research paper about advertisements

• Browse All Articles
• Newsletter Sign-Up

Advertising →

• 20 Jun 2023
• Cold Call Podcast

Elon Musk’s Twitter Takeover: Lessons in Strategic Change

In late October 2022, Elon Musk officially took Twitter private and became the company’s majority shareholder, finally ending a months-long acquisition saga. He appointed himself CEO and brought in his own team to clean house. Musk needed to take decisive steps to succeed against the major opposition to his leadership from both inside and outside the company. Twitter employees circulated an open letter protesting expected layoffs, advertising agencies advised their clients to pause spending on Twitter, and EU officials considered a broader Twitter ban. What short-term actions should Musk take to stabilize the situation, and how should he approach long-term strategy to turn around Twitter? Harvard Business School assistant professor Andy Wu and co-author Goran Calic, associate professor at McMaster University’s DeGroote School of Business, discuss Twitter as a microcosm for the future of media and information in their case, “Twitter Turnaround and Elon Musk.”

• 06 Jan 2021
• Working Paper Summaries

Aggregate Advertising Expenditure in the US Economy: What's Up? Is It Real?

We analyze total United States advertising spending from 1960 to 2018. In nominal terms, the elasticity of annual advertising outlays with respect to gross domestic product appears to have increased substantially beginning in the late 1990s, roughly coinciding with the dramatic growth of internet-based advertising.

• 15 Sep 2020

Time and the Value of Data

This paper studies the impact of time-dependency and data perishability on a dataset's effectiveness in creating value for a business, and shows the value of data in the search engine and advertisement businesses perishes quickly.

• 19 May 2020
• Research & Ideas

Why Privacy Protection Notices Turn Off Shoppers

It seems counterintuitive, but website privacy protection notices appear to discourage shoppers from buying, according to Leslie John. Open for comment; 0 Comments.

• 02 Mar 2020
• What Do You Think?

Are Candor, Humility, and Trust Making a Comeback?

SUMMING UP: Have core leadership values been declining in recent years? If so, how do we get them back? James Heskett's readers provide answers. Open for comment; 0 Comments.

• 06 Aug 2019

Super Bowl Ads Sell Products, but Do They Sell Brands?

Super Bowl advertising is increasingly about using storytelling to sell corporate brands rather than products. Shelle Santana discusses why stories win (or fumble) on game day. Open for comment; 0 Comments.

• 27 Jul 2019

Does Facebook's Business Model Threaten Our Elections?

America's 2016 presidential election was the target of voter manipulation via social media, particularly on Facebook. George Riedel thinks history is about to repeat itself. Open for comment; 0 Comments.

• 10 Oct 2018

The Legacy of Boaty McBoatface: Beware of Customers Who Vote

Companies that encourage consumers to vote online should be forewarned—they may expect more than you promise, according to research by Michael Norton, Leslie John, and colleagues. Open for comment; 0 Comments.

• 27 Sep 2018

Large-Scale Demand Estimation with Search Data

Online retailers face the challenge of leveraging the rich data they collect on their websites to uncover insights about consumer behavior. This study proposes a practical and tractable model of economic behavior that can reveal helpful patterns of cross-product substitution. The model can be used to simulate optimal prices.

• 18 Jun 2018

Warning: Scary Warning Labels Work!

If you want to convince consumers to stay away from unhealthy diet choices, don't be subtle about possible consequences, says Leslie John. These graphically graphic warning labels seem to do the trick. Open for comment; 0 Comments.

• 18 Sep 2017

'Likes' Lead to Nothing—and Other Hard-Learned Lessons of Social Media Marketing

A decade-and-a-half after the dawn of social media marketing, brands are still learning what works and what doesn't with consumers. Open for comment; 0 Comments.

• 26 Jul 2017

The Revolution in Advertising: From Don Draper to Big Data

The Mad Men of advertising are being replaced by data scientists and analysts. In this podcast, marketing professor John Deighton and advertising legend Sir Martin Sorrell discuss the positives and negatives of digital marketing. Open for comment; 0 Comments.

• 13 Mar 2017

Hiding Products From Customers May Ultimately Boost Sales

Is it smart for retailers to display their wares to customers a few at a time or all at once? The answer depends largely on the product category, according to research by Kris Johnson Ferreira and Joel Goh. Open for comment; 0 Comments.

• 06 Mar 2017

Why Comparing Apples to Apples Online Leads To More Fruitful Sales

The items displayed next to a product in online marketing displays may determine whether customers buy that product, according to a new study by Uma R. Karmarkar. Open for comment; 0 Comments.

• 13 Feb 2017

Paid Search Ads Pay Off for Lesser-Known Restaurants

Researchers Michael Luca and Weijia Dai wanted to know if paid search ads pay off for small businesses such as restaurants. The answer: Yes, but not for long. Open for comment; 0 Comments.

• 08 Dec 2016

How Wayfair Built a Furniture Brand from Scratch

What was once a collection of 240 home furnishing sites is now a single, successful brand, Wayfair.com. How that brand developed over time and the challenges and opportunities presented by search engine marketing are discussed by Thales Teixeira. Open for comment; 0 Comments.

• 04 May 2016

What Does Boaty McBoatface Tell Us About Brand Control on the Internet?

SUMMING UP. Boaty McBoatface may have been shot down as the social-media sourced name of a research vessel, but James Heskett's readers are up to their hip-boots in opinions on the matter. Open for comment; 0 Comments.

• 02 May 2016

Why People Don’t Vote--and How a Good Ground Game Helps

Recent research by Vincent Pons shows that campaigners knocking on the doors of potential voters not only improves overall turnout but helps individual candidates win more of those votes. Open for comment; 0 Comments.

• 21 Mar 2016

Can Customer Reviews Be 'Managed?'

Consumers increasingly rely on peer reviews on TripAdvisor and other sites to make purchase decisions, so it makes sense that companies have a stake in wanting to shape those opinions. But can they? Thales Teixeira says a good product trumps all. Open for comment; 0 Comments.

• 28 Oct 2015

A Dedication to Creation: India's Ad Man Ranjan Kapur

How do you build a brand amid the uncertainties and opportunities of a developing market? Harvard Business School Professor Sunil Gupta shares lessons learned from Ranjan Kapur, an iconic figure in the Indian advertising industry. Open for comment; 0 Comments.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Behav Sci (Basel)

Consumer Behaviour to Be Considered in Advertising: A Systematic Analysis and Future Agenda

Ahmed h. alsharif.

1 Azman Hashim International Business School, Universiti Teknologi Malaysia, Skudai 81310, Johor, Malaysia

Nor Zafir Md Salleh

Shaymah ahmed al-zahrani.

2 Department of Economic & Finance, College of Business Administration, Taif University, Taif 21944, Saudi Arabia

Ahmad Khraiwish

3 Department of Marketing, Faculty of Business, Applied Science Private University (ASU), Amman 11931, Jordan

Associated Data

Not applicable.

In the past decade, neurophysiological and physiological tools have been used to explore consumer behaviour toward advertising. The studies into brain processes (e.g., emotions, motivation, reward, attention, perception, and memory) toward advertising are scant, and remain unclear in the academic literature. To fill the gap in the literature, this study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol to extract relevant articles. It extracted and analysed 76 empirical articles from the Web of Science (WoS) database from 2009–2020. The findings revealed that the inferior frontal gyrus was associated with pleasure, while the middle temporal gyrus correlated with displeasure of advertising. Meanwhile, the right superior-temporal is related to high arousal and the right middle-frontal-gyrus is linked to low arousal toward advertisement campaigns. The right prefrontal-cortex (PFC) is correlated with withdrawal behaviour, and the left PFC is linked to approach behaviour. For the reward system, the ventral striatum has a main role in the reward system. It has also been found that perception is connected to the orbitofrontal cortex (OFC) and ventromedial (Vm) PFC. The study’s findings provide a profound overview of the importance of brain processes such as emotional processes, reward, motivation, cognitive processes, and perception in advertising campaigns such as commercial, social initiative, and public health.

1. Introduction

Self-report has been adopted in marketing activities to evaluate and identify consumer responses toward stimuli in the marketing sector, such as advertising practices. According to Carrington, et al. [ 1 ], the self-report methods reflect/measure the conscious responses (e.g., perception, approach/withdrawal attitudes, and thoughts) toward advertising. In fact, self-report cannot measure the unconscious or subconscious responses, which represent the majority of consumer responses, such as decision-making. Thus, researchers and marketers have adopted neuroscience tools such as electroencephalography (EEG) in the marketing field, to better understand the unconscious responses of consumers [ 2 , 3 , 4 ], which has led to an emerging new field, the so-called “Neuromarketing”. Smidts [ 5 ] was the first business researcher who coined the term “neuromarketing”, in 2002. Neuromarketing is a hybrid field containing numerous areas/fields such as psychology, marketing, and neuroscience [ 6 ]. According to Fortunato, et al. [ 7 ], the thanks for spreading this term was given to the Bright House Company when it created the first fMRI laboratory for marketing research.

In the hyper-competitive environment, neuromarketing is a mainstay in advertising because it has an opportunity to gauge consumers’ neural responses as emotional responses toward advertising; thereby, it is a revolutionary field for a better understanding of the subconscious and unconscious consumer responses. According to the literature, the rapid technological progress in neuroscience technology led to a better understanding of consumers’ behaviour in several contexts, such as, but not limited to, advertising [ 8 , 9 , 10 , 11 ]. Therefore, marketing and advertising leaders have used this technology to boost innovation and success in marketing and advertising, by controlling and minimizing task conflict [ 12 ]. According to Ramsoy [ 13 ], neuromarketing tools have been divided into four clusters, as follows: (1) neuroimaging tools such as electroencephalography (EEG), magnetoencephalography (MEG), functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and transcranial magnetic stimulation (TMS); (2) physiological tools such as galvanic skin responses (GSR), eye-tracking (ET), electrocardiogram (ECG), and electromyography (EMG); (3) self-report methods such as surveys, observations, focus groups, and interviews; (4) behavioural measurements, such as the implicit association test (IAT). For instance, neuroimaging tools have been used to gauge emotions, attention, and memory regarding advertising [ 14 , 15 ]. At the same time, physiological tools have been used to gauge the physiological responses of consumers, such as, but not limited to, visual fixation in-store at the purchasing point, thereby providing valuable and fruitful insights into the attitudes of consumers (i.e., approach, withdrawal) [ 16 , 17 , 18 ]. Behavioural measurements are used to measure the reaction time of consumers toward stimuli, and self-reporting is used to measure the conscious behaviour of consumers toward stimuli such as approach/withdrawal attitudes [ 18 ].

According to the literature, the first fMRI investigation in neuromarketing was carried out by McClure, et al. [ 19 ], and largely contributed to the practical studies of neuromarketing [ 18 ]. Therefore, neuromarketing research has received attention from both academia and the industrial world as a means of filling the gap in traditional marketing methods and overcoming the limitations by reducing consumer social-bias (e.g., consumer choices can be affected by others) [ 7 ]. However, understanding the global trends in advertising research within the neuromarketing field is still unclear in academic studies. In addition, to date, no investigation has determined the current neurophysiological and physiological techniques that have been used in studying the unconscious/subconscious responses of consumers toward advertising such as YouTube video scenes, TV ads, public health ads (antismoking), initiative ads (encouraging the use of seat belts in cars). To sum that, this study tries to analyse the extracted articles in depth, to provide a precise and concise conclusion. The contributions of this work are summarized as follows:

• Provides a profound evaluation of the current advertising research that has been used to investigate unconscious and subconscious consumer behaviour, such as emotional dimensions, perceptions, reward processes, and approach/withdrawal motivation toward advertising.
• Provides an overview of the current neurophysiological and physiological tools that were used in advertising within the neuromarketing context between 2009 and 2020.

In this vein, the current paper provides an inclusive overview of neuromarketing research and the current research objectives. Section two presents the methodology and data-collection process. Section three presents the growth of the publication, topics of interest and a thematic analysis. A discussion of the study’s findings is presented in Section four. Section five presents a concise conclusion and the implications of our work. Finally, Section six presents the limitations and future directions.

2. Materials and Methods

This review study is designed to select empirical articles from the Web of Science (WoS) database in advertising within the neuromarketing context, to fill the gap in the literature. The reason behind choosing the WoS database over Scopus is the availability of cleaner data, which means reducing the duplications as compared to the Scopus database; additionally, WoS includes publications in top-tier journals [ 20 ]. The first step was to follow the instructions of the PRISMA protocol to select the empirical articles which used neuroimaging and physiological tools to investigate consumer responses to advertising research within neuromarketing [ 21 ]. The reason behind the use of the PRISMA protocol to select the relevant articles for this study were that it has been widely used in social science and business to extract and select articles related to the study, for example, online learning [ 22 ], neuromarketing [ 23 ], and service and healthcare [ 24 , 25 ].

For the second step, we used the content analysis of selected empirical articles for this study. The above processes will provide us with a profound insight into the advancement in advertising research by identifying and analysing the general and specific areas. Additionally, providing us with a better understanding of advertising research that used neuroimaging and physiological tools/methods and which can be considered when we are conducting further research into advertising research. Therefore, the findings of this study provide a guide for new scholars who are interested in advertising research within a neuromarketing field.

Relevant empirical articles have been selected by using the following query applied to the title, abstract, and keywords: (((“neuromarketing” OR “consumer neuroscience”) AND (“advertising” OR “advertisement”) AND (neuroimaging OR physiological))). This study extracted 76 empirical articles relevant to this review paper from January 2009 to December 2020. This study focused on empirical journal-articles, in comparison to conferences and book chapters, which generally undergo a much more rigorous review process and therefore improve the credibility of published research in journals [ 26 ]. Figure 1 demonstrates the selection of articles processed for this study, as follows:

• Methods: neuroimaging and physiological tools;
• Publication year: January 2009 to December 2020;
• Language: English;
• Document type: original articles (chapters of books, articles from conferences, reviews, and proceedings books were excluded).

PRISMA protocol to extract empirical articles for this systematic study.

3.1. Growth of the Publication

Seventy-six articles were extracted from the WOS which used neuromarketing tools. According to the literature, McClure, et al. [ 19 ] published the first neuromarketing study in 2004, when they investigated the neural correlates of consumer behaviour (e.g., choice, decision-making) toward two brands (Coca-Cola vs. Pepsi Cola). However, the first investigation into advertising was carried out by Morris, et al. [ 27 ]. They found the gyri regions of the brain were highly related to pleasure/displeasure and high/low arousal. From January 2009 to December 2020, there was a fluctuation in the number of published articles in advertising. In 2020, it was the highest number of annual publications with thirteen articles, as depicted in Figure 2 .

By exploring the relevant articles to develop this review article, it was possible to classify the global academic-research trends and advancement in advertising and neuromarketing, as follows: (i) neuroimaging and physiological tools used in advertising; and (ii) consumers’ brain processes to be considered in advertising, such as emotions, motivation, reward process, attentions, perception, and memory. By reviewing the selected articles, we can enrich our understanding, to achieve the objectives of this study. Table 1 shows the summary of the neuroimaging and physiological tools that were used to investigate the brain processes of consumers toward advertising research.

Classifications of neuroimaging and physiological tools used in advertising research.

3.2. Topics of Interest and Thematic Analysis

3.2.1. neuroimaging and physiological tools used in advertising.

We found that the neural-response recording-tool most used in advertising was EEG. EEG is an electrical and also a non-invasive technique to gauge the unconscious/subconscious responses of consumers toward ads by recording the voltage changes of frequencies at scalp regions [ 28 , 29 , 30 , 31 ]. According to [ 32 ], there are five frequency bands (e.g., delta, theta, alpha, beta, and gamma). The EEG tool uses a 10–20 global system, which is used to describe the electrode locations on the volunteer’s scalp, for example, (Fp), (F), (P), (O), (C), and (T) describe the prefrontal, frontal, parietal, occipital, central, and temporal, respectively. Moreover, it uses the same number of electrodes on both side of the volunteer’s head (i.e., right and left side) [ 33 , 34 ]. In addition, it has an excellent temporal accuracy (estimated in milliseconds (ms)) and a poor spatial accuracy (estimated at 1 cm 3 at the scalp regions) [ 35 , 36 , 37 ]. In addition, it is not expensive or noisy [ 38 ]. fMRI and fNIRS are non-invasive, metabolic tools. Both are used to record oxygenated and deoxygenated haemoglobin in the brain [ 39 , 40 ]. fMRI has a superior spatial accuracy (estimated at 1–10 mm 3 of the deep structure of the brain) compared with fNIRS, which has poor spatial accuracy (estimated at 4 cm of cortical-activity regions) [ 41 ]. Meanwhile, both have acceptable temporal accuracy (estimated in seconds) [ 42 , 43 ]. fMRI and fNIRS have been used in marketing research to record the subconscious/unconscious responses of consumers (e.g., preference, perceptions, purchase decisions, choices) toward marketing stimuli [ 41 , 43 , 44 ]. fNIRS is a portable, novel, promising, and silent neuroimaging technique, which is also cheaper than fMRI [ 39 , 45 , 46 ].

ET is a portable technique that is used to gauge physiological reactions such as eye movements, pupil dilation, saccade, and fixation toward the stimuli of marketing [ 18 ]. According to Hoffman [ 47 ], it is used for measuring eye movements and the attention of consumers, which is beneficial for psychology and neurological research. Eye fixations last between 200 and 350 ms while reading a text and watching video scenes, respectively, while 200 ms refers to the duration of saccadic eye-movements [ 48 ]. The set of fixations and saccades is named the scan route, and analyses visual perception and cognitive purpose [ 49 ]. Pupil dilation with a longer blink-period gives us better information about processing [ 18 ]. The GSR and ECG tools are used to gauge the emotional reactions of consumers toward ads [ 50 ]. For example, sweating level is recorded by the GSR tool and the heart rate/heartbeat is measured by the ECG tool [ 50 , 51 , 52 ]. In addition, both of them can measure the autonomic nervous system and evaluate the internal emotional status of consumers [ 53 ]. Therefore, GSR and ECG are convenient and reliable techniques for measuring the emotional status of consumers and changes in skin conductance, respectively [ 54 ]. IAT is able to identify the customers’ attitudes toward marketing stimuli such as brands or ads (e.g., like/dislike) by recording the reaction time of customers [ 18 ]. EMG is used to measure the reactions shown on individuals’ faces (e.g., pleasure/displeasure, …, etc.) toward advertising [ 55 ], because facial-expression analysis is significant for marketers and advertisers, because faces can provide beneficial information about perceptions of customers toward ads in terms of measuring visible and invisible facial-muscle reactions [ 29 ].

3.2.2. Brain Processes to Be Considered in Advertising

Emotion and feelings.

The feeling is a relatively conscious aspect of emotional status [ 56 ], which derives from individuals’ judgments such as level of pleasure or unpleasure toward advertising [ 31 ]; it is likely the best way to understand and explain the physiological responses of the consumer toward ads [ 31 , 57 ]. Many research studies have affirmed that ad-elicited feelings are strong indexes of consumers’ response toward advertising [ 58 ]. On the opposite side of the spectrum, emotion is an unconscious aspect of emotional status which correlates to an automatic somatic response such as increased heart beat in some conditions (fright, anger) [ 56 , 59 ], which is crucial for making decisions, learning, and solving problems [ 60 ]. Advertisers and marketers can use both in advertising to captivate consumers’ attention, thereby enhancing purchase intention. Emotions are accompanied with changes in the autonomic nervous system (i.e., zygomatic facial muscles, corrugator facial muscles, and heart-rate), which can provide rich information about the emotional status of consumers. Therefore, the study of emotions has attracted many researchers and advertisers [ 61 ].

Emotion is constructed from a neural network in the brain, which performs basic psychological activities/functions (e.g., memory, perception, salience detection) [ 62 ]. Therefore, emotion is defined as the set of changes in the individual’s physiological and subconscious and unconscious responses, based on the individual’s experiences [ 63 , 64 ]. In addition, emotion is a relationship between humans and the environment, including multiple subcomponents (e.g., physiological, behavioural, appraisal, and expression [ 65 ]. The cognitive and neurological frameworks of the role of emotion in decision-making has been investigated more through Damasio’s theory, which is also known as the somatic marker hypothesis [ 66 , 67 ]. Consequently, researchers have agreed on two dimensions for measuring emotional responses toward stimuli: (i) valence/balance (i.e., pleasure or displeasure, depression or excitement), (ii) arousal (e.g., high or low) ( Figure 3 ) [ 28 , 68 , 69 ]. Figure 3 shows that the valence indicates either positive or negative emotional-status which is evoked by external stimuli such as advertising. In addition, valence is measured from the positive to the negative side. On the other side of the spectrum, arousal indicates the level of excitement; whether high or low, it is measured from high to low levels [ 68 , 70 , 71 ].

Dimensions model of emotions [ 68 ].

Table 2 shows several methods to measure emotions and feelings toward marketing stimuli. Previous studies have used, for example, self-report and physiological methods to analyse the emotional responses of consumers toward advertising [ 72 ]. For example, the EMG and self-report investigation of Lajante, et al. [ 73 ] gauge the consumer’s pleasure/displeasure toward advertisements. The findings of the experiment revealed that pleasure/displeasure had positively impacted the attitudes of customers toward advertisements. The ECG, EDA, and questionnaire study of Baraybar-Fernández, et al. [ 50 ] found that audio and visual ads with sad messages have more influence on participants. Barquero-Pérez, et al. [ 53 ] used the ECG, EDA and questionnaire in their experiment. They found that each type of ad generated a different emotion, such as surprise, anger, and so forth. In addition, physiological tools have been used to investigate the effectiveness of online advertising (e.g., YouTube) [ 51 ]. For example, Guixeres, et al. [ 51 ] used brain response, ECG, and ET to investigate the relationship between ad effectiveness and the number of views on YouTube. They found that there is a solid relationship between ad effectiveness and the number of views on YouTube. Herrador, et al. [ 74 ] conducted an EDA experiment and the findings revealed that both groups of participants (male and female) indicated a strong initial activation; moreover, they noticed a reduced activation during the most significant part of video material in the male group. Venkatraman, et al. [ 28 ] used several neuromarketing tools to evaluate participant’ responses to a 30-second TV advertisement. The findings revealed that the activity in the ventral striatum could be the predictor of response to advertising. Grigaliunaite and Pileliene [ 75 ] conducted an experiment by using ET and they found that emotional and rational advertising appeals influence consumers’ responses in various ways. The IAT and ET investigation of Pileliene and Grigaliunaite [ 76 ] found that warm-colour temperature attracts more visual attention to the advertisement; thereby generating a positive implicit attitude and inducing the buying intentions toward the advertised product, compared with cool-colour temperature advertisements, whether the spokesperson is a female or male celebrity. Similarly, Grigaliunaite and Pileliene [ 77 ] found that negative smoking images reflected a negative implicit attitude/behaviour of individuals toward those images and smoking behaviour, increasing the influence on individuals’ intention to whether to quit or not to start smoking. The ERP, ET, and questionnaire investigation of Pileliene and Grigaliunaite [ 78 ] found that a well-known female spokesperson has a significant impact on the effectiveness of fast-moving consumer-goods advertising.

Neuromarketing tools to measure emotions and feelings.

There has been an interesting growth in understanding the non-verbal responses of emotional status toward advertising by using neuroscience methods such as fMRI, EEG, fNIRS [ 31 , 51 , 79 , 80 , 81 , 82 ]. For example, Plichta, et al. [ 45 ] conducted an fNIRS experiment to investigate the detection of sensory activity by measuring emotional signals in the auditory field. The findings revealed that pleasurable/unpleasurable sounds increased the activity in the auditory cortex, compared to neutral sounds. Gier, et al. [ 41 ] conducted the fNIRS experiment to explore whether the fNIRS tool had the ability to predict the success of point-of-sale elements by measuring the neural signals of brain regions such as the dlPFC. The findings revealed that the fNIRS has the ability to predict the success elements at the point of sale, relying on the cortical-activity effect.

The EEG investigations of Vecchiato, et al. [ 83 ], and Vecchiato, et al. [ 84 ] found that activity in the right frontal alpha is associated with pleasure/liking ads, while the left frontal alpha correlated to displeasure/disliking ads. Additionally, Vecchiato, et al. [ 85 ] found that there were gender (i.e., male and female) differences in interest toward commercial categories and scenes in two ads. The EEG experiment of Harris, et al. [ 86 ] found that emotional advertisements are more effective than rational advertisements, which leads to a positive change in decision-making, inducing donation, and liking. The findings of Chen, et al. [ 87 ] revealed that e-cigarette ads increased the smoking desire; additionally, e-cigarette ads increased activity in the left middle-frontal-gyrus, the right medial-frontal-gyrus, the right parahippocampus, the left insula, the left lingual gyrus/fusiform gyrus, the right inferior-parietal-lobule, the left posterior-cingulate, and the left angular-gyrus. Wang, et al. [ 88 ] and Royo González, et al. [ 89 ] found that the narrative approach of ads and exposure to branding products have a favourable influence on the consumers’ preferences and excitement. The fMRI investigations of Morris, et al. [ 27 ] and Shen and Morris [ 90 ] found that pleasure and displeasure are correlated with more activity in the inferior frontal- and middle temporal-gyrus, respectively, while low and high arousal is associated with the right superior-temporal- and right middle-frontal-gyrus, consecutively. Leanza [ 91 ] used EEG and found that some of the emotive features of the virtual reality (VR) experience significantly influenced consumers’ preferences. Ramsoy, et al. [ 92 ] found that arousal and cognitive load were highly connected to subsequently stated travel-preferences; moreover, consumers’ subconscious emotional and cognitive responses are not identical to subjective travel-preference. Shestyuk, et al. [ 93 ] found that the EEG is a convenient tool to predict the success of TV programs and determine cognitive processes. Wang, et al. [ 94 ] conducted an experiment to propose a generative-design method using EEG measurements. The findings revealed that the product image that was generated with preference EEG-signals had more preference than the product image generated without preference EEG-signals. Kim, et al. [ 95 ] conducted an experiment to identify the effect of visual art (e.g., Mondrian’s and Kandinsky’s artworks) on consumers’ preferences, by using EEG. The findings showed that the visual effects induced high emotional arousal, which might promote heuristic decision-making. Mengual-Recuerda, et al. [ 96 ] found that food served by a chef positively influences emotions, while dishes with special presentations attract more attention than traditional dishes. The EEG study of Eijlers, et al. [ 31 ] found that arousal is positively connected to prominent ads in the wider population and negatively to consumer attitudes toward these ads.

According to Lang and Bradley [ 97 ], emotions and motivation processes are highly intersected and correlated. Chiew and Braver [ 98 ] and Pessoa [ 99 ] found that cognition and consumers’ behaviours are highly affected by motivational processes. For example, positive motivational stimuli will urge individuals toward achieving goals (e.g., obtain or predict a reward by performing a task correctly) [ 100 ], while negative motivational stimuli can lead to distraction [ 101 ].

Pessoa [ 102 ] and Raymond [ 103 ] suggested that motivational processes are a compass of consumers’ attitudes toward external stimuli to engage with the environment and achieve goals. Higgins [ 104 ] suggested two dimensions for measuring motivational processes such as withdrawal and approach attitudes. Researchers and practitioners attempted to investigate the neural responses of motivational processes to better understand consumers’ behaviours toward advertisements and products [ 83 ]. For example, Cherubino, et al. [ 105 ] carried out an experiment using EEG to investigate the relationship between the PFC and motivational dimensions. The findings revealed that the PFC is related to motivational dimensions, wherein the right PFC correlated to withdrawal attitudes and the left PFC related to approach attitudes. The EEG investigations of Pozharliev, et al. [ 106 ] and Zhang, et al. [ 107 ] recorded the brain responses of subjects toward luxury products (motivations). The findings showed that social motivations have a vital role in influencing the purchase of luxury products in order to satisfy social goals (at least one goal). The EEG investigation of Bosshard, et al. [ 108 ] found that liked brands reflect more motivational aspects and activity signals in the right parietal-cortices than disliked brands.

Therefore, there is a strong relationship between the activation of the PFC and motivational dimensions toward marketing stimuli such as advertisements [ 109 ]. Therefore, marketing researchers and practitioners have to focus on the motivational processes of consumers to orient the marketing mix (e.g., target-appropriate audiences and markets, increasing the effectiveness of ads and products) [ 110 ]. According to previous studies, NM research has used the approach/withdrawal attitude to evaluate TV ads [ 111 ]. Therefore, approach/withdrawal motivational attitudes are highly significant for marketing and advertising research.

Reward Processing

According to the literature, it is highly significant for researchers and practitioners to study and know the neural responses that are responsible for reward processing, such as money, food, and social activities [ 112 , 113 , 114 , 115 ]. This is because the positive rewards such as gaining money, foods, or other types of rewards, enhance the accuracy and cognitive task [ 116 , 117 , 118 ] through modifying the early attentional process. Anderson, et al. [ 101 ] demonstrated that visual features (e.g., product design) that are correlated to reward, will make the consumer prioritize, therefore attracting the consumer’s attention automatically. For example, the design preference of a product or brand can increase the activity in the regions which are responsible for reward processing, thereby, causing more activation in regions of motivations that might impact consumers’ purchase decisions [ 29 ]. Many studies concentrated on the individual’s response toward a monetary reward by studying the approach/avoidance attitude [ 112 , 113 ]. For example, Bechara, et al. [ 119 ] carried out an experiment named the “Iowa Gambling Task” by using GSR to investigate the influence of reward on decision-making. They divided participants into two groups, the healthy group and the group with lesions in the vmPFC. The findings revealed that healthy participants sweated more, which led them to infer that participants had a negative emotional experience toward picking up cards from a monetary losing deck; meanwhile, the lesion group picked up cards regardless of whether they were cards with monetary wins or losses. Consequently, reward highly influenced decision-making [ 112 , 120 , 121 ].

Many researchers have confirmed that the importance of the striatum activity in reward processing, wherein the components of the striatum such as the caudate nucleus, nucleus accumbens (NAcc), and putamen play a central role in expectation and evaluation of reward [ 115 , 122 , 123 ]. For example, Galvan [ 124 ] and Geier, et al. [ 125 ] carried out an experiment to investigate the relationship between reward processing and the striatum. Their findings revealed that the ventral striatum (VS) has a key role in the prediction of reward. Jung, et al. [ 126 ] found that the rewards, memory, semantics, and attention regions in the brain were lit up when viewing a combination of a celebrity face and a car, compared with viewing a combination of a non-celebrity face and a car. In addition, car favourableness correlated positively with activation in the left anterior-insula, left OFC, and left higher-order visual cortex in the OL. Padmanabhan, et al. [ 127 ] investigated the influence of the reward system on attention processes. Their findings showed that incentives facilitate cognitive control. Previous neuroimaging studies demonstrated that rewards activate the ventral medial prefrontal cortex (vmPFC) and ventral striatum [ 128 , 129 , 130 ]. The ventral striatum has been discussed before as a part of the reward system [ 131 ]. Therefore, findings suggest that neurodevelopmental changes in the striatum systems may contribute to changes in how reward influences attentional processes [ 56 ].

Attention is defined as the way “people tend to seek, accept and consume the messages that meet their interests, beliefs, values, expectations and ideas, and overlook the messages that are incompatible with this system” [ 132 ]. It has also been defined as selective perception [ 133 ]. Selective perception is associated with filtering out information and concentrating on significant information (e.g., different aspects of stimulus or different stimuli) [ 134 ]. For instance, consumers are exposed to nearly 10 million bits of visual information (e.g., ads, images, sound, video, and colour) per second through their senses (e.g., eyes, ears, skin) daily. Most input information goes by unnoticed, with consumers able to process almost 40 bits of input information per second [ 29 , 135 ]. This lead us to infer that attention has a strong influence on consumer behaviour in how consumers represent, perceive and process information and thus select and prioritize information. Attentional and emotional processes are relatively intersected/connected, and emotion is considered a reliable and effective source for attracting consumers ’ attention [ 136 , 137 ]. For example, the activation in the amygdala (AMY) and cingulate cortex (CC) in the brain is related to emotional stimuli.

Attention is a significant brain process, which has a central role in measuring the effectiveness of advertising campaigns; thereby, it is an indicator of consumer’s behaviour and the effectiveness of advertising [ 138 ]. According to the literature, the majority of researchers have agreed on two systems to measure attention toward advertising: (i) bottom-up, and (ii) top-down, systems [ 28 , 139 , 140 ]. Bottom-up (visual saliency/exogenous/ involuntary) attention is the type of attentional system which is initiated by external stimuli such as colour, voice, promotion, faces, text, novelty, brightness, and so forth, which lead to a process in which information in external stimuli is received automatically. Top-down (goal-driven/endogenous/voluntary) attention is the other type of attentional system, which is initiated by internal and external goals and expectation; thereby, it is necessary to focus all your mental power toward the goal that you are looking to achieve, thereby filtering goals to achieve your goals ( Figure 4 ) [ 2 , 140 , 141 ].

The bottom-up and top-down attention processes [ 59 ].

For this reason, the underlying brain reactions of attention and visual processing are highly interesting for advertising. Moreover, the anterior cingulate cortex (ACC) is highly related to the function of top-down and bottom-up attentional systems [ 142 , 143 ]. For example, Smith and Gevins [ 144 ] found that the occipital lobe (OL) is associated with attentional processes toward TV advertisements. The EEG investigation of Kong, et al. [ 145 ] found that variation in activity in the cerebral hemisphere related to the cognitive task can help to determine the success or lack of success of the advertisement. A recent fMRI investigation by Casado-Aranda, et al. [ 146 ] found that the correspondence between advertising and gender voice (male, female) induces attention regions in the brain. Ananos [ 147 ] carried out experiment using ET to investigate the attention level and processing of information in advertising (content recognition) among elderly and young people groups. Their findings revealed that the attention level among both groups was the same, but recognition by the young-people group was higher than that of the elderly-people group for untraditional advertising. Guixeres, et al. [ 51 ] conducted an experiment to investigate the effectiveness of an ad (e.g., a recall ad) and the number of views on YouTube channels, using neural networks and neuroscience-based metrics (e.g., brain response, ECG, and ET). Their findings suggest an important relationship between neuroscience metrics and self-report of ad effectiveness (e.g., recall ad) and the number of views on YouTube. Cuesta-Cambra, et al. [ 148 ] investigated how information is processed and learned and how visual attention takes place. Their findings indicated that the visual activity of men has different patterns from women, and does not influence subsequent recall, wherein recall relies on the emotional value of ads and simplicity, while complex ads need more visual fixation and are therefore hard to remember. In addition, the importance of the playful component of memory and low-involvement processing were confirmed by EEG. Treleaven-Hassard, et al. [ 149 ] examined the engagement of the consumer with interactive TV ads with a particular brand’s logo compared with non-interactive TV ads. The findings revealed that brands linked with interactive ads attract more automatic attention. Boscolo, et al. [ 81 ] conducted an experiment using ET and questionnaires to investigate differences in the visual attention between genders (male and female), toward print ads. Their findings revealed that there is difference in visual attention in the case of male, while no differences were noticed in the case of females.

According to Simson [ 150 ], studies into the perception of value and how it is formed reflect what is known in marketing theory, wherein the marketing-mix elements can be changed to influence the perceived value of a product. However, studies on how attention systems impact consumers ’ perception and actions have been limited to consumer report and behavioural studies, which depend on a rational report; this is not enough to explain attention processes, wherein there are two attentional systems influencing consumers’ perceptions (e.g., top-down and bottom-up attention system) ( Figure 5 ) [ 59 ]. Consumer perception is the first step in engagement with marketing stimuli or any other stimuli in the environment [ 151 ]. Hogg, et al. [ 152 ] defined perception as “the process by which marketing stimuli are selected, organised, and interpreted”. Therefore, individuals add meaning and interpret it in a certain way, which leads to the perceptions of the individual’s findings for each one. As stated by Belch and Belch [ 153 ] perception processing is extremely reliant on internal processes such as prior knowledge (experiences), current goals, beliefs, expectations, needs, and moods, and also external stimuli such as colour, orientation, intensity, and movement [ 59 ]. Although this explains the process of how consumer perceptions are formed, the exact the part concerning the explanation of sensations and the internal and unique assignation of meaning to sensations is what lies concealed, and remains unexplained in detail in the current consumer-behaviour literature. However, it is widely believed that this process is driven by the unconscious.

Two attentional systems impact the consumers’ perceptions.

Cartocci, et al. [ 154 ] and Modica, et al. [ 155 ] conducted experiments to estimate the accuracy measurement of the cerebral and emotional perception of social advertising campaigns (i.e., antismoking) using several methods such as EEG, GSR, and ECG. The findings showed that the antismoking campaign which was characterized by a symbolic communication style gained the highest approach-values, as evaluated by the approach/withdrawal index. Meanwhile, an image based on “fear-arousing appeal” and with a narrative style reported the highest and lowest effort-values index, respectively. The fMRI investigation of Falk, et al. [ 156 ] predicted the out-of-sample (population) effectiveness of quit-smoking ads. The findings revealed that activity in the prior mPFC was largely predictive of the success of various advertising campaigns in the real world. Plassmann, et al. [ 157 ] carried out an experiment to study the perception of pleasantness in the taste of wines, using the fMRI tool. Their findings showed a stronger activation in the medial OFC (mOFC) regions in the brain when subjects believed they are drinking expensive wine, showing that the mOFC is responsible for experiencing pleasantness. This led to infer that the pleasantness report was correlated with perceived value and price of product more than taste itself. Neuroscientists have found that the OFC and ventromedial prefrontal cortex (vmPFC) are involved in decision-making, through the perceived value of products [ 158 ]. Nuñez-Gomez, et al. [ 159 ] carried out an experiment using EEG to examine how advertising material is perceived by two groups (e.g., healthy group and group with Asperger syndrome). The findings revealed that there are large difference between these groups in their perception of emotion and their attention variables. Gong, et al. [ 160 ] carried out an experiment to identify the influence of sales promotion (e.g., gift-giving, discount) on the perception of consumers and purchase decisions by using EEG/ERP. The findings revealed that discount promotions have more impact on purchase decisions than gift-giving sales promotions.

Memory is defined as an ongoing learning-process, which has input and output functions in the brain [ 161 , 162 ]. The input function encodes information, while the output function retrieves information, and this is very important for advertising research [ 137 , 163 ]. For example, recall and recognition advertising-information is a retrieving function [ 28 ]. Myers and DeWall [ 162 ] and Atkinson and Shiffrin [ 163 ] presented the multistore model of memory, which includes three steps, as follows: (i) a sensory memory, (ii) short-term memory (STM), and (iii) long-term memory (LTM) ( Figure 6 ) [ 164 ]. Brain processes related to memory have revealed certain valuable outcomes, as to the factors which influence the consumers’ behaviour ,such as recall- and recognition-advertising [ 165 ]. Input and output functions in the memory are highly important for marketers and advertisers, due to each function having an awareness and unawareness aspect in the brain [ 137 , 166 ]. Memory and emotion are highly connected to each other. For example, previous studies confirmed that the emotional events are usually remembered more than neutral events, and especially if emotions correspond to events at that moment [ 167 ].

The information phases in the memory’s stages [ 164 ].

The memory process has been widely studied, and it has concluded that the hippocampus (HC), located in the temporal lobe (TL), plays a major role in generating and processing memories [ 165 ]. Additionally, activation of the HC has a strong relationship with LTM and STM, which highly impacts consumers’ purchase decisions [ 168 , 169 ]. In addition, the AMY is located next and close to the HC, which is significant for the memory system [ 165 ]. For example, the EEG investigation by Rossiter, et al. [ 170 ] found that the left hemisphere is correlated with encoding in the LTM. Similarly, the EEG investigations by Astolfi, et al. [ 171 ] used EEG to determine the brain regions that were triggered by the successful memory-encoding of TV ads. They found stronger activity in the cortical regions. Morey [ 172 ] investigated the impact of advertising message on recognition memory. The findings revealed stronger activity in the gamma band, which directly affected memory. The fMRI investigation by Bakalash and Riemer [ 173 ] and Seelig, et al. [ 174 ] measured the brain regions of memory ads. The findings revealed that stronger activity in the amygdala (AMY) and frontotemporal regions is associated with memorable and unmemorable ads. Similarly, [ 175 ] carried out experiments to investigate the content of ads and the activity of frontal regions and memory. The findings showed that the content of ads increased the activity in the frontal regions and the input function (encoding) of memory.

The study of these mental processes such as emotion and feelings, attention, memory, reward processing, motivation, and perception are highly important considerations for advertising research.

4. Discussion

A total of 76 articles were extracted and analysed, wherein the content analysis of the relevant articles revealed that the annual and the accumulative number of publications has been growing since 2009, reaching its peak in 2020 with twelve empirical articles that used neuroimaging, physiological, and self-report techniques to study the consumers’ brain processes such as, but not limited to, emotions toward the stimuli of marketing such as advertising. We followed the PRISMA protocol to select the relevant empirical articles for this study as brain processes such as emotions, feelings, motivation, reward, attention, and memory need to be considered in advertising research. The findings of the study revealed that the neuroimaging tool that is used most in studying the brain processes of consumers is the EEG, with 38 empirical articles, followed by the fMRI with 20 articles; it was also noticed that only four articles used the fNIRS tool in neuromarketing research. In addition, for physiological tools, it was observed that five techniques were used in neuromarketing studies to investigate consumer responses toward the stimuli of marketing such as advertising. The ET was used in 14 articles alongside neuroimaging tools such as EEG, wherein ET is the most used tool, followed by GSR with 12 articles; it was also used alongside other physiological tools such as ECG and EMG. Finally, self-report (i.e., surveys, interviews, focus groups, and observation) was used in seven articles.

This study found that the brain processes to be considered most in advertising research are emotions, feelings, attention, memory, perception, approach/withdrawal motivation, and reward processing. The findings demonstrated that the strongest activity in the inferior-frontal- and middle-temporal-gyri are associated with pleasure and displeasure, while the activity in the right superior-temporal and the right middle-frontal-gyrus correlated with high or low arousal [ 90 ]. As well as this, we found the OL connected with the attention system [ 144 ], and the HC, located in the temporal lobe (TL), plays a major role in generating and processing memories [ 165 ]. In addition, the VS, which is located in the basal ganglia plays a central role in the reward system; for example, the key functions of VS (i.e., control movement and planning) have a vital role in the reward system, while the components of VS such as the putamen, caudate nucleus, and nucleus accumbens (NAcc) have a central role in the assessment of consumer expectations, compared to the actual reward received [ 123 ]. In addition, the ventral tegmental region is considered a part of the reward system, which passes the neurotransmitter dopamine to other brain areas, thereby affecting goal-seeking behaviour [ 123 ]. For motivation, it was found that the anterior cerebral hemispheres play a central role in withdrawal and approach motivation; for example, the increase in activation in the right PFC is linked to withdrawal behaviour, while the increase in activation in the left PFC is associated with approach behaviour [ 105 , 176 ]. Finally, in accordance with the literature, it was found that the OFC and vmPFC regions play vital roles in perception (i.e., perceived value) [ 158 ].

5. Conclusions and Implications

Implication of the research findings for theory and practice: Theoretically, the current findings can be divided into three areas, as follows: firstly, neuroscientific techniques and methods enable the capture/measurement of the activity signals of the brain and body relating to consumers’ responses (e.g., emotions and feelings, attention, memory, perception, reward processing, and motivation) toward advertising campaigns. For example, neuroimaging tools (e.g., fMRI, EEG/ERP, fNIRS) enable the recording of the neural signals of the mental responses (e.g., pleasure/displeasure, low/high arousal, advertising recall and recognition) toward advertising, which can be beneficial for advertisers and marketers in creating more effective advertising campaigns to attract, captivate, and impact consumers. Meanwhile, physiological tools (e.g., ET, GSR, EMG, and ECG) enable researchers to gauge the physiological reactions of the consumer, such as pupil dilation, fixation, eye movements, saccade, heart rate, blood pressure, sweating level, and reaction time toward advertising. Secondly, neuroimaging and physiological tools will help advertisers and scholars to identify the weak elements in advertising which lead to withdrawal behaviour and to address these, besides identifying the strengths which lead to approach behaviour, and to enhance them. Thirdly, many articles have concentrated on detecting the neural and physiological responses of emotions, feelings, attention, memory, reward processing, motivation and perception toward advertising such as the presenter’s features (i.e., celebrity), because these processes play a key role in the decision-making of consumers (i.e., purchasing decisions). Additionally, some advertising research concentrated on gender voice (i.e., male, female), ads appeal, faces of celebrity, social campaigns (i.e., using seat belts in the car), and public health (i.e., anti-smoking campaigns). These areas can provide a reasonable explanation of the neural and physiological correlates of emotions and feelings (e.g., pleasure/displeasure, high/low arousal), attention (e.g., top-down, bottom-up), memory (e.g., encoding, retrieving), motivation (e.g., approach/withdrawal), reward processing, and perception (e.g., perceived value of ads) to be considered in advertising research. Thus, an application of this research perhaps offers reasonable explanations of how advertising works in consumers’ minds, and the relationship between the neural correlates of consumers’ brain and physiological responses toward advertising, thereby better understanding consumers’ behaviour, which leads to the creation of more attractive advertising for political, social and business sectors.

General Conclusion : Neuromarketing is a promising field, not only to study and solve the commercial issues such as the weaknesses of advertising campaigns and to reduce the wastage of advertising budgets, but also to create more effective advertising campaigns in social, political, and public-health sectors, in order to increase human awareness. In today’s hyper-competitive environments among advertising agencies, each agency seeks to find the most beneficial methods to beat competitors and be the first priorities in the consumer’s mind. Thus, advertisers have adopted neuroscientific methods in their research to study, analyse, and predict the neural and physiological responses of consumers toward the stimuli of marketing (i.e., advertising), thereby identifying the most important mental and physiological responses to be considered in advertising research to raise advertising effectiveness. Most studies in advertising research have determined the main mental processes of interest for advertising research, such as emotions and feelings, attention, memory, reward processing, motivation, and perception.

The findings of the study suggest that neuroscientific methods and techniques are significant to gauge the brain and physiological reactions of consumers toward the stimuli of marketing, such as advertising research. For example, neuroimaging tools are able to gauge the neural-activity signals of the consumer’s brain. At the same time, physiological tools can gauge physiological reactions such as eye movements, sweating level, and fixation.

6. Limitations and Future Directions

This paper tried to minimize the limitations in methodology; however, some limitations occurred and provided several directions for further research. This research concentrated on the English articles that were published in open-access journals from 2009 to 2020 and were listed in the WOS database. Therefore, this paper overlooked non-English articles, non-open-access articles, and other documents such as books, review papers, conference proceedings, and so forth. Thus, this paper is not free of bias. For future directions, we hope to overcome the obstacles in the future, which include the cost of research and techniques, lack of labs and facilities, use of time (e.g., data interpretation, recruiting participants, and so forth), and increased investment and funding in neuromarketing research and technique [ 177 ]. We encourage researchers and marketers from emerging countries to enter this embryonic field and leave their footprint by publishing articles for future works. In addition, we suggest that researchers and scholars identify the influence of advertising on consumers persuasion, engagement, and excitement, as well as the contributions of neuromarketing research to various domains (e.g., social sciences, public health, politics, and stock exchanges). We believe that this review study provides a profound overview of the global academic-trends in neuromarketing research, using the neuroimaging and physiological studies in advertising to study the brain processes of consumers. Thus, it provides valuable and reliable insights into the appropriate brain processes to be considered in future research.

Acknowledgments

The authors would like to thank Universiti Teknologi Malaysia (UTM), Azman Hashim International Business School (AHIBS); Taif University, Department of Economic & Finance, College of Business Administration; and Applied Science Private University (ASU), Department of Marketing for supporting this study.

Funding Statement

This research received no external funding.

Author Contributions

A.H.A., conceptualization, methodology, writing—original draft preparation, result discussion, and data curation; N.Z.M.S., supervision, review and editing; S.A.A.-Z., review, editing, and methodology; A.K., review and results discussion. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Informed consent statement, data availability statement, conflicts of interest.

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

• Research Priorities
• Advanced search

Advanced Search

Volume 64, Issue 2, 2024

• Tips for Advertising Professionals Submitting to JAR Colin Campbell
• The Impact of Advertising on a Company’s Stock Price Shuba Srinivasan and Dominique M. Hanssens
• Searching for the Holy Grail of Sponsorship-Linked Marketing Jonathan A. Jensen
• Ignite the Scent—The Effectiveness of Implied Explosion in Perfume Ads Olivier Droulers, Junwei (Jerry) Yu, and Sophie Lacoste-Badie
• Why Shorter Advertisement Breaks Reduce Radio Advertisement Avoidance Aaron Michelon, Steven Bellman, Margaret Faulkner, Justin Cohen, and Johan Bruwer
• Musical Manipulation of Visual Scenes in Video, Film, and TV Advertisements Sebastian Silas, David John Baker, and Daniel Müllensiefen

SPECIAL ISSUE CONTINUATION: PROSOCIAL ADVERTISING MESSAGES

• Why a Single Pro-Environmental Appeal Works to Promote Behavioral Change Matthew Pittman, Tyler Milfeld, and Kibum Youn
• How Virtual Reality Can Increase Effectiveness of Prosocial Advertising Ou Li, Zhigeng Pan, Han Qiu, and Da Qian

View full current issue

LATEST RESEARCH

New Digital First articles  added on AR and VR.

FREE BEST PAPERS

JAR Best Paper winners are open access.

Featured Articles

• When to Combine Television With Online Campaigns September 05, 2017
• Examining Consumers' Multiplatform Usage And Its Contribution to Their Trust in Advertising February 24, 2017
• What Makes Online Promotional Games Go Viral? June 06, 2017
• Measuring Television In the Programmatic Age March 07, 2017
• The Market Value Of Celebrity Endorsement October 27, 2016
• How Direct-to-Consumer Advertising for Prescription Drugs Affects Consumers' Welfare November 23, 2016
• Consumer Response To Gay and Lesbian Imagery December 05, 2016

DIGITAL FIRST: LATEST RESEARCH FROM JAR

• You have access Are Virtual Influencers More Effective In Real-World Advertisement Settings? Jeongmin Ham , Sitan Li , Pratik Shah , Matthew S. Eastin Journal of Advertising Research Jun 2024, PAP_015; DOI: 10.2501/JAR-2024-015
• You have access How Virtual Mirrors in Advertising Influence Attitudes about Beauty Products Hui-Fei Lin , Benjamin Yeo , Tsai-Yi Lu Journal of Advertising Research Jun 2024, PAP_016; DOI: 10.2501/JAR-2024-016
• You have access Innovations in Marketing Analytics Journal of Advertising Research Mar 2017, JAR-2017-011; DOI: 10.2501/JAR-2017-013
• Do Billboard Advertisements Drive Customer Retention?
• Effectiveness and Efficiency of TV’s Brand-Building Power: A Historical Review
• Why the Experiential View Is Vital To Marketing Communications Research Now
• When Brands Go Dark: A Replication and Extension
• What Do We Know About Celebrity Endorsement in Advertising?

• Election 2024
• Entertainment
• Newsletters
• Photography
• Personal Finance
• AP Investigations
• AP Buyline Personal Finance
• AP Buyline Shopping
• Press Releases
• Israel-Hamas War
• Russia-Ukraine War
• Global elections
• Asia Pacific
• Latin America
• Middle East
• Election Results
• Delegate Tracker
• AP & Elections
• Auto Racing
• 2024 Paris Olympic Games
• Movie reviews
• Book reviews
• Financial Markets
• Business Highlights
• Financial wellness
• Artificial Intelligence
• Social Media

Hillary Clinton to release essay collection about personal and public life

This cover image released by Simon & Schuster shows “Something Lost, Something Gained: Reflections on Life, Love and Liberty” by Hillary Rodham Clinton. The book will be released Sept. 17. (Simon & Schuster via AP)

• Copy Link copied

Hillary Clinton’s next book is a collection of essays, touching upon everything from marriage to politics to faith, that her publisher is calling her most personal yet.

Simon and Schuster announced Tuesday that Clinton’s “Something Lost, Something Gained: Reflections on Life, Love and Liberty” will be released Sept. 17.

Among the topics she will cover: Her marriage to former President Bill Clinton, her Methodist faith, adjusting to private life after her failed presidential runs, her friendships with other first ladies and her takes on climate change, democracy and Vladimir Putin.

“The book reads like you’re sitting down with your smartest, funniest, most passionate friend over a long meal,” Clinton’s editor, Priscilla Painton, said in a statement.

“This is the Hillary Americans have come to know and love: candid, engaged, humorous, self-deprecating — and always learning.”

Clinton, the former first lady, U.S. senator and secretary and presidential candidate, will promote her book with a cross country tour. “Something Lost, Something Gained” comes out two months before Bill Clinton’s memoir about post-presidential life, “Citizen.”

Financial terms were not disclosed. Clinton was represented by Washington attorney Robert Barnett, whose other clients have included former President George W. Bush and former President Barack Obama.

Clinton’s previous books include such bestsellers as “It Takes a Village,” “Living History” and “What Happened.”

• Open access
• Published: 18 June 2024

Using GPT-4 to write a scientific review article: a pilot evaluation study

• Zhiping Paul Wang 1 ,
• Priyanka Bhandary 1 ,
• Yizhou Wang 1 &
• Jason H. Moore 1  

BioData Mining volume  17 , Article number:  16 ( 2024 ) Cite this article

508 Accesses

5 Altmetric

Metrics details

GPT-4, as the most advanced version of OpenAI’s large language models, has attracted widespread attention, rapidly becoming an indispensable AI tool across various areas. This includes its exploration by scientists for diverse applications. Our study focused on assessing GPT-4’s capabilities in generating text, tables, and diagrams for biomedical review papers. We also assessed the consistency in text generation by GPT-4, along with potential plagiarism issues when employing this model for the composition of scientific review papers. Based on the results, we suggest the development of enhanced functionalities in ChatGPT, aiming to meet the needs of the scientific community more effectively. This includes enhancements in uploaded document processing for reference materials, a deeper grasp of intricate biomedical concepts, more precise and efficient information distillation for table generation, and a further refined model specifically tailored for scientific diagram creation.

Peer Review reports

Introduction

A comprehensive review of a research field can significantly aid researchers in quickly grasping the nuances of a specific domain, leading to well-informed research strategies, efficient resource utilization, and enhanced productivity. However, the process of writing such reviews is intricate, involving multiple time-intensive steps. These include the collection of relevant papers and materials, the distillation of key points from potentially hundreds or even thousands of sources into a cohesive overview, the synthesis of this information into a meaningful and impactful knowledge framework, and the illumination of potential future research directions within the domain. Given the breadth and depth of biomedical research—one of the most expansive and dynamic fields—crafting a literature review in this area can be particularly challenging and time-consuming, often requiring months of dedicated effort from domain experts to sift through the extensive body of work and produce a valuable review paper [ 1 , 2 ].

The swift progress in Natural Language Processing (NLP) technology, particularly with the rise of Generative Pre-trained Transformers (GPT) and other Large Language Models (LLMs), has equipped researchers with a potent tool for swiftly processing extensive literature. A recent survey indicates that ChatGPT has become an asset for researchers across various fields [ 3 ]. For instance, a PubMed search for “ChatGPT” yielded over 1,400 articles with ChatGPT in their titles as of November 30th, 2023, marking a significant uptake just one year after ChatGPT’s introduction.

The exploration of NLP technology’s capability to synthesize scientific publications into comprehensive reviews is ongoing. The interest in ChatGPT’s application across scientific domains is evident. Studies have evaluated ChatGPT’s potential in clinical and academic writing [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 ], and discussions are underway about its use as a scientific review article generator [ 11 , 12 , 13 ]. However, many of these studies predate the release of the more advanced GPT-4, which may render their findings outdated. In addition, there is no study specifically evaluating ChatGPT (GPT-4) for writing biomedical review papers.

As the applications of ChatGPT are explored, the scientific community is also examining the evolving role of AI in research. Unlike any tool previously utilized in the history of science, ChatGPT has been accorded a role akin to that of a scientist, even being credited as an author in scholarly articles [ 14 ]. This development has sparked ethical debates. While thorough evaluations of the quality of AI-generated scientific review articles are yet to be conducted, some AI tools, such as Scopus AI [ 15 ], are already being employed to summarize and synthesize knowledge from scientific literature databases. However, these tools often come with disclaimers cautioning users about the possibility of AI generating erroneous or offensive content. Concurrently, as ChatGPT’s potential contributions to science are probed, concerns about the possible detrimental effects of ChatGPT and other AI tools on scientific integrity have been raised [ 16 ]. These considerations highlight the necessity for more comprehensive evaluations of ChatGPT from various perspectives.

In this study, we hypothesized that ChatGPT can compose text, tables and figures for a biomedical research paper using two cancer research papers as benchmarks. To test this hypothesis, we used the first paper [ 17 ] to prompt ChatGPT to generate the main ideas and summarize text. Next, we used the second paper [ 18 ] to assess its ability to create tables and figures/graphs. We simulated the steps a scientist would take in writing a cancer research review and assessed GPT-4’s performance at each stage. Our findings are presented across four dimensions: the ability to summarize insights from reference papers on specific topics, the semantic similarity of GPT-4 generated text to benchmark texts, the projection of future research directions based on current publications, and the synthesis of context in the form of tables and graphs. We conclude with a discussion of our overall experience and the insights gained from this study.

Review text content generation by ChatGPT

The design of this study aims to replicate the process a scientist undergoes when composing a biomedical review paper. This involves the meticulous collection, examination, and organization of pertinent references, followed by the articulation of key topics of interest into a structured format of sections, subsections, and main points. The scientist then synthesizes information from the relevant references to develop a comprehensive narrative. A primary objective of this study is to assess ChatGPT’s proficiency in distilling insights from references into coherent text. To this end, a review paper on sex differences in cancer [ 17 ] was chosen as a benchmark, referred to as BRP1 (Benchmark Review Paper 1). Using BRP1 for comparison, ChatGPT’s content generation was evaluated across three dimensions: (1) summarization of main points; (2) generation of review content for each main point; and (3) synthesis of information from references to project future research directions.

Main point summarization

The effectiveness of GPT-4 in summarizing information was tested by providing it with the 113 reference articles from BRP1 to generate a list of potential sections for a review paper. The generated sections were then compared with BRP1’s actual section titles for coverage evaluation (Fig.  1 (A)). Additionally, GPT-4 was tasked with creating possible subsections using the BRP1 section titles and reference articles, which were compared with the actual subsection titles in BRP1.

Review content generation

The review content generation test involved comparing GPT-4’s ability to summarize a given point with the actual text content from BRP1 (Fig.  1 (B)). BRP1 comprises three sections with seven subsections, presenting a total of eight main points. The corresponding text content for each point was manually extracted from BRP1. Three strategies were employed for GPT-4 to generate detailed elaborations for these main points: (1) providing a point only in a prompt for baseline content generation; (2) feeding all references used by BRP1 to GPT-4 for reference-based content generation; (3) using only the references corresponding to a main point, i.e., articles being referred in a subsection of BRP1, for content generation to make a main point. The semantic similarity of the text content generated by these strategies was then compared with the manually extracted content from BRP1.

( A ) GPT-4 summarizes sections and subsections; ( B ) GPT-4 generated review content evaluation

Projections on future research

The section on “outstanding questions” in the Concluding Remarks of BRP1 serves a dual purpose: it summarizes conclusions and sets a trajectory for future research into sex differences in cancer. This is a common feature in biomedical review papers, where a forward-looking analysis is synthesized from the main discussions within the paper. The pivotal inquiry is whether ChatGPT, without further refinement, can emulate this forward projection using all referenced articles. The relevance of such a projection is contingent upon its alignment with the main points and references of the review. Moreover, it raises the question of whether the baseline GPT-4 LLM would perform comparably.

To address these queries, all references from BRP1 were inputted into GPT-4 to generate a section akin to Concluding Remarks, encompassing a description of sex differences in cancer, future work, and potential research trajectories. Additionally, three distinct strategies were employed to assess GPT-4’s ability to formulate specific “outstanding questions,” thereby evaluating ChatGPT’s predictive capabilities for future research. These strategies involved uploading all BRP1 reference articles to GPT-4 for projection: (1) without any contextual information; (2) with the inclusion of BRP1’s main points; (3) with a brief description of broad areas of interest. The outputs from these strategies, along with the base model’s output—GPT-4 without reference articles—were juxtaposed with BRP1’s original “outstanding questions” for comparison.

Data process

Chatgpt query.

In initiating this study, we utilized the ChatGPT web application ( https://chat.openai.com/ ). However, we encountered several limitations that impeded our progress:

A cap of ten file uploads, which restricts the analysis of content synthesized from over ten articles.

A file size limit of 50 MB, hindering the consolidation of multiple articles into a single file to circumvent the upload constraint.

Inconsistencies in text file interpretation when converted from PDF format, rendering the conversion of large PDFs to smaller text files ineffective.

Anomalies in file scanning, where ChatGPT would occasionally process only one of several uploaded files, despite instructions to utilize all provided files.

Due to these constraints, we transitioned to using GPT-4 API calls for all tests involving document processing. The GPT-4 API accommodates up to twenty file uploads simultaneously, efficiently processes text files converted from PDFs, and demonstrates reliable file scanning for multiple documents. The Python code, ChatGPT prompts, and outputs pertinent to this study are available in the supplementary materials.

The web version of ChatGPT cannot read from all the PDFs uploaded and is able to process only a subset of them. However, the API version of ChatGPT was set up to be able to upload and process 20 PDFs at a time. Several validation tests were carried out to make sure that it is able to read from all of them equally well. One common validation test was to ask ChatGPT if it could reiterate the Methods section of the 18th PDF and so on. This test was carried out randomly multiple times with a different PDF each time to see if ChatGPT is truly able to upload and process the PDFs.

Text similarity comparison

To assess text content similarity, we employed a transformer network-based pre-trained model [ 19 ] to calculate the semantic similarity between the original text in BRP1 and the text generated by GPT-4. We utilized the util.pytorch_cos_sim function from the sentence_transformers package to compute the cosine similarity of semantic content. Additionally, we conducted a manual validation where one of the authors compared the two texts and then categorized the similarity between the GPT-4 generated content and the original BRP1 content into three distinct levels: semantically very similar (Y), partially similar (P), and not similar (N).

Reproducibility and plagiarism evaluation

The inherent randomness in ChatGPT’s output, attributable to the probabilistic nature of large language models (LLMs), necessitates the validation of reproducibility for results derived from ChatGPT outputs. To obtain relatively consistent responses from ChatGPT, it is advantageous to provide detailed context within the prompt, thereby guiding the model towards the desired response. Consequently, we replicated two review content generation tests, as depicted in Fig.  1 (B)—one based on point references and the other on the GPT-4 base model—one week apart using identical reference articles and prompts via API calls to GPT-4. The first test aimed to evaluate the consistency of file-based content generation by GPT-4, while the second assessed the base model. We compared the outputs from the subsequent run with those from the initial run to determine the reproducibility of the text content generated by ChatGPT.

Prior to considering the utilization of ChatGPT for generating content suitable for publication in a review paper, it is critical to address potential plagiarism concerns. The pivotal question is whether text produced by GPT-4 would be flagged as plagiarized by anti-plagiarism software. In this study, GPT-4 generated a substantial volume of text, particularly for the text content comparison test (Fig.  1 (B)). We subjected both the base model-generated review content and the reference-based GPT-4 review content to scrutiny using iThenticate to ascertain the presence of plagiarism.

Table and figure generation by ChatGPT

Review papers often distill the content from references into tables and further synthesize this information into figures. In this study, we evaluated ChatGPT’s proficiency in generating content in tabular and diagrammatic formats, using benchmark review paper 2 (BRP2) [ 18 ] as a reference, as illustrated in Fig.  2 . The authors of BRP2 developed the seminal Cancer-Immunity Cycle concept, encapsulated in a cycle diagram, which has since become a structural foundational for research in cancer immunotherapy.

Table content generation

Analogous to the file scan anomaly, ChatGPT may disproportionately prioritize one task over others when presented with multiple tasks simultaneously. To mitigate this in the table generation test, we adopted a divide-and-conquer approach, submitting separate GPT-4 prompts to generate content for each column of the table. This strategy facilitated the straightforward assembly of the individual outputs into a comprehensive table, either through GPT-4 or manual compilation.

In BRP2, eleven reference articles were utilized to construct a table (specifically, Table  1 of BRP2) that categorized positive and negative regulators at each stage of the Cancer-Immunity Cycle. These articles were compiled and inputted into ChatGPT, prompting GPT-4 to summarize information for corresponding table columns: Steps, Stimulators, Inhibitors, Other Considerations, and Example References. The content for each column was generated through separate GPT-4 API calls and subsequently compared manually with the content in the original BRP2 table. The semantic similarity and manual validations were carried out for each row of the Table  1 from BRP2. With the API version, we uploaded the references cited within the corresponding row in the table and used that to generate the contents of the row.

Diagram creation

ChatGPT is primarily designed for text handling, yet its capabilities in graph generation are increasingly being explored [ 20 ]. DALL-E, the model utilized by ChatGPT for diagram creation, has been trained on a diverse array of images, encompassing various subjects, styles, contexts, and including scientific and technical imagery. To direct ChatGPT towards producing a diagram that closely aligns with the intended visualization, a precise and succinct description of the diagram is essential. Like the approach for table generation, multiple prompts may be required to facilitate incremental revisions in the drawing process.

In this evaluation, we implemented three distinct strategies for diagram generation, as demonstrated in Fig.  2 . Initially, the 11 reference articles used for table generation were also employed by GPT-4 to generate a description for the cancer immunity cycle, followed by the creation of a diagrammatic representation of the cycle by GPT-4. This approach not only tested the information synthesis capability of GPT-4 but also its diagram drawing proficiency. Secondly, we extracted the paragraph under the section titled ‘The Cancer-Immunity Cycle’ from BRP2 to serve as the diagram description. Terms indicative of a cyclical structure, such as ‘cycle’ and ‘step 1 again,’ were omitted from the description prior to its use as a prompt for diagram drawing. This tested GPT-4’s ability to synthesize the provided information into an innovative cyclical structure for cancer immunotherapy. Lastly, the GPT-4 base model was tasked with generating a cancer immunity mechanism and its diagrammatic representation without any given context. The diagrams produced through these three strategies were scrutinized and compared with the original cancer immunity cycle figure in BRP2 to assess the scientific diagram drawing capabilities of GPT-4.

GPT-4 table generation and figure creation

Results and discussions

Main point summary.

As depicted in Fig.  1 A, GPT-4 generated nine potential sections for a proposed paper entitled ‘The Spectrum of Sex Differences in Cancer,’ utilizing the 113 reference articles uploaded, which encompassed all three sections in BRP1. Upon request to generate possible subsections using BRP1 section titles and references, GPT-4 produced four subsections for each section, totaling twelve subsections that encompassed all seven subsections in BRP1. Detailed information regarding GPT-4 prompts, outputs, and comparisons with BRP1 section and subsection titles is provided in the supplementary materials.

The results suggest that ChatGPT can effectively summarize the key points from a comprehensive list of documents, which is particularly beneficial when composing a review paper that references hundreds of articles. With ChatGPT’s assistance, authors can swiftly summarize a list of main topics for further refinement, organization, and editing. Once the topics are finalized, GPT-4 can easily summarize different aspects for each topic, aiding authors in organizing the subsections. This indicates a novel approach to review paper composition that could be more efficient and productive than traditional methods. It represents a collaborative effort between ChatGPT and the review writer, with ChatGPT sorting and summarizing articles, and the author conducting high-level and creative analysis and editing.

During this evaluation, one limitation of GPT-4 was identified: its inability to provide an accurate list of articles referenced for point generation. This presents a challenge in developing an automated pipeline that enables both information summarization and file classification.

Figure  3 illustrates a sample of the text content generation, including the original BRP1 text, the prompt, and ChatGPT’s output. The evaluation results for GPT-4’s review content generation are presented in Table  1 (refer to Fig.  1 B). When generating review content using corresponding references as in BRP1, GPT-4 achieved an average similarity score of 0.748 with the original content in BRP1 across all main points. Manual similarity validation confirmed that GPT-4 generated content that was semantically similar for all 8 points, with 6 points matching very well (Y) and 2 points matching partially (P). When utilizing all reference articles for GPT-4 to generate review content for a point, the mean similarity score was slightly lower at 0.699, with a manual validation result of 5Y3P. The results from the GPT-4 based model were comparable to the corresponding reference-based results, with a mean similarity score of 0.755 and a 6Y2P manual validation outcome.

Text generation using GPT4 with specific references ( A ) Original section in BRP1 ( B ) Prompt for same section ( C ) Response from GPT4

As the GPT-4 base model has been trained on an extensive corpus of scientific literature, including journals and articles that explore sex differences in cancer, it is plausible for it to generate text content similar to the original review paper, even for a defined point without any contextual input. The performance when using corresponding references is notably better than when using all references, suggesting that GPT-4 processes information more effectively with relevant and less noisy input.

The similarity score represents only the level of semantic similarity between the GPT-4 output and the original review paper text. It should not be construed as a measure of the quality of the text content generated by GPT-4. While it is relatively straightforward to assess the relevance of content for a point, gauging comprehensiveness is nearly impossible without a gold standard. However, scientific review papers are often required in research areas where such standards do not yet exist. Consequently, this review content similarity test merely indicates whether GPT-4 can produce text content that is semantically akin to that of a human scholar. Based on the results presented in Table  1 , GPT-4 has demonstrated adequate capability in this regard.

In this evaluation, GPT-4 initially synthesized content analogous to the Concluding Remarks section of BRP1 by utilizing all reference articles, further assessing its capability to integrate information into coherent conclusions. Subsequently, GPT-4 projected future research directions using three distinct methodologies. The findings, as detailed in Table  2 , reveal that GPT-4’s content generation performance significantly increased from 0.45 to 0.71 upon the integration of all pertinent references, indicating that the provision of relevant information markedly enhances the model’s guidance. Consequently, although GPT-4 may face challenges in precisely replicating future research due to thematic discrepancies, equipping it with a distinct theme can empower it to produce content that more accurately represents the intended research trajectory. In contrast, the performance of the GPT-4 base model remained comparably stable, regardless of additional contextual cues. Manual verification confirmed GPT-4’s ability to synthesize information from the provided documents and to make reasonably accurate predictions about future research trajectories.

Reproducibility

The comparative analysis of GPT-4 outputs from different runs is presented in Table  3 . Based on previous similarity assessments, a similarity score of 0.7 is generally indicative of a strong semantic correlation in the context of this review paper. In this instance, GPT-4 outputs using corresponding references exhibited an average similarity score of 0.8 between two runs, while the base model scored 0.9. A manual review confirmed that both outputs expressed the same semantic meaning at different times. Consequently, it can be concluded that GPT-4 consistently generates uniform text responses when provided with identical prompts and reference materials.

An intriguing observation is that the GPT-4 base model appears to be more stable than when utilizing uploaded documents. This may suggest limitations in GPT-4’s ability to process external documents, particularly those that are unstructured or highly specialized in scientific content. This limitation aligns with our previous observation regarding GPT-4’s deficiency in cataloging citations within its content summaries.

Plagiarism check

The plagiarism assessment conducted via iThenticate ( https://www.ithenticate.com/ ) yielded a percentage score of 34% for reference-based GPT-4 content generation and 10% for the base model. Of these percentages, only 2% and 3%, respectively, were attributed to matches with the original review paper (BRP1), predominantly due to title similarities, as we maintained the same section and subsection titles. A score of 34% is typically indicative of significant plagiarism concerns, whereas 10% is considered minimal. These results demonstrate the GPT-4 base model’s capacity to expound upon designated points in a novel manner, minimally influenced by the original paper. However, the reference-based content generation raised concerns due to a couple of instances of ‘copy-paste’ style matches from two paragraphs in BRP1 references [ 21 , 22 ], which contributed to the elevated 34% score. In summary, while the overall content generated by ChatGPT appears to be novel, the occurrence of sporadic close matches warrants scrutiny.

This finding aligns with the theoretical low risk of direct plagiarism by ChatGPT, as AI-generated text responses are based on learned patterns and information, rather than direct ‘copy-paste’ from specific sources. Nonetheless, the potential for plagiarism and related academic integrity issues are of serious concern in academia. Researchers have been exploring appropriate methods to disclose ChatGPT’s contributions in publications and strategies to detect AI-generated content [ 23 , 24 , 25 ].

Table construction in scientific publications often necessitates a more succinct representation of relationships and key terms compared to text content summarization and synthesis. This requires ChatGPT to extract information with greater precision. For the five columns of information compiled by GPT-4 for Table  1 in BRP2, the Steps column is akin to summarizing section and subsection titles in BRP1. ‘ Stimulators’ and ‘ Inhibitors’ involve listing immune regulation factors, demanding more concise and precise information extraction. ‘ Other Considerations ’ encompasses additional relevant information, while ‘ Example References ’ lists citations.

For the Steps column, GPT-4 partially succeeded but struggled to accurately summarize information into numbered steps. For the remaining columns, GPT-4 was unable to extract the corresponding information accurately. Extracting concise and precise information from uploaded documents for specific scientific categories remains a significant challenge for GPT-4, which also lacks the ability to provide reference citations, as observed in previous tests. All results, including GPT prompts, outputs, and evaluations, are detailed in the supplementary materials.

In summary, GPT-4 has not yet achieved the capability to generate table content with the necessary conciseness and accuracy for information summary and synthesis.

Figure creation

In the diagram drawing test, we removed all terms indicative of a cyclical graph from the diagram description in the prompt to evaluate whether GPT-4 could independently recreate the original, pioneering depiction of the cancer immune system cycle. We employed three strategies for diagram generation, as depicted in Fig.  2 , which included: (1) using a diagram description generated from references and incorporated into the drawing prompt; (2) using the description from BRP2; (3) relying on the GPT-4 base model. The resulting diagrams produced by GPT-4 are presented in Fig.  4 , with detailed information provided in the supplementary materials.

( A ) Original figure ( B ) reference description ( C ) BRP2 description ( D ) base model

These diagrams highlight common inaccuracies in GPT-4’s drawings, such as misspelled words, omitted numbers, and a lack of visual clarity due to superfluous icons and cluttered labeling. Despite these issues, GPT-4 demonstrated remarkable proficiency in constructing an accurate cycle architecture, even without explicit instructions to do so.

In conclusion, while GPT-4 can serve as a valuable tool for conceptualizing diagrams for various biomedical reactions, mechanisms, or systems, professional graph drawing tools are essential for the actual creation of diagrams.

Conclusions

In this study, we evaluated the capabilities of the language model GPT-4 within ChatGPT for composing a biomedical review article. We focused on four key areas: (1) summarizing insights from reference papers; (2) generating text content based on these insights; (3) suggesting avenues for future research; and (4) creating tables and graphs. GPT-4 exhibited commendable performance in the first three tasks but was unable to fulfill the fourth.

ChatGPT’s design is centered around text generation, with its language model finely tuned for this purpose through extensive training on a wide array of sources, including scientific literature. Consequently, GPT-4’s proficiency in text summarization and synthesis is anticipated. When specifically comparing the API GPT model performance on a section providing specific references (references only limited to that section) and all references from the entire paper, the model does better when it is given specific references because providing all references could bring in a lot of noise. One more thing to note is that the prompt specifically mentions not to use external knowledge and hence it must process over a hundred publications and discover relevant information for the section and then compose a reply. This could explain why giving specific references improves performance over giving all references. Remarkably, the GPT-4 base model’s performance is on par with, or in some cases, slightly surpasses that of reference-based text content generation, owing to its training on a diverse collection of research articles and web text. Hence, when given a prompt and some basic points, it performs well since it already possesses all the information needed to generate an appropriate response. Furthermore, reproducibility tests have demonstrated GPT-4’s ability to generate consistent text content, whether utilizing references or solely relying on its base model.

In addition, we assessed GPT-4’s proficiency in extracting precise and pertinent information for the construction of research-related tables. GPT-4 encountered difficulties with this task, indicating that ChatGPT’s language model requires additional training to enhance its ability to discern and comprehend specialized scientific terminology from literature. This improvement necessitates addressing complex scientific concepts and integrating knowledge across various disciplines.

Moreover, GPT-4’s capability to produce scientific diagrams does not meet the standards required for publication. This shortfall may stem from its associated image generation module, DALL-E, being trained on a broad spectrum of images that encompass both scientific and general content. However, with ongoing updates and targeted retraining to include a greater volume of scientific imagery, the prospect of a more sophisticated language model with improved diagrammatic capabilities could be a foreseeable advancement.

To advance the assessment of ChatGPT’s utility in publishing biomedical review articles, we executed a plagiarism analysis on the text generated by GPT-4. This analysis revealed potential issues when references were employed, with GPT-4 occasionally producing outputs that closely resemble content from reference articles. Although GPT-4 predominantly generates original text, we advise conducting a plagiarism check on ChatGPT’s output before any formal dissemination. Moreover, despite the possibility that the original review paper BRP1 was part of GPT-4’s training dataset, the plagiarism evaluation suggests that the output does not unduly prioritize it, considering the extensive data corpus used for training the language model.

Our study also highlights the robust performance of the GPT-4 base model, which shows adeptness even without specific reference articles. This observation leads to the conjecture that incorporating the entirety of scientific literature into the training of a future ChatGPT language model could facilitate the on-demand extraction of review materials. Thus, it posits the potential for ChatGPT to eventually author comprehensive summary and synthesis-based scientific review articles. ChatGPT did not offer any citations for the PDFs that were provided to it at the time this work was written. Therefore, it is advised in such a situation to go section by section, supply a single paper, and obtain a summary of that publication alone so that the user can write a few phrases for that portion and properly credit the paper. On the other hand, the user can supply all articles for commonly recognized knowledge to produce a well-rounded set of statements that require a set of citations.

ChatGPT’s power and versatility warrant additional exploration of various facets. While these are beyond the scope of the current paper, we will highlight selected topics that are instrumental in fostering a more science oriented ChatGPT environment. Holistically speaking, to thoroughly assess ChatGPT’s proficiency in generating biomedical review papers, it is imperative to include a diverse range of review paper types in the evaluation process. For instance, ChatGPT is already equipped to devise data analysis strategies and perform data science tasks in real-time. This capability suggests potential for generating review papers that include performance comparisons and benchmarks of computational tools. However, this extends beyond the scope of our pilot study, which serves as a foundational step toward more extensive research endeavors.

Ideally, ChatGPT would conduct essential statistical analyses of uploaded documents, such as ranking insights, categorizing documents per insight, and assigning relevance weights to each document. This functionality would enable scientists to quickly synthesize the progression and extensively studied areas within a field. When it comes to mitigating hallucination, employing uploaded documents as reference material can reduce the occurrence of generating inaccurate or ‘hallucinated’ content. However, when queries exceed the scope of these documents, ChatGPT may still integrate its intrinsic knowledge base. In such cases, verifying ChatGPT’s responses against the documents’ content is vital. A feasible method is to cross-reference responses with the documents, although this may require significant manual effort. Alternatively, requesting ChatGPT to annotate its output with corresponding references from the documents could be explored, despite being a current limitation of GPT-4.

To address academic integrity concerns, as the development of LLMs progresses towards features that could potentially expedite or even automate the creation of scientific review papers, the establishment of a widely accepted ethical practice guide becomes paramount. Until such guidelines are in place, it remains essential to conduct plagiarism checks on AI-generated content and transparently disclose the extent of AI’s contribution to the published work. The advent of large language models like Google’s Gemini AI [ 26 ] and Perplexity.ai has showcased NLP capabilities comparable to those of GPT-4. This, coupled with the emergence of specialized models such as BioBert [ 27 ], BioBART [ 28 ], and BioGPT [ 29 ] for biomedical applications, highlights the imperative for in-depth comparative studies. These assessments are vital for identifying the optimal AI tool for particular tasks, taking into account aspects such as multimodal functionalities, domain-specific precision, and ethical considerations. Conducting such comparative analyses will not only aid users in making informed choices but also promote the ethical and efficacious application of these sophisticated AI technologies across diverse sectors, including healthcare and education.

Data availability

All source codes, GPT4 generated contents and results are available at the GitHub repository.

Dhillon P. How to write a good scientific review article. FEBS J. 2022;289:3592–602.

Article   CAS   PubMed   Google Scholar  

Health sciences added to the Nature Index. | News | Nature Index. https://www.nature.com/nature-index/news/health-sciences-added-to-nature-index .

Van Noorden R, Perkel JM. AI and science: what 1,600 researchers think. Nature. 2023;621:672–5.

Article   PubMed   Google Scholar  

Ariyaratne S, Iyengar KP, Nischal N, Chitti Babu N, Botchu R. A comparison of ChatGPT-generated articles with human-written articles. Skeletal Radiol. 2023;52:1755–8.

Kumar AH. Analysis of ChatGPT Tool to assess the potential of its utility for academic writing in Biomedical Domain. Biology Eng Med Sci Rep. 2023;9:24–30.

Article   Google Scholar  

Alkaissi H, McFarlane SI. Artificial Hallucinations in ChatGPT: implications in Scientific writing. Cureus 15, e35179.

Meyer JG, et al. ChatGPT and large language models in academia: opportunities and challenges. BioData Min. 2023;16:20.

Article   PubMed   PubMed Central   Google Scholar  

Mondal H, Mondal S. ChatGPT in academic writing: maximizing its benefits and minimizing the risks. Indian J Ophthalmol. 2023;71:3600.

Misra DP, Chandwar K. ChatGPT, artificial intelligence and scientific writing: what authors, peer reviewers and editors should know. J R Coll Physicians Edinb. 2023;53:90–3.

Ali SR, Dobbs TD, Hutchings HA, Whitaker IS. Using ChatGPT to write patient clinic letters. Lancet Digit Health. 2023;5:e179–81.

Alshami A, Elsayed M, Ali E, Eltoukhy AEE, Zayed T. Harnessing the power of ChatGPT for automating systematic review process: Methodology, Case Study, limitations, and future directions. Systems. 2023;11:351.

Huang J, Tan M. The role of ChatGPT in scientific communication: writing better scientific review articles. Am J Cancer Res. 2023;13:1148–54.

PubMed   PubMed Central   Google Scholar  

Haman M, Školník M. Using ChatGPT to conduct a literature review. Account Res. 2023;0:1–3.

ChatGPT listed as author on research papers. many scientists disapprove. https://www.nature.com/articles/d41586-023-00107-z .

Scopus AI. Trusted content. Powered by responsible AI. www.elsevier.com https://www.elsevier.com/products/scopus/scopus-ai .

Conroy G. How ChatGPT and other AI tools could disrupt scientific publishing. Nature. 2023;622:234–6.

Rubin JB. The spectrum of sex differences in cancer. Trends Cancer. 2022;8:303–15.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Chen DS, Mellman I. Oncology meets immunology: the Cancer-Immunity cycle. Immunity. 2013;39:1–10.

Reimers N, Gurevych I, Sentence-BERT. Sentence Embeddings using Siamese BERT-Networks. in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) (eds. Inui, K., Jiang, J., Ng, V. & Wan, X.) 3982–3992Association for Computational Linguistics, Hong Kong, China, (2019). https://doi.org/10.18653/v1/D19-1410 .

Jin B et al. Large Language Models on Graphs: A Comprehensive Survey. Preprint at https://doi.org/10.48550/arXiv.2312.02783 (2024).

Polkinghorn WR, et al. Androgen receptor signaling regulates DNA repair in prostate cancers. Cancer Discov. 2013;3:1245–53.

Broestl L, Rubin JB. Sexual differentiation specifies Cellular responses to DNA damage. Endocrinology. 2021;162:bqab192.

ChatGPT and Academic Integrity Concerns. Detecting Artificial Intelligence Generated Content | Language Education and Technology. https://www.langedutech.com/letjournal/index.php/let/article/view/49 .

Gao CA et al. Comparing scientific abstracts generated by ChatGPT to original abstracts using an artificial intelligence output detector, plagiarism detector, and blinded human reviewers. 2022.12.23.521610 Preprint at https://doi.org/10.1101/2022.12.23.521610 (2022).

Homolak J. In reply: we do not stand a ghost of a chance of detecting plagiarism with ChatGPT employed as a ghost author. Croat Med J. 2023;64:293–4.

Article   PubMed Central   Google Scholar  

Introducing Gemini. Google’s most capable AI model yet. https://blog.google/technology/ai/google-gemini-ai/#sundar-note .

Lee J, et al. BioBERT: a pre-trained biomedical language representation model for biomedical text mining. Bioinformatics. 2020;36:1234–40.

Yuan H et al. Association for Computational Linguistics, Dublin, Ireland,. BioBART: Pretraining and Evaluation of A Biomedical Generative Language Model. in Proceedings of the 21st Workshop on Biomedical Language Processing (eds. Demner-Fushman, D., Cohen, K. B., Ananiadou, S. & Tsujii, J.) 97–109 (2022). https://doi.org/10.18653/v1/2022.bionlp-1.9 .

Luo R, et al. BioGPT: generative pre-trained transformer for biomedical text generation and mining. Brief Bioinform. 2022;23:bbac409.

Download references

This research received no specific grant from any funding agency.

Author information

Authors and affiliations.

Department of Computational Biomedicine, Cedars Sinai Medical Center, 700 N. San Vicente Blvd, Pacific Design Center, Suite G-541, West Hollywood, CA, 90069, USA

Zhiping Paul Wang, Priyanka Bhandary, Yizhou Wang & Jason H. Moore

You can also search for this author in PubMed   Google Scholar

Contributions

Z. Wang authored the main manuscript text, while P. Bhandary conducted most of the tests and collected the results. Y. Wang and J. Moore contributed scientific suggestions and collaborated on the project. All authors reviewed the manuscript.

Corresponding author

Correspondence to Jason H. Moore .

Ethics declarations

Supplements.

All supplementary materials are available at https://github.com/EpistasisLab/GPT4_and_Review .

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Wang, Z.P., Bhandary, P., Wang, Y. et al. Using GPT-4 to write a scientific review article: a pilot evaluation study. BioData Mining 17 , 16 (2024). https://doi.org/10.1186/s13040-024-00371-3

Download citation

Received : 26 April 2024

Accepted : 11 June 2024

Published : 18 June 2024

DOI : https://doi.org/10.1186/s13040-024-00371-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

BioData Mining

ISSN: 1756-0381

• General enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Published: 09 November 2023

Observation of the Mars O 2 visible nightglow by the NOMAD spectrometer onboard the Trace Gas Orbiter

• J.-C. Gérard   ORCID: orcid.org/0000-0002-8565-8746 1 ,
• L. Soret   ORCID: orcid.org/0000-0003-4093-330X 1 ,
• I. R. Thomas   ORCID: orcid.org/0000-0003-3887-6668 2 ,
• B. Ristic   ORCID: orcid.org/0000-0002-9635-1125 2 ,
• Y. Willame 2 ,
• C. Depiesse 2 ,
• A. C. Vandaele 2 ,
• F. Daerden   ORCID: orcid.org/0000-0001-7433-1839 2 ,
• B. Hubert 1 ,
• J. P. Mason 3 ,
• M. R. Patel   ORCID: orcid.org/0000-0002-8223-3566 3 &
• M. A. López-Valverde 4  

Nature Astronomy volume  8 ,  pages 77–81 ( 2024 ) Cite this article

675 Accesses

1 Citations

605 Altmetric

Metrics details

• Atmospheric chemistry
• Atmospheric dynamics

On Mars, atomic oxygen controls the carbon dioxide radiative cooling of the upper atmosphere and the presence of an ozone layer near the poles. To remotely probe meridional transport of O atoms from the summer to the winter hemisphere and the descending flow in the winter polar regions, the O 2 Herzberg II atmospheric emission could be used as a proxy. This emission is quite weak on Earth’s nightside, but it is prominent in the Venus night airglow, and it has not previously been observed on Mars. Here we report the limb detection of the O 2 Herzberg II visible bands in the Mars nightglow with the NOMAD ultraviolet–visible spectrometer onboard the European Space Agency’s Trace Gas Orbiter. The emission layer reaches up to hundreds of kilorayleighs in the limb viewing geometry. It is mainly located between 40 km and 60 km at high latitudes during the winter season, consistent with three-body recombination of oxygen atoms. This O 2 nightglow should be observable from a Martian orbiter as well as from the Martian surface with the naked eye under clear sky conditions. These observations pave the way to future global observations of the Martian atmospheric circulation with simpler lower-cost instrumentation.

This is a preview of subscription content, access via your institution

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 12 digital issues and online access to articles

111,21 € per year

only 9,27 € per issue

Buy this article

• Purchase on Springer Link
• Instant access to full article PDF

Prices may be subject to local taxes which are calculated during checkout

Similar content being viewed by others

Detection of green line emission in the dayside atmosphere of Mars from NOMAD-TGO observations

Direct detection of atomic oxygen on the dayside and nightside of Venus

Far-ultraviolet aurora identified at comet 67P/Churyumov-Gerasimenko

Data availability.

The NOMAD-UVIS spectra can be downloaded from ESA’s SA archives at https://archives.esac.esa.int/psa/#!Table%20View/NOMAD=instrument (select UVIS from the list of instruments and ‘Level 3 Calibrated’ from the processing level). Observed limb intensities and model calculations supporting Fig. 4 are available from BIRA-IASB repository at https://doi.org/10.18758/71021084 or from the corresponding author upon reasonable request.

Barth, C. A. et al. Mariner 6 and 7 ultraviolet spectrometer experiment: upper atmosphere data. J. Geophys. Res. 76 , 2213–2227 (1971).

Article   ADS   Google Scholar  

Gérard, J. C. et al. Detection of green line emission in the dayside atmosphere of Mars from NOMAD-TGO observations. Nat. Astron. 4 , 1049–1052 (2020).

Gérard, J. C. et al. First observation of the oxygen 630 nm emission in the Martian dayglow. Geophys. Res. Lett. 48 , e2020GL092334 (2021).

Aoki, S. et al. Density and temperature of the upper mesosphere and lower thermosphere of Mars retrieved from the O  i 557.7 nm dayglow measured by TGO/NOMAD. J. Geophys. Res. 127 , e2022JE007206 (2022).

Soret, L. et al. The Mars oxygen visible dayglow: a Martian year of NOMAD/UVIS observations. J. Geophys. Res. 127 , e2022JE007220 (2022).

Schneider, N. M. et al. Imaging of Martian circulation patterns and atmospheric tides through MAVEN/IUVS nightglow observations. J. Geophys. Res. 125 , e2019JA027318 (2020).

Bertaux, J. L., Gondet, B., Lefèvre, F., Bibring, J. P. & Montmessin, F. First detection of O 2 1.27 μm nightglow emission at Mars with OMEGA/MEX and comparison with general circulation model predictions. J. Geophys. Res. 117 , E00J04 (2012).

Fedorova, A. et al. The O 2 nightglow in the martian atmosphere by SPICAM onboard of Mars-Express. Icarus 219 , 596–608 (2012).

Clancy, R. T. et al. Extensive MRO CRISM observations of 1.27 μm O 2 airglow in Mars polar night and their comparison to MRO MCS temperature profiles and LMD GCM simulations. J. Geophys. Res. 117 , E00J10 (2012).

Article   Google Scholar  

Clancy, R. T. et al. First detection of Mars atmospheric hydroxyl: CRISM near-IR measurement versus LMD GCM simulation of OH Meinel band emission in the Mars polar winter atmosphere. Icarus 226 , 272–281 (2013).

Patel, M. R. et al. NOMAD spectrometer on the ExoMars trace gas orbiter mission: part 2—design, manufacturing, and testing of the ultraviolet and visible channel. Appl. Opt. 56 , 2771–2782 (2017).

Vandaele, A. C. et al. An integrated suite of three spectrometers for the ExoMars trace gas mission: technical description, science objectives and expected performance. Space Sci. Rev. 214 , 1–47 (2018).

López-Valverde et al. Investigations of the Mars upper atmosphere with ExoMars Trace Gas Orbiter. Space Sci. Rev. 214 , 29 (2018).

Krupenie, P. H. The spectrum of molecular oxygen. J. Phys. Chem. Ref. Data 1 , 423–534 (1972).

Krasnopolsky, V. A. in Venus (eds Hunten D. M. et al.) 459–483 (Univ. Arizona Press, 1983).

Lawrence, G. M., Barth, C. A. & Argabright, V. Excitation of the Venus night airglow. Science 195 , 573–574 (1977).

Slanger, T. G. & Copeland, R. A. Energetic oxygen in the upper atmosphere and the laboratory. Chem. Rev. 103 , 4731–4766 (2003).

Bougher, S. W. & Borucki, W. J. Venus O 2 visible and IR nightglow: implications for lower thermosphere dynamics and chemistry. J. Geophys. Res. 99 , 3759–3776 (1994).

Krasnopolsky, V. A. Excitation of the oxygen nightglow on the terrestrial planets. Planet. Space Sci. 59 , 754–766 (2011).

Bougher, S. W., Hunten, D. M. & Roble, R. G. CO 2 cooling in terrestrial planet thermospheres. J. Geophys. Res. 99 , 14609–14622 (1994).

Montmessin, F. & Lefèvre, F. Transport-driven formation of a polar ozone layer on Mars. Nat. Geosci. 6 , 930–933 (2013).

Forget, F. et al. Improved general circulation models of the Martian atmosphere from the surface to above 80 km. J. Geophys. Res. 104 , 24155–24175 (1999).

Omholt, A. The Optical Aurora (Springer, 1971); https://doi.org/10.1007/978-3-642-46269-6

Lilensten, J. Prediction of blue, red and green aurorae at Mars. Planet. Space Sci. 115 , 48–56 (2015).

Willame, Y. et al. Calibration of the NOMAD-UVIS data. Planet. Space Sci. 218 , 105504 (2022).

García Muñoz, A., Mills, F. P., Slanger, T. G., Piccioni, G. & Drossart, P. Visible and near‐infrared nightglow of molecular oxygen in the atmosphere of Venus. J. Geophys. Res. 114 , E12002 (2009).

García Muñoz, A. et al. Limb imaging of the Venus O 2 visible nightglow with the Venus Monitoring Camera. Geophys. Res. Lett. 40 , 2539–2543 (2013).

Gérard, J. C., Soret, L., Migliorini, A. & Piccioni, G. Oxygen nightglow emissions of Venus: vertical distribution and collisional quenching. Icarus 223 , 602–608 (2013).

Migliorini, A. et al. The characteristics of the O 2 Herzberg II and Chamberlain bands observed with VIRTIS/Venus Express. Icarus 223 , 609–614 (2013).

Smith, G. P. & Robertson, R. Temperature dependence of oxygen atom recombination in nitrogen after ozone photolysis. Chem. Phys. Lett. 458 , 6–10 (2008).

Nair, H., Allen, M., Anbar, A. D., Yung, Y. L. & Clancy, R. T. A photochemical model of the Martian atmosphere. Icarus 111 , 124–150 (1994).

Millour, E. et al. The Mars Climate Database (Version 6.1). EPSC Abstr. 16 , EPSC2022-786 (2022).

Download references

Acknowledgements

B.H. is research associate of the Belgian Fund for Scientific Research (FNRS). ExoMars is a space mission of ESA and Roscosmos. The NOMAD experiment is led by the IASB-BIRA, assisted by co-principal-investigator teams from Spain (IAA-CSIC), Italy (INAF-IAPS) and the United Kingdom (The Open University). This project acknowledges funding from BELSPO, with the financial and contractual coordination by the ESA Prodex Office (PEA grant numbers 4000140863, 4000121493 and 4000129683). M.A.L.-V. was supported by grant number PGC2018-101836-B-100 (MCIU/AEI/FEDER, EU) and CEX2021-001131-S funded by MCIN/AEI/10.13039/501100011033. We also acknowledge support from the UK Space Agency through grant numbers ST/V002295/1, ST/V005332/1, ST/Y000234/1 and ST/X006549/1’. We thank the ESA TGO team and its project scientists H. Svedhem and C. Wilson for supporting these observations.

Author information

Authors and affiliations.

LPAP, STAR Institute, Université de Liège, Liège, Belgium

J.-C. Gérard, L. Soret & B. Hubert

Royal Belgian Institute for Space Aeronomy, Brussels, Belgium

I. R. Thomas, B. Ristic, Y. Willame, C. Depiesse, A. C. Vandaele & F. Daerden

School of Physical Sciences, The Open University, Milton Keynes, UK

J. P. Mason & M. R. Patel

Instituto de Astrofìsica de Andalucía/CSIC, Granada, Spain

M. A. López-Valverde

You can also search for this author in PubMed   Google Scholar

Contributions

J.-C.G. and L.S. conceived the study and wrote the paper. Y.W., C.D., I.R.T. and J.P.M. calibrated the UVIS data and prepared the datasets. Observation planning was managed by J.-C.G., L.S., B.R., and M.R.P. A.C.V. is the NOMAD principal investigator, M.R.P. is NOMAD co-principal investigator. All authors contributed to discussion and comments on the paper.

Corresponding author

Correspondence to J.-C. Gérard .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Astronomy thanks Antonio García Muñoz and Guillaume Gronoff for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Cite this article.

Gérard, JC., Soret, L., Thomas, I.R. et al. Observation of the Mars O 2 visible nightglow by the NOMAD spectrometer onboard the Trace Gas Orbiter. Nat Astron 8 , 77–81 (2024). https://doi.org/10.1038/s41550-023-02104-8

Download citation

Received : 27 March 2023

Accepted : 18 September 2023

Published : 09 November 2023

Issue Date : January 2024

DOI : https://doi.org/10.1038/s41550-023-02104-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Airglow and aurora in the martian atmosphere: contributions by the mars express and exomars tgo missions.

• Francisco González-Galindo
• Jean-Claude Gérard
• Franck Montmessin

Space Science Reviews (2024)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

🏆 Tennessee wins first-ever title

💪 How the Vols conquered the 2024 MCWS

👀 Tennessee's Dreiling is MOP

🎥 Game 3 highlights

Paris 2024 Olympics

Ncaa.com | june 24, 2024, the college careers of the usa women's basketball team at the 2024 olympics, including awards and stats.

Twelve former college basketball stars descend on Paris this summer with the hopes of winning the United States an eighth consecutive gold medal in women's basketball in the Olympics. 

Here's a brief background on all 12 players, including notable stats and achievements from their college basketball careers. All WNBA stats through June 11, 2024.

Team USA roster:

• #11 | Napheesa Collier | UConn
• #7 | Kahleah Copper | Rutgers
• #8 | Chelsea Gray | Duke
• #15 | Brittney Griner | Baylor
• #6 | Sabrina Ionescu | Oregon
• #4 | Jewell Loyd | Notre Dame
• #5 | Kelsey Plum | Washington
• #10 | Breanna Stewart | UConn
• #12 | Diana Taurasi | UConn
• #14 | Alyssa Thomas | Maryland
• #9 | A'Ja Wilson | South Carolina
• #13 Jackie Young | Notre Dame

Napheesa Collier

Collier finished her UConn career as one of five Huskies players in the 2,000 points and 1,000 rebounds club, winning the 2016 NCAA title. She was taken No. 6 overall by the Minnesota Lynx in the 2019 WNBA draft and won the gold medal with Team USA in 2020.

School:  UConn Years:  2015-19 Career averages:  16.1 ppg, 2.5 apg, 8.1 rpg, 1.5 spg, 1.7 bpg Current WNBA team:  Minnesota Lynx WNBA career averages:  17.0 ppg, 2.9 apg, 7.8 rpg, 1.7 spg, 1.2 bpg

College career achievements

• Two-time WBCA First Team All-American (2017, '19)
• Two-time AP First Team All-American (2017, '19)
• Two-time USBWA First Team All-American (2017, '19)
• Two-time American Athletic Conference Player of the Year (2017, '19)
• Three-time American Athletic Conference First Team (2017, ’18, '19)
• 2019 Katrina McClain Award winner
• 2019 American Athletic Conference Defensive Player of the Year
• 2019 NCAA Albany Regional Most Outstanding Player
• 2019 American Athletic Conference Tournament Most Outstanding Player
• 2018 AP Third Team All-American
• 2018 NCAA Final Four All-Tournament Team
• 2017 NCAA Bridgeport Regional Most Outstanding Performer
• 2016 American Athletic Conference All-Freshman Team

Kahleah Copper

Copper finished her college career with the third-most points all time in program history (1,872) at Rutgers. After being drafted No. 7 overall in 2016 by the Washington Mystics, she was then traded to the Chicago Sky. There, she won 2021 WNBA Finals MVP as the team won its first title in franchise history.

School:  Rutgers Years:  2012-16 Career averages:  14.1 ppg, 1.3 apg, 5.8 rpg, 1.1 spg, 0.3 bpg Current WNBA team:  Phoenix Mercury WNBA career averages:  11.8 ppg, 1.3 apg, 3.6 rpg, 0.6 spg, 0.2 bpg

•  Second Team All-Big Ten by both the coaches and media (2016)
• Big Ten All-Academic honoree (2016)
• Second Team All-Big Ten by both the coaches and media (2015)
• Second Team All-Met (2015)
• Big Ten All-Academic honoree (2015)
• WNIT’s Most Valuable Player (2014)
• First Team All-American Athletic Conference (2014)
• First Team All-Met (2014)
• Big East All-Freshman Team (2013)

Chelsea Gray

Gray left Duke as arguably the best player in school history, despite her final two seasons with the Blue Devils marred by injury. Gray's playmaking ability led to her being selected with the 11th pick in the 2014 WNBA draft. Her long professional career has already yielded a 2020 Olympic gold, three WNBA championships and the 2022 WNBA Finals MVP.

School:  Duke Years:  2010-14 Career averages:  11.1 ppg, 5.0 apg, 4.4 rpg, 2.7 spg, 0.2 bpg Current WNBA team:  Las Vegas Aces WNBA career averages:  12.4 ppg, 5.0 apg, 3.1 rpg, 1.1 spg, 0.2 bpg

• WBCA Coaches All-America (2013)
• Associated Press Second Team All-America (2013)
• Co-ACC Player of the Year by the head coaches (2013)
• John R. Wooden finalist (2013)
• All-ACC First Team selection (2013)
• All-ACC Defensive Team (2013)
• Finalists for the Nancy Lieberman Award (2012)
• USBWA finalist for National Player of the Year (2012)
• Posted a school-record 201 assists (2012)
• All-ACC Freshman Team honoree (2011)

Brittney Griner

Griner compiled one of the most dominant college careers in history in four seasons at Baylor. A two-time Olympic gold medalist, Griner also won a national title in Baylor and a WNBA championship with the Phoenix Mercury.

School:  Baylor Years:  2009-13 Career averages:  22.2 ppg, 1.6 apg, 8.8 rpg, 0.5 spg, 5.1 bpg Current WNBA team:  Phoenix Mercury WNBA career averages:  17.7 ppg, 1.9 apg, 7.4 rpg, 0.5 spg, 2.7 bpg

• Finished college career with an NCAA record (men or women) 748 career blocks, second on NCAA Division I women's all-time scoring list (3,283), second on NCAA Division I women's all-time most games scoring in double figures chart with 146 performances, Big 12's all-time leading scorer, NCAA women's record 18 dunks, including three in one game (Florida State, 3/26/13).
• Big 12 Player of the Year (2011, 2012, 2013)
• Wade Trophy (2012, 2013)
• Naismith Trophy (2012, 2013)
• Wooden Award (2012, 2013)
• AP Player of Year (2012, 2013)
• Ann Meyers Drysdale (USBWA) Player of the Year (2012, 2013)
• AP All-American (unanimous) (2012, 2013)
• Consensus First-Team All-American (2011, 2012, 2013)
• WBCA Defensive Player of the Year (2011, 2012, 2013)
• Big 12 championship MOP (2011, 2012, 2013)
• Big 12 Defensive Player of the Year (2010, 2011, 2012, 2013)
• All-Big 12 First Team (2010, 2011, 2012, 2013)
• All-Big 12 Defensive Team (2010, 2011, 2012, 2013)
• Big 12 Female Athlete of the Year (2012)
• MOP NCAA Championship, Final Four (2012)
• MOP NCAA Championship, Des Moines Regional (2012)
• MOP of Big 12 Championship (2011, 2012)
• Wooden Award, Wade Trophy, Naismith Award, Honda Award finalist (2011)
• USBWA Freshman of the Year (2010)
• AP All-American (second team) (2010)
• Big 12 Freshman of the Year (2010)
• All-Big 12 Freshman Team (2010)

Sabrina Ionescu

The NCAA's all-time leader in triple-doubles will make her Olympic debut for Team USA this summer. Ionescu's four seasons at Oregon were legendary, and led to her selection with the No. 1 pick in the 2020 WNBA draft by the New York Liberty.

School:  Oregon Years:  2016-20 Career averages:  18.0 ppg, 7.7 apg, 7.3 rpg, 1.5 spg, 0.3 bpg Current WNBA team:  New York Liberty WNBA career averages:  15.9 ppg, 5.8 apg, 5.9 rpg, 0.9 spg, 0.3 bpg

• AAU James E. Sullivan Award Co-Winner as the nation's most outstanding amateur athlete (2020)
• Wooden Award Winner (2019, 2020)
• AP First-Team All-America (2019, 2020)
• Pac-12 Player of the Year (2018, 2019, 2020)
• Nancy Liebermann Award as nation's top point guard (2018-20)
• National Freshman of the Year (2017)
• First player (female or male) in NCAA history with 2,000 career points (2,562), 1,000 career rebounds (1,040) and 1,000 career assists (1,091).  
• Only the second player in NCAA history (female or male) with 2,000+ points and 1,000+ assists, joining Courtney Vandersloot (Gonzaga). 
• NCAA all-time leader (women's and men's) with 26 career triple-doubles; previous record was 12 by Kyle Collinsworth (BYU, 2012-16). 
• NCAA single-season record eight triple-doubles in both 2018-19 and 2019-20. 
• Finished career at No. 4 on NCAA women's all-time assists list.
• Only the fifth player in NCAA women's history to surpass 1,000 career assists, and the 12th to reach 900 career assists. 
• Pac-12 women's and men's all-time assists leader (Previous: Gary Payton, 938).
• Oregon all-time leader in points, assists, three-point field goals (329) and double-figure scoring games (134). 
• Tied for fifth in NCAA women's history with 134 double-figure scoring games.
• No. 2 in Pac-12 history in double-figure scoring games, No. 3 in three-pointers, No. 5 in points and No. 19 in rebounds. 
• No. 2 in Oregon history in career free-throw percentage (85.1) and field goals made (905). 
• No. 3 in Oregon history in career steals (207).
• No. 4 in Oregon history in double-doubles (58), three-point field-goal percentage (42.2) and games played (142).
• Pac-12 Tournament all-time leader in points (250), assists (86) and field goals made (90); two-time Pac-12 Tournament most outstanding player (2018, 2020). 

Jewell Loyd

Add Lloyd to the list of players already with an Olympic gold medal on their resume. Prior to winning gold with Team USA in 2020, Loyd was the first overall pick in the WNBA draft in 2015 out of Notre Dame. Loyd's four-year career with the Irish saw two first-team All-American selections as one of the most explosive scorers in college basketball.

School:  Notre Dame Years:  2012-15 Career averages:  17.0 ppg, 2.4 apg, 5.7 rpg, 1.4 spg, 0.4 bpg Current WNBA team:  Seattle Storm WNBA career averages:  16.7 ppg, 3.2 apg, 3.6 rpg, 1.2 spg, 0.2 bpg

• espnW National Player of the Year (2015)
• John R. Wooden Award Runner-Up (2015)
• Unanimous AP All-America First Team (2015)
• ACC Player of the Year (2015)
• ACC Tournament Most Valuable Player (2014, 2015)
• All-ACC Defensive Team (2015)
• NCAA Women’s Final Four All-Tournament Team (2014, 2015)
• AP All-America Second Team (2014)
• USBWA National Freshman of the Year (2013)
• BIG EAST Freshman of the Year (2013)

Kelsey Plum

Plum's 57-point senior night broke the NCAA scoring record as part of an illustrious four-year career at Washington. Plum's 3,527 points in her career were the most in NCAA history when she graduated. Plum also set the single season scoring record in 2016-17 with 1,109 points.

School:  Washington Years:  2013-17 Career averages:  25.4 ppg, 3.8 apg, 4.3 rpg, 1.4 spg, 0.2 bpg Current WNBA team:  Las Vegas Aces WNBA career averages:  13.9 ppg, 4.0 apg, 2.4 rpg, 0.9 spg, 0.1 bpg

• No. 2 all-time NCAA scoring leader with 3,527 points
• Broke single-season NCAA scoring record with 1,109 points in 2016-17
• Broke 33-year-old NCAA career free throw record (912)
• AP Player of the Year (2017)
• espnW Player of the Year (2017)
• Pac-12 Player of the Year (2017)
• WBCA and USWBA All-American (2016, 2017)
• Unanimous first team All-American by AP (2017)
• Unanimous first team All-American by espnW (2017)
• Unanimous All-Pac-12 selection
• Won the Dawn Staley, Ann Meyers Drysdale, Nancy Lieberman awards, Naismith and Wade Trophy (2017)
• Tom Hansen Conference Medal recipient (2017)
• Pac-12 All-Academic honorable mention (2015, 2016, 2017)
• Seattle Female Sports Star of the Year (2017)
• Wooden Award All-American (2016)
• AP Third Team All-American and Naismith Award Semifinalist (2016)
• Pac-12 All-Tournament Team (2016)
• Second Team All-America (2016)
• Two-time WBCA All-Region Team (2015, 2016)
• Four-time All-Pac-12 Team (2014, 2015, 2016, 2017)
• Washington all-time three-point leader, second all-time in Pac-12
• Pac-12 all-time single-season scoring leader
• Pac-12 record holder for free throws made (season and career)
• Pac-12 single-game scoring record of 57 points vs. Utah
• 2014 Pac-12 Freshman of the Year

Breanna Stewart

UConn's Stewart is arguably the most accomplished player in college basketball history. She was a four-time national champion at UConn and won most outstanding player in the NCAA tournament all four seasons of her college career. Stewart is also a two-time Olympic gold medalist with Team USA.

School:  Connecticut Years:  2012-16 Career averages:  17.6 ppg, 2.8 apg, 7.8 rpg, 1.5 spg, 2.7 bpg Current WNBA team:  New York Liberty WNBA career averages:  20.7 ppg, 3.1 apg, 8.8 rpg, 1.4 spg, 1.5 bpg

• Associated Press Player of the Year (2014, 2015, 2016)
• Naismith Trophy Winner (2014, 2015, 2016)
• 2016 Wade Trophy Winner (2015, 2016)
• 2016 John R. Wooden Award Winner (2015, 2016)
• 2016 USBWA Player of the Year (2014, 2015, 2016)
• Consensus First Team All-American (2014, 2015, 2016)
• American Athletic Conference Player of the Year (2014, 2015, 2016)
• Final Four Most Outstanding Performer (2013, 2014, 2015, 2016)
• American Athletic Conference Championship MOP (2014, 2016)
• Sullivan Award Winner (2016)
• Only player in history to be honored as the Final Four Most Outstanding Player more than twice
• Second place on UConn's all-time scoring list with 2,676 career points
• The only player in NCAA history to block at least 400 shots and dish-out at least 400 assists
• UConn's all-time leader in blocked shots at 414

Diana Taurasi

A legend of basketball, Taurasi will compete for her sixth gold medal with Team USA. Taurasi had a legendary career at UConn before taking over pro basketball, with three national championships while playing for Geno Auriemma.

School:  Connecticut Years:  2000-04 Career averages:  15.0 ppg, 4.5 apg, 4.4 rpg, 1.2 spg, 1.0 bpg Current WNBA team:  Phoenix Mercury WNBA career averages:  19.1 ppg, 4.2 apg, 3.9 rpg, 0.9 spg, 0.6 bpg

• Naismith national player of the year (2003, 2004)
• NCAA national champion (2002-2004)
• NCAA Tournament most outstanding player (2003, 2004)
• Wade Trophy (2003)
• Honda Sports Award winner (2003, 2004)
• USBWA Women's National Player of the Year (2003)
• AP College Player of the Year (2003)
• Nancy Lieberman Award Winner (2003, 2004)
• Big East Player of the Year (2003, 2004)

Alyssa Thomas

After leading Maryland to the Final Four in her final year, Alyssa Thomas was drafted No. 4 overall in 2014. After a draft-day trade to the Connecticut Sun, Thomas has played with the team for her entire WNBA career.

School:  Maryland Years:  2010-14 Career averages:  17.5 ppg, 3.6 apg, 9.1 rpg, 1.8 spg, 0.4 bpg Current WNBA team:  Connecticut Sun WNBA career averages:  12.4 ppg, 4.3 apg, 7.5 rpg, 1.5 spg, 0.3 bpg

• WBCA Coaches' All-American (2012)
• USBWA All-American (2012)
• AP All-America First Team (2012)
• ACC Player of the Year (2012, 2013)
• ACC Tournament MVP (2012)
• All-ACC Tournament Team (2012, 2013)
• All-ACC Team (Second Team, 2011; First Team, 2012, 2013)
• Wooden Ballott (2012, 2013)
• Naismith Preseason list (2012, 2013)
• Wade Watch list (2012, 2013)
• Preseason All-ACC Team (2011)
• ACC All-Freshman Team (2010)
• ACC Rookie of the Week (7X)
• ACC Player of the Week (7X)
• ESPNW National Player of the Week (1/14/13)
• ECAC Rookie of the Year (2011)
• FullCourtPress.com All-American (2012)
• FullCourtPress.com All-Freshman Second Team (2011)
• All-time Maryland Terrapins leader in points (2356), rebounds (1,235), and double-doubles (66)

A'Ja Wilson

A gold medal, two WNBA championships and an NCAA championship are just a small part of A'Ja Wilson's impressive basketball resume. At South Carolina, Wilson was a three-time consensus first-team All-American and won NCAA tournament MOP in 2017.

School:  South Carolina Years:  2014-18 Career averages:  17.3 ppg, 1.4 apg, 8.7 rpg, 1.0 spg, 2.6 bpg Current WNBA team:  Las Vegas Aces WNBA career averages:  20.3 ppg, 2.2 apg, 8.8 rpg, 1.1 spg, 1.9 bpg

• Consensus National Player of the Year (2018)
• Wooden Award Winner (2018)
• Naismith Trophy Winner (2018)
• Wade Trophy Winner (2018)
• Honda Sport Award – Basketball (2018)
• All-America First Team (2017, 2018)
• SEC Player of the Year first three-time selection in league history (2016, 2017, 2018)
• SEC Defensive Player of the Year (2016, 2018)
• All-SEC First Team (2016, 2017, 2018)
• SEC Tournament MVP (2017, 2018)
• NCAA Final Four MOP (2017)

Jackie Young

Young was the leader of Notre Dame's 2018 national championship team as a sophomore guard. After her junior season at Notre Dame, she was selected No. 1 overall in the WNBA Draft.

School:  Notre Dame Years:  2016-19 Career averages:  12.4 ppg, 3.5 apg, 6.3 rpg, 1.2 spg, 0.4 bpg Current WNBA team:  Las Vegas Aces WNBA career averages:  13.2 ppg, 3.9 apg, 4.0 rpg, 1.1 spg, 0.2 bpg

• NCAA champion (2018)
• ACC tournament MVP (2019)
• Second-team All-ACC (2019)
• All-ACC Freshman team (2017)

NCAA Student-Athletes at the 2024 Paris Olympics 🥇🥈🥉  

College careers of NCAA student-athletes at 2024 Paris Olympics:  🏀 Basketball: Men's team USA | Women's team USA | Women's 3x3 | Men's 3x3   🏐 Volleyball: U.S. Women's  | ⛳️ Golf: USA men's | USA women's 🥇🥈🥉 📺  WATCH: College highlights of former-NCAA Olympians: 🤼‍♂️ Wrestling:   Kyle Dake | Zain Retherford | Aaron Brooks 🏐  Women's volleyball: Kathryn Plummer  | Dana Rettke | Avery Skinner 🏀 Men's hoops:  Steph Curry | Devin Booker |  Anthony Davis 🏀 Women's hoops: Sabrina Ionescu  | Kelsey Plum |   A'ja Wilson ⛳️ Men's g olf: Scottie Scheffler 🔥History:  2022 NCAA student-athlete Olympic medal winners  

Highlights from the college careers of the USA women's golf roster at the 2024 Olympics

• Where the 2024 U.S. Olympic women's 3x3 basketball team played in college

2024 USA men's golf Olympic roster: College careers, highlights

March madness.

• 📆 2024 March Madness schedule
• ❓ How the field of 68 is picked
• 👀 Best performances in tournament history

Women's Final Four Most Outstanding Players from 1982 to present

The 13 highest-scoring individual performances in March Madness women's history

Women's basketball championship history

Schools with the most DI women's basketball national championships

Best performances in NCAA women's basketball tournament history

Di women's basketball news.

• The college careers of the USA women's basketball team at the 2024 Olympics, including awards and stats
• Four years after Iowa home debut, Caitlin Clark finds new wave of fans, obstacles, attention in first WNBA home game
• Iowa women's basketball head coach Lisa Bluder announces retirement
• ROOTED: Episode 5 - Bloom
• Caitlin Clark's full career March Madness highlights (2021-2024)
• 2024 WNBA Draft top picks' March Madness highlights
• Colleges & conferences with the most players taken in the 2024 WNBA draft
• Every HBCU player ever selected in the WNBA draft
• College basketball teams with the most WNBA 1st-round draft picks

Follow NCAA Women's Basketball

• Share full article

Advertisement

Supported by

Charles M. Blow

On Juneteenth, Freedom Came With Strings Attached

By Charles M. Blow

Opinion Columnist

Last week at a Juneteenth concert on the South Lawn of the White House, Vice President Kamala Harris said that on June 19, 1865, after Union troops arrived in Galveston, Texas, “The enslaved people of Texas learned they were free.” On that day, she said, “they claimed their freedom.”

With those words, Harris, who stood alongside President Biden when he admirably signed the legislation that made Juneteenth a federal holiday, expressed a common oversimplification, one born of our tendency to conjugate history’s complexities: Although it’s a mark of progress to commemorate the end of American slavery, it’s imperative that we continue to underscore the myriad ways in which Black freedom was restricted long after that first Juneteenth.

To start, there is some debate over whether most of the estimated 250,000 enslaved people in Texas at the time didn’t know about the Emancipation Proclamation. As the Harvard professor Henry Louis Gates Jr. told me recently, “I have never met a scholar who believes that’s true.”

But more important, emancipation was not true freedom — not in Texas and not in most of the American South, where a vast majority of Black people lived. It was quasi freedom. It was an ostensible freedom. It was freedom with more strings attached than a marionette.

Most Black people couldn’t claim their freedom on June 19, 1865, because their bodies (and their free will) were still being policed to nearly the same degree and with the same inveterate racism that Southern whites had aimed at them during slavery.

The laws governing the formerly enslaved “were very restrictive in terms of where they could go, what kind of jobs they could have, where they could live in certain communities,” said Daina Ramey Berry, the dean of humanities and fine arts at the University of California, Santa Barbara, and the author of “The Price for Their Pound of Flesh: The Value of the Enslaved, From Womb to Grave, in the Building of a Nation.”

We are having trouble retrieving the article content.

Please enable JavaScript in your browser settings.

Thank you for your patience while we verify access. If you are in Reader mode please exit and  log into  your Times account, or  subscribe  for all of The Times.

Thank you for your patience while we verify access.

Already a subscriber?  Log in .

Want all of The Times?  Subscribe .