Evaluative Conditioning: The How and Why of Forming Opinions without Facts.

Have you ever wondered why we judge a book by its cover? Here, you will read about how and why we form first opinions without evidence and pass judgment on things we barely know. A process termed evaluative conditioning tries to give an explanation on why the book with the beautiful cover is nicely written.

Evaluative Conditioning

Evaluative Conditioning: Why we judge a book by its cover.

Every day, each of us is exposed to several different, and often new, experiences. This is inevitable in a world as dynamic and diverse as ours. A new shop opened three blocks down the road, Katie has a new boyfriend and your favorite musician just released his new album. Sometimes, we then catch ourselves judging things we barely know and giving opinions on matters we scarcely heard of. I’m sure that new shop won’t last much longer than the previous one, Katie’s boyfriend is most likely a handsome lad and the new album? Well, that’s undoubtedly going to be terrific! However, are our predictions true? And on what ground do we so confidently exclaim what arguably has no basis? These questions are of importance for both psychological and industrial research. How we evaluate our surrounding influences our behavior towards it. That said, whoever understands its underlying mechanisms and manages to willingly direct them could, in turn, manipulate a patient into ceasing ill behavior or a client into buying a certain product.

Classical conditioning: learning based on experience

Evaluative Conditioning

Classical Conditioning: also known as Pavlovian conditioning

One mechanism of interest was termed evaluative conditioning. It was named in accordance to the previously established classical conditioning. The latter, famous for its initial experiments by Pavlov featuring dog salivation, describes a learning process that creates expectations based on reoccurring patterns. If a situation is always followed by the same event, just experiencing the first will be enough to foretell that the other will also set in.

So, for example, when you see the signaling light of a car’s left blinker, you automatically expect it to take a left turn. This is the natural sequence of events, as we have learned to know it: blinking light then turning the car.

Classical conditioning

First empirical research on this dogma was conducted on dogs. Each time they were fed, the same ringing noise of a bell would precede the serving of their food. At first, the dogs would happily welcome the meal, spittle flowing into their mouth once the food was presented to them. Then, after some time, the saliva would start dripping just by the mere sound of the bell. Over time, the dogs had learned that when the bell starts ringing, food will soon follow. Formally, the ringing of the bell was categorized as a conditioned stimulus (CS), while the serving of food was described as an unconditioned stimulus (US). Thereafter, it was concluded that the process termed classical conditioning occurs when one stimulus, an unconditioned stimulus, is reliably preceded by another stimulus, a neutral stimulus. The meaning of the latter unconditioned stimulus is applied to the neutral stimulus. Once the association is set, the neutral stimulus is termed conditioned stimulus, as it triggers the same reaction pattern as the unconditioned one. If we apply this to our example with the car, then the blinking light would be the conditioned stimulus, while the subsequent turning would be the unconditioned. Furthermore, the blinkers would be meaningless unless they were to reliably indicate the turning of the car.

Afterward, several studies were conducted on classical conditioning elucidating its properties. As turns out, the stimuli do not have to co-occur every time, just frequent enough. Furthermore, the required rate depends on the nature of the stimuli, as some can be stronger indicators than others. The research was also conducted on the stability of the effect. Here, it was shown that once the stimuli reliably stop cooccurring, the association ceases to exist. From that point onwards, the conditioned stimulus returns to being a neutral stimulus.

Evaluative conditioning: opinions based on experience

Evaluative Conditioning

Evaluative Conditioning

While the basis of evaluative conditioning is found in classical conditioning, its roots dig deeper. It emerged from research on attitudes. As it became clear that evaluative conditioning could be one mechanism influencing the formation and change of attitudes, a research field of its own was established. Evaluative conditioning follows similar rules to classical conditioning, as both have an unconditioned and a conditioned stimulus. Here, however, one event is not followed by the other, but both occur at the same time. Also, instead of predicting an immediate event and preparing the appropriate reaction to it, a long-term influence is predicted and the appropriate stance towards it chosen.

Opinions influence behavior

It goes like this: the rating of something on a two-dimensional scale (such as good and bad, likes and dislikes), officially called valence, influences the behavior towards it. You approach something you like and distance yourself from something you do not like. Hereby, imminent harmful or even life-threatening events must be disregarded, as a defense, and self-preserving mechanisms would influence behavior in ways beyond just attitudes. You approach or distance yourself from something that may have a long-term positive or negative influence on you. You recently got to know Thomas, however, you do not like Thomas. He has opinions you disagree with. Although he does nothing harmful, you guess that he would still be a “bad” influence on you. Therefore, you try to distance yourself from him. (Sorry to all Thomas’, it is just an example, please do not take it personally.)

Baseless assumptions?

If then an unknown factor appears together with an already judged factor, the process called evaluative conditioning uses the assessment of the known one to predict the long-term influence of the new event. In other words, this is a mechanism that uses the categorization of a known target to sort a somehow connected but still unknown target into the same subjective two-dimensional scale (e.g. good and bad, like and dislike).

For ease of comparison, the events were similarly named to those of classical conditioning. First, we have an unconditioned stimulus which is either positively or negatively valenced. Then, we have the conditioned stimulus with a neutral valence, or at least a lesser valence, than that of its unconditioned counterpart. When both stimuli are then viewed together, seemingly connected to each other, the opinion of the unconditioned stimulus is then applied to the conditioned one.

Sorry Thomas, but to keep to our example: if this previously mentioned Thomas appears with a friend of his, you will most likely be not too keen on getting to know this new fellow. He will probably have a similar conviction to the one you disagree with and would, therefore, be the same “bad” influence as Thomas. Thus, he is just as unlikable as his friend. Hereby, a cognitive association between Thomas and this stranger was created that categorized the unknown person similarly to the known Thomas, thereby triggering a distancing stance towards the new subject. It goes without saying that we have no factual evidence of this newly met person to be as “bad” as the first, furthermore we do not know for certain if Thomas would influence us in a wrong way.

Properties of evaluative conditioning

It seems that the attitude adopted is always that of the more extreme opinion. To clarify, if a slightly negative and a strongly positive viewed stimulus appear together, the slightly negative will surely be judged more positively upon that. If your best friend Rebecca suddenly introduces you to an acquaintance of hers, surprisingly the same stranger you saw earlier chatting with Thomas in a friendly manner, then you will probably change your mind and give him a chance. Rebecca is awesome, no chance in hell that this guy could be a disappointment. Sure, he seemed friendly with Thomas earlier, but if he is with Rebecca, then he will be all right.

However, it should be noted, especially regarding our example with poor Thomas, that it is still under debate whether this change of valence occurs consciously or subconsciously, or if it even could be prevented through conscious knowledge. Even more, no hard evidence was presented on how we were to judge a positive or negative conditioned stimulus after several presentations with neutral stimuli. Unfortunately, some studies showed that after repeated co-occurrence the formerly conditioned stimulus turns neutral again, while other showed that the valence evaluation resists this so-called extinction phase. At last, the timing seems to play a role in the process. It was shown that evaluative conditioning works best when both stimuli appear at the same time. Nonetheless, if the conditioned stimulus is presented shortly before or after the unconditioned stimulus, the conditioning can still take place.

Implications of evaluative conditioning

Evaluative Conditioning

Evaluative Conditioning

Knowledge about opinion formation and their change is a serious topic and like many others must be approached with responsibility. Private companies are most likely already conducting research in the field, as efficiently associating a product with something pleasant could greatly boost their sales. There are two main problems with this:

  1. First, these findings would not be accessible to neither the public nor other researchers.
  2. The second problem we face is that its properties are not fully understood. This means that whoever uncovers them could influence the public without a specialist’s notice.

For example, if it were to turn out that evaluative conditioning functions only subconsciously, commercials would be viewed while the tv-program is running, instead of in-between, or products would appear more often in the background of a movie. These changes would seem insignificant to the eyes of an unknowledgeable observer, while actually heavily influencing sales. That said, research funds must be invested into topics such as opinion formation, thereby preventing their abuse for personal gain (as a possible propaganda tool, for instance) while utilizing their enormous potential.

Think of the benefit for health care. Unhealthy behavior (such as certain addictions) could be cured or productivity and motivation increased through according to associations. So, if someone had a horrifying phobia that impeded his life’s quality, for example, the poor botanist Steven with his sudden fear of spiders after an unfortunate vacation, simple associations with strongly positive topics could hastily cure his unpleasant situation. Or, as media addiction is a seemingly increasing problem among our youths, this knowledge could be used to reintroduce the fun of the real-life through gradually improving the attitude towards real experiences. Furthermore, these findings could bear a countermeasure against prejudice, that would benefit the whole of humanity.

Nonetheless, we have a long way ahead of us until then. The current findings can be contradicting and confusing. Most of the research was done on humans which bears certain risks such as demand awareness, among others. Demand awareness can create artificial results due to the test subjects suspecting what results are looked for and reacting accordingly. It should be noted that findings from other research fields suggest that animals also like and dislike, just as us, and have ways of expressing these feelings. I would suggest an additional approach through animal experiments. This would extinguish some of the risks, while at the same time rendering the results more comparable to other experiments on evaluative conditioning and classical conditioning. However, we must acknowledge the research done on the topic and encourage the researchers to keep up their work, stay creative and not give up when all seems to play out differently. The first step to shorten the distance to the above-mentioned vision and prevent abuse of important findings is to create awareness of this field so that more researchers are acknowledged funds to look further into evaluative conditioning.

Check out the following articles to get further knowledge on the topic and give credit to the diligent men and women who worked hard for this knowledge:


Bethell, E. J. (2015). A “how-to” guide for designing judgment bias studies to assess captive animal welfare. Journal of applied animal welfare science, 18 Suppl 1, S18-42. doi:10.1080/10888705.2015.1075833

Bohner, G., & Dickel, N. (2011). Attitudes and attitude change. Annual review of psychology, 62, 391–417. doi:10.1146/annurev.psych.121208.131609

Davey, G. C. (1994). Is evaluative conditioning a qualitatively distinct form of classical conditioning? Behavior research and therapy, 32(3), 291–299.

De Houwer, J., Thomas, S., & Baeyens, F. (2001). Association learning of likes and dislikes: A review of 25 years of research on human evaluative conditioning. Psychological bulletin, 127(6), 853–869. doi:10.1037//0033-2909.127.6.853

Field, A. P. (2000). Evaluative conditioning is pavlovian conditioning: Issues of definition, measurement, and the theoretical importance of contingency awareness. Consciousness and cognition, 9(1), 41–49. doi:10.1006/ccog.1999.0412

Gawronski, B., & Bodenhausen, G. V. (2006). Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change. Psychological bulletin, 132(5), 692–731. doi:10.1037/0033-2909.132.5.692

Havermans, R. C., & Jansen, A. (2007). Evaluative conditioning: A review and a model. Netherlands journal of psychology, 63(2), 31–41. doi:10.1007/BF03061060

Rozin, P., Wrzesniewski, A., & Byrnes, D. (1998). The elusiveness of evaluative conditioning. Learning and motivation, 29(4), 397–415. doi:10.1006/lmot.1998.1012