The effects of the first person perspective on emotions and attitudes:TL;DR

In 2015, I did some in-depth research and experimentation on the use of the first person perspective in commercials (this wasn't for fun, it was for my master's thesis). Because of all the sweaty palm inducing GoPro videos floating around YouTube, I got the idea that this style of filming could be used in commercials to make the viewing experience more exciting, differentiate the commercial from all the other noise consumers are exposed to, and maybe have some unexpected psychological effects.

My plan was to gather as much information on the topic as I could find, then use that research to design a well informed experiment which would identify the differences in how people perceive first person vs third person commercials. Before any research took place, I needed a central question to help me keep focus throughout all this, what I came up with was as follows: What effect does filming a commercial in the first person perspective have on consumers' emotions and attitudes towards a product?

The first set back I experienced was the complete lack of related research available. I'm sure that big companies, such as GoPro, have put a lot of resources into researching this type of thing, but having a budget of just under zero dollars, I wasn't able to find or buy any of it. So this became exploratory research. I broke down what it was I was looking for into manageable chunks; psychology of emotion, psychology of attitudes, consumers' emotional and attitudinal responses to advertising, and the psychology of perspective.

I could (and did) write in great depth on what I found, but for the sake of brevity below are chapter summaries of my secondary research along with a TL;DR (too long; didn't read). You can even scroll to the very end for a summary of my findings in bold font.


“In conclusion, emotion is a complex phenomenon that begins with physiological activity within various brain structures, notably, the thalamus, hypothalamus, amygdala, and the hippocampus. The feelings produced by emotional states are the results of both physiological and cognitive processes, which has made it very difficult for researchers to determine the specific underlying order of events. Earlier theories regarding this order of events (James-Lange, Cannon-Bard) proposed that the entire process was physiological, and therefore the feelings that resulted were essentially preprogrammed within the human brain. Later theories (Schachter, Lazarus) proposed that cognition not only helps to develop emotional states, but plays a very large role in doing so. For anyone interested in influencing emotional states, the addition of cognition to the emotional process is very important because cognition can easily be influenced, or manipulated. Influencing an individual’s cognition of a stimulus has been proven to alter their emotional appraisal- this is referred to as cognitive mediation.”

TL;DR: You can influence an individual’s emotional response to a stimulus by manipulating their cognition of the stimulus. Is it possible to manipulate someone's emotional response to a stimulus? Yes.

A chart showing the Lazarus model of emotion.
This is Lazarus' model. It asserts that by manipulating the appraisal process (via cognitive mediation), the emotional and physiological response outcomes can be manipulated.


“In conclusion, an attitude is a positive or negative evaluation of a stimulus. Attitudes are very useful for predicting and explaining behavior, likewise, behaviors are very useful for predicting attitudes. People are aware of both their attitudes and their behaviors and literally strive for consistency between them, meaning they will adjust their behaviors to match their attitudes or vise-versa. Manipulating peoples’ attitudes is a very common practice, and is typically accomplished by attempting to appeal to an individual cognitively (persuasion). Newer theories, notably the theory of transportation, propose that it is possible to manipulate attitudes without appealing to cognitive processes. In these cases, the level of imagery is what determines the degree of attitude change, and when the level of imagery is high, viewers will accept the facts of the narrative world over the facts of the real world. In many cases this will occur by a viewer identifying with the focal character in a narrative, and their opinions will adjust cause them to become more homogeneous with the focal character.”

TL;DR: There are two ways to influence an individual's attitudes towards a subject, appealing to the individual's cognition (persuasion), or employing high levels of imagery in some type of narrative (book, movie, etc.) causing the individual to emulate the attitudes of the focal subject. Is it possible to manipulate someone's attitudes towards an object? Scarily so.

Consumers' Emotional and Attitudinal Responses to Advertising

“In conclusion, the four studies analyzed in this chapter all study viewer/listener responses to advertisements by analyzing attitudes and emotions as dependent variables. A few different methods were employed to measure emotional response: scoring a feeling based on intensity, the PAD (Pleasure Arousal Dominance) emotional state model, and a modified version of Edell and Burke’s (1987) feelings inventory.

Each method has advantages and disadvantages, for instance, asking respondents only to score the intensity which they feel certain emotions is very simple and easy for them to understand, but when exploratory research is being conducted, i.e. the possible influence on emotion is completely unknown, this is insufficient because of the complexity of emotions and the simplicity of this method. Edell and Burke’s inventory of feelings, on the other hand, is a much more in depth analysis of feelings, and the feelings in the inventory are designed to be relevant to advertising. This method is far superior than simply scoring the intensity of a handful of feelings, but again, when considering the possible effects that a stimulus could have on an individual, this method is just too simple- it has a list of very specific feelings, and can only determine if an individual is experiencing one of those particular feelings.

The purpose of this thesis (in part) is to determine if there is any difference in emotional appraisal, not to determine if there is a difference in the appraisal of this particular predetermined list of feelings. The PAD emotional state model therefore stands out as the most appropriate method to apply, because it asserts that any emotional state that a person can experience can be identified and given a spatial location within the model. This method is far more thorough than the other methods, and gives researchers the ability to find a result they might not have been looking for.

Within these four case studies, there was much more homogeneity in the methodology applied to measure attitudes. In each study a semantic differential was used, and in all cases (with the exception of Olney, Holbrook, & Batra (1991)) the items chosen for the semantic differential were standard bi-polar answers (good/bad, boring/interesting, etc.). The semantic differential items chosen by Olney, Holbrook, and Batra (1991) on the other hand were designed to describe attitudes towards a product and it provides a much more detailed result. Instead of just determining if respondents’ attitudes are favorable or unfavorable, this method determines an individual’s attitudes in three separate dimensions: hedonism, utilitarianism, and interestingness. Like with the PAD emotional state model, this is a very encompassing method which will provide much more depth in its findings than other methods.”

TL;DR: The best method available to quantify viewers' emotions is the PAD (Pleasure, Arousal, Dominance) emotional state model. Although it wasn't mentioned in the above conclusion, this model can be improved by removing the Dominance dimension. The best method for quantifying attitudes is using a semantic differential scale among different dimensions of attitudes. Also not mentioned above, the two most distinct dimensions of attitudes are hedonic and utilitarian. These additional details weren't in this summary because I used this chapter's information to guide further research in my methodology.

Psychology of Perspective (The Canonical Perspective)

“In conclusion, the canonical perspective is a ’better’ perspective, or viewing point, of a focal object. This perspective is found to be considered both more informative, and more aesthetically pleasing. Additionally, images viewed from the canonical perspective provide enhanced recall memory compared to images from the non-canonical perspective. Canonical perspectives are formed from frequency, meaning the more often an individual views an object from a particular perspective, the more informative and aesthetically pleasing they find that perspective to be. It can be inferred, therefore, that the perspective from which an individual views an object affects their appraisal thereof.”

TL;DR: Just read the damn paragraph.

Images Ghose and Liu used to compare canonical and non-canonical perspectives.
Ghose and Liu’s canonical (column 1) and non-canonical (column 2) images (Source:; Ghose & Liu 2013)

The Experiment

Finally it was time to start the experiment. I would be creating two separate nondescript car commercials, one filmed from the first person perspective, the other from the third person perspective. There was no text on screen, no voice over, and the same audio track in both. Obviously it was important that the only discernible difference between the two commercials was the camera perspective, this includes the weather conditions and sun's position in the sky. I'd have to move fast. I planned the scene to be short and simple, 6 shots, 10 seconds long. For the first person commercial, I mounted the camera so it hung from my forehead down in between my eyes. Driving like this was hilariously dangerous, and possibly unnecessary. For the third person perspective I used what I found to be the 6 most typical commercial shots in which the camera was bound to the car, because there needed to be what I referred to as “a first person equivalent” for each shot. I had been looking forward to filming these commercials for a very long time (over a year) and I was surprised when execution time came that it only took about 25 damn minutes. I needed only a few days to edit the commercials, create an online questionnaire, and spam the internet for participants. There were two forms of the questionnaire, control (third person perspective) and experimental (first person perspective). Which version a participant saw was based on whether their birthday was on an odd or even day. Below are frames from each of the 6 shots, first person on the left, third person on the right.

The storyboard of Dan Carqueville's car commercials comparing the first and third person perspectives in advertising.


After 12 days, 159 people completed the questionnaire. Besides demographic data, there were 3 primary sections to the questionnaire designed to quantify emotions, attitudes towards the product, and level of evoked imagery (this would help identify if the first person perspective is the canonical perspective). Participants would answer the questionnaire after viewing one of the two commercials. I analyzed the data with R and was struggling to find any statistically significant differences between responses from the control and experimental questionnaires. Finally, some findings popped out: When viewing the first person perspective commercial... 15-24 year olds experienced lower levels of pleasure, 15-24 year olds had reduced utilitarian attitudes towards the product, and men had reduced hedonic attitudes towards the product.

I was happy to find any statistically significant results, but pretty bummed out that my intuition was wrong. The longer I sat to think about it, the more questions I had. Would the results have been reversed if I was driving a sports car instead of an economy car with a 60hp engine? What if I was doing something more exciting, like climbing the side of a mountain? Did they think my driving gloves were stupid? Is the third person perspective actually the canonical perspective for commercials?

This was exploratory research, so the fundamental purpose was to provide a foundation for further research. It would be very interesting to see this experiment recreated with different products and scenarios. Regardless of which perspective outperforms the other, I'm certain there would continue to be statistically significant differences. As the world's (self proclaimed) leading expert in this field of study, I believe further research would discover that for certain products or scenarios, one camera perspective would have a stronger impact on attitudes and emotions than the other.