functional specialization of different regions can change in response to experience and environmental factors” (OpenAI, 2023; see Appendix A for the full transcript).
The in-text citations and references above are adapted from the reference template for software in Section 10.10 of the Publication Manual (American Psychological Association, 2020, Chapter 10). Although here we focus on ChatGPT, because these guidelines are based on the software template, they can be adapted to note the use of other large language models (e.g., Bard), algorithms, and similar software.
The reference and in-text citations for ChatGPT are formatted as follows:
Let’s break that reference down and look at the four elements (author, date, title, and source):
Author: The author of the model is OpenAI.
Date: The date is the year of the version you used. Following the template in Section 10.10, you need to include only the year, not the exact date. The version number provides the specific date information a reader might need.
Title: The name of the model is “ChatGPT,” so that serves as the title and is italicized in your reference, as shown in the template. Although OpenAI labels unique iterations (i.e., ChatGPT-3, ChatGPT-4), they are using “ChatGPT” as the general name of the model, with updates identified with version numbers.
The version number is included after the title in parentheses. The format for the version number in ChatGPT references includes the date because that is how OpenAI is labeling the versions. Different large language models or software might use different version numbering; use the version number in the format the author or publisher provides, which may be a numbering system (e.g., Version 2.0) or other methods.
Bracketed text is used in references for additional descriptions when they are needed to help a reader understand what’s being cited. References for a number of common sources, such as journal articles and books, do not include bracketed descriptions, but things outside of the typical peer-reviewed system often do. In the case of a reference for ChatGPT, provide the descriptor “Large language model” in square brackets. OpenAI describes ChatGPT-4 as a “large multimodal model,” so that description may be provided instead if you are using ChatGPT-4. Later versions and software or models from other companies may need different descriptions, based on how the publishers describe the model. The goal of the bracketed text is to briefly describe the kind of model to your reader.
Source: When the publisher name and the author name are the same, do not repeat the publisher name in the source element of the reference, and move directly to the URL. This is the case for ChatGPT. The URL for ChatGPT is https://chat.openai.com/chat . For other models or products for which you may create a reference, use the URL that links as directly as possible to the source (i.e., the page where you can access the model, not the publisher’s homepage).
You may have noticed the confidence with which ChatGPT described the ideas of brain lateralization and how the brain operates, without citing any sources. I asked for a list of sources to support those claims and ChatGPT provided five references—four of which I was able to find online. The fifth does not seem to be a real article; the digital object identifier given for that reference belongs to a different article, and I was not able to find any article with the authors, date, title, and source details that ChatGPT provided. Authors using ChatGPT or similar AI tools for research should consider making this scrutiny of the primary sources a standard process. If the sources are real, accurate, and relevant, it may be better to read those original sources to learn from that research and paraphrase or quote from those articles, as applicable, than to use the model’s interpretation of them.
We’ve also received a number of other questions about ChatGPT. Should students be allowed to use it? What guidelines should instructors create for students using AI? Does using AI-generated text constitute plagiarism? Should authors who use ChatGPT credit ChatGPT or OpenAI in their byline? What are the copyright implications ?
On these questions, researchers, editors, instructors, and others are actively debating and creating parameters and guidelines. Many of you have sent us feedback, and we encourage you to continue to do so in the comments below. We will also study the policies and procedures being established by instructors, publishers, and academic institutions, with a goal of creating guidelines that reflect the many real-world applications of AI-generated text.
For questions about manuscript byline credit, plagiarism, and related ChatGPT and AI topics, the APA Style team is seeking the recommendations of APA Journals editors. APA Style guidelines based on those recommendations will be posted on this blog and on the APA Style site later this year.
Update: APA Journals has published policies on the use of generative AI in scholarly materials .
We, the APA Style team humans, appreciate your patience as we navigate these unique challenges and new ways of thinking about how authors, researchers, and students learn, write, and work with new technologies.
American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000
Comments are disabled due to your privacy settings. To re-enable, please adjust your cookie preferences.
Subscribe to the APA Style Monthly newsletter to get tips, updates, and resources delivered directly to your inbox.
Welcome! Thank you for subscribing.
Browse APA Style writing guidelines by category
Full index of topics
remarkRemark \newsiamremark hypothesisHypothesis \newsiamthm claimClaim \newsiamremark exampleExample \newsiamremark notationNotation \newsiamthm resultResult \newsiamthm assumptionAssumption \headers Operator Learning Using Random FeaturesNicholas H. Nelsen and Andrew M. Stuart
Supervised operator learning centers on the use of training data, in the form of input-output pairs, to estimate maps between infinite-dimensional spaces. It is emerging as a powerful tool to complement traditional scientific computing, which may often be framed in terms of operators mapping between spaces of functions. Building on the classical random features methodology for scalar regression, this paper introduces the function-valued random features method. This leads to a supervised operator learning architecture that is practical for nonlinear problems yet is structured enough to facilitate efficient training through the optimization of a convex, quadratic cost. Due to the quadratic structure, the trained model is equipped with convergence guarantees and error and complexity bounds, properties that are not readily available for most other operator learning architectures. At its core, the proposed approach builds a linear combination of random operators. This turns out to be a low-rank approximation of an operator-valued kernel ridge regression algorithm, and hence the method also has strong connections to Gaussian process regression. The paper designs function-valued random features that are tailored to the structure of two nonlinear operator learning benchmark problems arising from parametric partial differential equations. Numerical results demonstrate the scalability, discretization invariance, and transferability of the function-valued random features method.
IMAGES
COMMENTS
The changing nature of research outputs has the potential to affect a wide range of organizations and people. A proactive stance could help drive research towards better practices in information storage, sharing and communication, but requires early action and shared goals at a sector level.
Research results refer to the findings and conclusions derived from a systematic investigation or study conducted to answer a specific question or hypothesis. These results are typically presented in a written report or paper and can include various forms of data such as numerical data, qualitative data, statistics, charts, graphs, and visual aids.
Outputs from Research A research output is the product of research . It can take many different forms or types. See here for a full glossary of output types. The tables below sets out the generic criteria for assessing outputs and the definitions of the starred levels, as used during the REF2021 exercise.
An article published in an academic journal can go by several names: original research, an article, a scholarly article, or a peer reviewed article. This format is an important output for many fields and disciplines. Original research articles are written by one or a number of authors who typically advance a new argument or idea to their field.
Scholarly/research outputs and activities represent the various outputs and activities created or executed by scholars and investigators in the course of their academic and/or research efforts. One common output is in the form of scholarly publications which are defined by Washington University as:
Learn how to write a clear and concise results section for your dissertation, with tips and examples to help you present your findings effectively.
1 Definition An output is an outcome of research and can take many forms. Research Outputs must meet the definition of Research.
This chapter explores what we mean by research project deliverables—particularly the difference between outputs and outcomes. This is an increasingly important distinction to funding bodies. Research outputs, which are key performance indicators for academics,...
Research Report is a written document that presents the results of a research project or study, including the research question, methodology, results, and conclusions, in a clear and objective manner.
Outputs resulting form creative practice as research, including the following subtypes. Artefact, Object, Craftwork. Artefacts, objects or craftworks, exhibited, commissioned or otherwise presented or offered for distribution or sale in the public domain, for example, visual arts, craft and cultural creations.
These mile-stones are often aligned with the output of key research outputs, such as papers, talks or reports, along the way and are likely to result in significant contributions, or individual, thesis chapters.
Abstract. Starting out from a broad concept of research output, this article looks at the question as to what research outputs can typically be expected fr
Learn what the research problem and problem statement are and how to write them. Plain-language explanation with clear, practical examples.
The theoretical research output prediction model highlights predictors such as 'professional activities' and 'individual skills and competence' for specific groupings.
Research proposal purpose Academics often have to write research proposals to get funding for their projects. As a student, you might have to write a research proposal as part of a grad school application, or prior to starting your thesis or dissertation.
A research paradigm is a philosophical framework that your research is based on. It offers a pattern of beliefs and understandings from which the theories and practices of your research project operate. A research paradigm consists of ontology, epistemology, and research methodology. Ontology answers the question: "What is reality?".
Learn about user research deliverables—from basic reports to engaging formats—with our guide. Actionable insights, catered to your audience.
2.4: Examples of Outputs vs. Outcomes. Try not to confuse outcomes with outputs. Outputs are the activities we do or accomplish that help achieve outcomes. Outcomes are the results of those activities for individuals, families, groups, or communities. Look at the following examples. The program trains and empowers community volunteers.
17 Research Proposal Examples. Written by Chris Drew (PhD) | January 12, 2024. A research proposal systematically and transparently outlines a proposed research project. The purpose of a research proposal is to demonstrate a project's viability and the researcher's preparedness to conduct an academic study.
Find out what a research proposal is and when you should write it. This article also has expert tips and advice on how to write one, along with research proposal examples.
Research output. From peer-reviewed papers to book chapters, monographs and conference proceedings. There are three ways to add your research outputs to Pure. This page refers to creating records from the templates that are available in Pure. You can also import your research outputs from an online source or from a BibTeX or RIS file.
Research objectives refer to the definitive statements made by researchers at the beginning of a research project detailing exactly what a research project aims to achieve. These objectives are explicit goals clearly and concisely projected
High Benefit. High Risk. 1) There is always a trade off to be made - when a therapy is very efficacious then there are safety issues and physicians and patients need to balance the two, and 2) Long term unknown adverse events may develop.
What marketing research methods are right for you? Here are nine of the most effective, when you should use them, and how to set them up for success.
The output value, land use structure, enterprise profile, employment rates, investments, air quality, rail transit system and other data points regarding sample industrial parks were collected by means of geofencing as well as through the creation of an enterprise credit information database and development area yearbook.
This post outlines how to create references for large language model AI tools like ChatGPT and how to present AI-generated text in a paper.
An operator is an input-output relationship such that each input and corresponding output is infinite-dimensional. For example, the mapping from the current temperature in a room to the temperature one hour later is an operator. ... Two different lines of research have emerged that address PDE approximation problems with scientific machine ...