Assimilation and Contrast Effects on Consumers= Choices

ABSTRACT - This article studies how consumers’ choices are influenced with assimilation and contrast effects. The participants were required to choose from a set of two options that were described on two attributes, and the two options were constructed such that one was superior on attribute A but was inferior on attribute B. It is hypothesized that when exposed to extreme context information about one attribute, the consumer will subjectively lessen the difference on that attribute in the choice set and be more likely to choose an option that is inferior on that attribute.

Citation:

Hao Shen (2002) ,"Assimilation and Contrast Effects on Consumers= Choices", in AP - Asia Pacific Advances in Consumer Research Volume 5, eds. Ramizwick and Tu Ping, Valdosta, GA : Association for Consumer Research, Pages: 156-158.

Asia Pacific Advances in Consumer Research Volume 5, 2002      Pages 156-158

ASSIMILATION AND CONTRAST EFFECTS ON CONSUMERS= CHOICES

Hao Shen, China Europe International Business School (CEIBS), China

ABSTRACT -

This article studies how consumers’ choices are influenced with assimilation and contrast effects. The participants were required to choose from a set of two options that were described on two attributes, and the two options were constructed such that one was superior on attribute A but was inferior on attribute B. It is hypothesized that when exposed to extreme context information about one attribute, the consumer will subjectively lessen the difference on that attribute in the choice set and be more likely to choose an option that is inferior on that attribute.

In experiment 1, two types of questionnaires were distributed randomly to all the subjects. In both questionnaires, subjects were asked to choose from a set of two options, "purchasing a book for $49" or "taking 15 minutes’ ride to purchase the book for $35". Experimental subjects were exposed to his classmate who purchased books for a large sum of money while control subjects were not.

In experiment 2, two types of questionnaires were distributed randomly to all the subjects. In both questionnaires, subjects were asked to choose from a set of two options, "a beefsteak that costs $40" or "a beefsteak that have a better taste but costs $80". Experimental subjects were exposed to "a delicate bottle of wine that costs 200" before taking the order while control subjects were exposed to "a delicate piece of pizza that costs $20" before taking the order.

Although the result of experiment 1 somehow contradicted the hypothesis, the result of experiment 2 successfully confirmed it. The explanations of such results and implication of the hypothesis were discussed.

Consumers often depend on some attributes of products to make an overall evaluation of them since complete product information is rarely available to consumers (Henry Assael 1987; Huber and McCann 1982). However, consumers often feel puzzled when they have to make a choice from a given product category, where all of products, though with different attribute values, are of no decisive advantage. Consider a consumer who wants to see a movie. There are two available tickets, one with a good seat but more expensive and the other cheaper but with a less good seat. If none of these two attributes, "seat" and "cost", are principal to this consumer, and if the overall evaluations of these two tickets are approximately the same, then he or she will find it hard to make a decision.

Psychologists find that on most occasions the purchase decisions of consumers are context-dependent (Simonson&Tversky, 1992;Thompson&John Edward, 1983). In this article, I want to discuss assimilation and contrast effects on consumer choice. On some occasions, information that is not related to the given product category will greatly influence the decision making of consumers. As to the example above, sometimes assimilation and contrast effects may help this consumer reach a resolution.

Assimilation and contrast effects will make people don’t judge things as they "really" are. "Contrast effect" reflects a displacement of judgment away from the implication of the available information while "assimilation effect" reflects a displacement of judgment toward the implication of the available information. (Thompson and John Edward, 1983)

Assimilation and contrast effects can be easily found in our life. Eiser (1994) described the contrast effect in this way: if you put one hand into one bowl of cold water and the other hand into a bowl of hot water, afterwards you put both hands into tepid water, then the hand that was in hot water will find the tepid water cold whereas the hand that was in cold water will find the tepid water hot, both contrary to what the water really feels.

Thompson and John Edward (1983) found that effect of assimilation and contrast also exists in social judgment, in their experiment, some dissimilar outgroups were manipulated and perceived by the participants as "accepted", "noncommittal" or "rejected", it suggested that in the presence of a more dissimilar outgroup, a contrast effect was observed for "accepted" outgroups while an assimilation effect was observed for "rejected" outgroups, that is, the "accepted" outgroups received a more positive evaluation while the "rejected" outgroups received a more negative evaluation. These findings imply that people will change the evaluation of moderate things if an extreme one is brought in.

A few explanations have been given to such effect. Some psychologists argue that the way the respondents anchor or define the response scale will influence their judgment. For example, people will judge a middle-sized animal (e.g. "wolf ") as small when the rating scale is subjectively anchored by an extremely animal (e.g. "elephant"). They will judge the same middle-sized animal as medium when the rating scale is subjectively anchored by a moderate large animal (e.g. "horse") (Winkielman and Piotr, 1997) Some studies indicate that extreme context information is more likely than moderate context information to be used as an anchor with which a target stimulus is contrasted, that is, people are more likely to contrast the target stimulus when an extreme context information are subjectively used. (Herr and Paul, 1983; Thompson ad John Edward, 1983)

Herr, Sherman, and Fazio (1983) also use the category theory to interpret the effect of assimilation and contrast. They suggest that when encountering an unfamiliar target object or product, an individual attempts to categorize it. If the features of the contextual cues and the target product share considerable overlap, the product will be categorized as a member of the same category as that activated by the contextual cues. Therefore, assimilation will be more likely when accessible contextual information shares many relevant features with the target, while contrast will be more likely when accessible contextual information shares only a few features with the target. (Herr, Sherman, and Fazio, 1983; Winkielman and Piotr, 1997)

On some occasions, the effect of assimilation and contrast greatly influence the purchase decision of consumers. In this article, I want to interpret how the effect of assimilation and contrast influence the way that consumers consider the attributes of the products.

When consumers are faced with a product category, which includes two products with different attribute values, they will try to determine the appropriate exchange rate between these attributes. As to the example above, considering the expensive ticket with a less good seat, the consumer will try to determine whether a negative rating on attribute "cost" can be exceeded by positive ratings on attribute "seat" as compared to the other ticket. The consumer will choose this ticket if the answer is "yes" or discard it if the answer is "no".

Suppose ticket 1 is $10 with row 20, and ticket 2 is $30 with row 5. The consumer will try to determine whether $10 more cost can be compensated or even exceeded by the benefit that the consumer subjectively think he or she can get from exchanging the seat at row 20 with that at row 5.

Now let us suppose that before this consumer make the decision, he or she see someone buy a car, which costs that person more than $10000. Can this extreme context information influence the decision this consumer makes? As lots of studies about assimilation and contrast effects indicate, the consumer will categorize "$10" and "$30" as the members of the same category, while contrasting the target "$10" and "$30" with the extreme cost "$10000", which is included in the context information. Thus the cost "$10" or "$30" will be judged as smaller than they "really" are, and the difference between "$10" and "$30" will also be judged as smaller. Therefore, in face of such information, the consumer will be more likely to choose the expensive ticket with better seat.

My hypothesis can be shown as follows:

H: In a choice task, when exposed to extreme context information about one attribute, the consumer will subjectively lessen the difference on that attribute in the choice set and be more likely to choose an option that is inferior on that attribute.

EXPERIMENT 1

Method

The experiment was conducted on undergraduate students in Fudan University. 125 participants were invited to take part in the experiments. Two types of questionnaire are randomly distributed to all of these 125 participants.

Question [1] (N=64)

Imagine that you are about to purchase a book for $49, just as you are waiting in the line, a classmate informs you that the book you wish to buy is on sale for $35 at another store, located 15 minutes’ ride away. Now it’s your turn to pay for the book, would you just buy that book in this store or go to another store to buy a cheaper book?

A. buy the book immediately B. go to buy a cheaper book

Question [2] (N=61)

Imagine that you are about to purchase a book for $49, just as you are waiting in the line, you come across a classmate who just come to the store buy textbooks for the whole class. He brings out $50000 from the bag and gives them to the salesman. It takes quite a while before the salesman counts such a large amount of money. This classmate tells you that the book you wish to buy is on sale for $35 at another store, located 15 minutes’ ride away. Now it’s your turn to pay for the book, would you just buy that book in this store or go to another store to buy a cheaper book?

A. Buy the book immediately B. Go to buy a cheaper book

Result and discussion

This experiment was a 2 by 2 design. One factor was the presence or absence of someone else buying a large quantity of the books. The other factor was the difference between the two options (take a 15 minutes’ ride to buy the book for $35 vs. buy the book for $49 with no effort ). It was so designed to test whether participants will be more likely to buy the book immediately when faced with a classmate buying books in a large sum. Considering option B, the participants would try to determine whether the cost can be saved from B will compensate or exceed the time and strength of 15 minutes’ ride.

When faced with a classmate buying books for $50000, the attribute cost "$49" and "$35" might be subjectively contrasted with "$50000". Thus, the participants might judge it smaller than it really is. Then the difference on the attribute cost might seem smaller and made participants feel gain less and be more likely to choose A rather than B in question [2].

However, the result was somehow in contradiction to this hypothesis. The following tabulation summarizes the results.

                                        condition 1         condition 2

                                           (n=64)              (n=61)

Option

A (49;No effort)                28 (44%)           17 (28%)

B (35;15 minutes’ ride)     36 (56%)             44 (72%)

In question [1], 44% of the 64 respondents chose A. As to question [2], only 28%of the 61 respondents chose A. As the result indicates, more people chose A in question [1] compared to question [2], contrary to my hypothesis. (x²=3.419, p>0.05 )

There is an explanation that can interpret how such result came out. Probably a consider number of respondents might have successfully guessed the intention of the experiment designer when asked to make a choice in question [2]. We can find that it is really not difficult for a respondent to do so.

Obviously, the context information is not relevant to the choice set, so probably the respondents would have thought why such irrelevant context information was given in the question [2]. In this way, the context information might trigger specific kinds of guesses on the part of the participants and then guesses might affect their responses to the choices. Thus, some of them would intentionally choose B when they guessed that the experimenter was intended to let them choose A.

Such explanation was to some extent proven by later studies on the participants. When asked whether they know the intention of the experiment designer, some of participants said "yes". Although some of these participants didn’t admit that they changed their mind after they successfully guessed my intention, it is still safe to conclude that such conduct may be one of the explanations why considerably more of them will choose B in question [2].

EXPERIMENT 2

In this experiment, we made efforts to improve the context information. Two types of questionnaires were made and were randomly distributed to participants.

Method

The experiment was conducted on undergraduate students from Fudan University and Shanghai International Studies University. 184 participants were invited to complete one questionnaire. Two types of questionnaires were randomly distributed to these 184 participants.

Question [1] (N=94)

Imagine that you come to a restaurant to have a dinner. When you just sit down, a waiter comes up to you with a delicate piece of pizza and asks you if that is what you have ordered. You say no but ask the waiter how much it costs. The waiter tells you it costs $20. After the waiter goes away, another waiter comes up to you to let you take the order. There are two available types of beefsteak, one costs $40 and the other $80, the more expensive beefsteak will be more tasty. Which types of beefsteak will you choose?

A. One that costs $80 B. One that costs $40

Question [2](N=90)

Imagine that you come to a restaurant to have a dinner. When you just sit down, a waiter comes up to you with a delicate bottle of wine and asks you if that is what you have ordered. You say no but ask the waiter how much it costs. The waiter tells you it costs $200. After the waiter goes away, another waiter comes up to you to let you take the order. There are two available types of beefsteak, one costs $40 and the other $80, the more expensive beefsteak will be more tasty. Which types of beefsteak will you choose?

A. One that costs $80 B. One that costs $40

Result and discussion

This experiment was a 2 by 2 design. One factor was the different feedback the participants may have when exposed to different context information (the presence of a delicate pizza that costs $20 vs. the presence of a delicate bottle of wine that costs $200). The other factor was the difference between the two options (beefsteak that costs $80 and more tasty vs. beefsteak that costs $40 and less tasty). Considering the option A, the participants would try to determine whether the negative ratings on attribute cost can be compensated or exceeded by the positive ratings on attribute taste.

After seeing a delicate wine which costs $200, the participants might have a contrast effect on attribute cost, thus subjectively lessening the difference on the attribute cost. In this way, participants might be more likely to choose A rather than B in question [2].

On the other hand, after seeing a delicate piece of pizza which costs $20, participants might categorize "$40" and "$20" into the members of the same category, while contrasting "$80" with "$40" and "$20". In this way, the participants would judge the difference on the attribute cost larger than it really is. Participants might be more likely to choose B rather than A in question [1].

Thus the effect of assimilation and contrast would make people be more likely to choose A in question [2] as compared to question [1].

The result confirmed the hypothesis. The following tabulation summarizes the results.

                                condition 1        condition 2

                                  (n=94 )             (n=90)

Option

A ($80;more tasty)     45 (48%)         61 (68%)

B ($40;tasty )             49 (52%)         29 (32%)

In question [1], 48% of the 94 respondents chose A. As to question [2], 68% of the other 90 respondents chose A. As the result indicates, the number of people who chose A in question [2] is significantly greater than that in question [1]. (x²=7.46, p<0.01)

GENERAL DISCUSSION

In this article, we studied assimilation and contrast effects on consumer choice. Although experiment 1 failed to find these effects, experiment 2 successfully confirmed the hypothesis.

It is interesting to think about why such effects were not apparent in experiment 1. As was suggested, the disappointing result of experiment 1 might be attributed to specific kinds of guesses on the part of participants. In experiment 1, experimental subjects were exposed to extreme context information that is actually irrelevant to the choice set. This information might trigger some guesses on participants and adversely affected their responses to the choices. Thus, experiment 1 can probably be improved.

In this way, experiment 2 has actually taken more consideration of the context information. In a restaurant, it is no surprise tat the waiter may deliver the wrong food to the guest by mistake. Thus, it may be more understandable to let participants be exposed to such information than to a person buying books for a large sum of money in the book- store, since the latter rarely happens.

Assimilation and contrast effects were successfully found in experiment 2. An important implication of these findings is that they demonstrate the value of context information the consumers are exposed to before they make purchase decisions. Marketers should strive to let consumers get away from extreme information about price if they want to attract consumers by low price. Thus, in supermarket, the products at discount or products that are completely priced should not be put beside those expensive products such as computers or cars.

These findings are also good explanations for why the sale of cheap products will be greatly improved if the shopping mall sales expensive products at discount. After consumers are attracted to buy expensive products, contrast effect will let them judge the price of cheap products less than they really are. Thus, these products will be more likely bought if they meet the demand of consumers.

REFERENCES

Brett Hart and Virginia A. Diehl (1994). "Position Reversal: Isolating The Key Factor In Assimilation and contrast." Journal of Psychology, Jan 94, Vol. 128 Issue 1, p71.

Daly, J. A., & Dickson-Markman, F. (1982). Contrast effects in evaluating essays. Journal of Educational Measurement, 19, 309-316.

Diederik A. Stapel and piotr (1998). "Assimilation And Contrast As A Function Of Context-target, Similarity, Distinctness, and Dimensional Relevance." Personality & Social Psychology Bulletin, Jun98, Vol. 24 Issue 6, p634.

George Loewenstein and Daniel Adler (1995). "A Bias In The Prediction Of tastes". The Economic Journal, 105 (July), 929-937.

Henry Assael (1987), "consumer behavior and marketing action", third edition, kent publishing company, 1987.

Herr, Paul Michael(1983), "Judgmental And Behavioral Consequences Of Category Activation". Dai-B 44/12, p. 3978, Jun 1984.

Herr, Paul Michael (1986). Consequences of priming: Judgment and behavior. Journal of Personality and Social Psychology, 51, 1006-1015.

Huber, Joel and John McCann (1982), "The Importance of Inferential Beliefs on Product Evaluation," Journal of Marketing Research, 19(August), 324-333.

Joan Meyers-Levy and Brian Sternthal (1993). "A Two-factor Explanation Of Assimilation and Contrast effects". Journal of Marketing Research, Aug93, Vol. 30 Issue 3, p359.

Ramadhar Singh, Wan Mei Choo and Li Li Poh (1998). "In-group Bias And Fair-Mindedness As Strategies Of Self-Presentation in intergroup Perception". Personality & Social Psychology Bulletin, Feb98, Vol. 24 Issue 2, p147.

Simonson,Itamar and Tversky,Amos (1992). "Choice in context: Tradeoff contrast and extremeness aversion. " Journal of Marketing, Aug92, Vol.29 Issue 3, p281.

Steven C. Clark and Victor A. Benassi (1997). "Contrast, Assimilation, And Base Rate Effects: Reconsideration Of The Manis-Paskewitz Judgment Model". Journal of Social Behavior & Personality, Mar97, Vol. 12 Issue 1, p1.

Thompson and John Edward (1983). "Assimilation And Contrast Effects in The Judgment Of Outgroups: An Application Of Social Judgment Theory To Intergroup Relations." Dai-B 44/5, p. 1644, Nov 1983.

----------------------------------------