Special Session Summary Whatcha Thinking? Mental Simulation in Consumer Contexts

Jennifer Edson Escalas, University of Arizona
Mary Frances Luce, University of Pennsylvania
[ to cite ]:
Jennifer Edson Escalas and Mary Frances Luce (2003) ,"Special Session Summary Whatcha Thinking? Mental Simulation in Consumer Contexts", in NA - Advances in Consumer Research Volume 30, eds. Punam Anand Keller and Dennis W. Rook, Valdosta, GA : Association for Consumer Research, Pages: 213-215.

Advances in Consumer Research Volume 30, 2003     Pages 213-215



Jennifer Edson Escalas, University of Arizona

Mary Frances Luce, University of Pennsylvania

The three papers in this session present experimental consumer research using mental simulation. Mental simulation is the imitative mental representation of some event or series of events (Taylor and Schneider 1989). The first paper, "Process vs. Outcome Thought-Focus and Advertising," (by Jennifer Edson Escalas and Mary Frances Luce) examines mental simulation in a print ad context and varies the focus of simulation on either the process of using the advertised product or the outcomes or benefits of using the product. The authors find that process- (versus outcome-) focused thought results in greater sensitivity to argument strength, but that this effect is reversed under high involvement; they conclude that process-focused thought involves qualitatively different cognitive processes than those typically studied in research on dual-processing theories (e.g. Elaboration Likelihood Model). The second paper, "SelfBOther Discrepancies in Counterfactual Mental Simulation," (by Parthasarathy Krishnamurthy and Anuradha Sivaraman) examines the structure of counterfactual thoughts, a form of mental simulation, focusing on how self- versus other-focus alters the process, and therefore the effect, of mental simulation. They conclude that we undo events involving others differently than those involving ourselves. The third paper, "The Impact of Simulating Satisfaction on Product Returns" (by Baba Shiv, Joel B. Huber, and Himanshu Mishra) finds a reduction in decision reversals when choice is preceded by anticipation of satisfaction compared to a "normal" (decision-focused) choice scenario. James R. Bettman (Duke University) acted as the discussant and proposed a "dual process" view of mental simulation based upon recent work in cognitive neuroscience and Tulving’s distinction between autonoetic consciousness (which is more "episodic" in nature) versus noetic consciousness (more "semantic").



Jennifer Edson Escalas, University of Arizona

Mary Frances Luce, University of Pennsylvania

Recent social psychological research demonstrates that certain types of mental simulation are particularly useful for helping individuals reach the future they envision (e.g., Taylor et al. 1998). More specifically, Taylor’s recent research indicates that the most successful simulations focus on the process of reaching a goal rather than on the outcomes or attainment of the goal. The purpose of our research is to explore the distinction between outcome- and process-focused mental simulations as a possible tool for designing persuasive advertisements.

Experiment 1 manipulates the focus of participants’ thoughts with pre-ad instructions in the context of a fictitious vitamin print advertisement. We find that process-focused thought results in significantly higher behavioral intentions than outcome-focused thought when ad arguments are strong. However, in the case of weak ad arguments, process-focused thought actually lowers behavioral intentions, compared to thoughts focusing on the end result or outcome of product usage. Thus, in addition to increasing the persuasive power of strong advertising claims, process-focused thinking appears to make individuals better, or more discerning, consumers who do not form behavioral intentions when it is inappropriate to do so.

We believe that the underlying explanation for these effects is plan formation, that is, participants in the process-focus condition create a step-by-step plan to achieve a favorable outcome. By imagining using the vitamins on a daily basis, it appears that participants visualize the behavioral steps necessary to achieve the benefits of using vitamins. Therefore, they are more willing to take these behavioral steps, and indicate greater intentions to buy the vitamins under conditions of strong arguments. We also believe that process-focused thought results in lower behavioral intentions under weak arguments because of associated planning processes. When the behavior’s ability to achieve the outcome is weak, consumers may be unable to formulate an actionable plan or they may more generally recognize the disadvantages of the plan when its formulation is encouraged by process-focused thought. When consumers reject developing a plan involving the behavior and outcome, significantly lower behavioral intentions result.

While Experiment 1 extends the social psychological research findings to an advertising context, the ad itself in that study does not influence the focus of participants’ thoughts; our manipulation consists of pre-ad instructions. In our second experiment, we examine a potential mechanism for influencing thought-focus within the ad itself: narrative advertising, that is, ads that tell stories. Narrative advertisements are well suited to influence the extent to which viewers focus on the process of using a product, as opposed to the outcome of using the product, because narratives consist of goal-oriented action sequences incorporating goals, processes, and outcomes.

In Experiment 2, we are able to manipulate participants’ thought-focus by modifying the scenes in a storyboard advertisement. A story emphasizing the process of using the product results in more process focused-thought than a story emphasizing usage outcomes. Furthermore, participants who saw a process-focused narrative exhibited higher ad and brand attitudes and greater behavioral intentions to purchase the advertised product (when ad arguments are strong). While watching a narrative ad that emphasizes the process of using a fictitious shampoo brand, it appears that participants think about the behavioral steps necessary to achieve the benefits of using the shampoo, thus creating a plan. Therefore, they are more willing to take these behavioral steps, and indicate higher attitudes and greater intentions to buy the shampoo.

Experiment 2 also examines the interactive effect of argument strength and thought-focus. Participants are more sensitive to argument strength while viewing a process-focused narrative ad than they are while viewing an outcome-focused narrative ad. Again, we believe that process-focused thought results in more sensitivity to argument strength because of the associated planning processes. When the behavior’s ability to achieve the outcome is weak, consumers may become particularly sensitive to the disadvantages of the arguments for the advertised behavior, resulting in significantly lower behavioral intentions.

Finally, Experiment 3 attempts to rule out elaboration as a potential alternative explanation for our process-focus results. This study crosses the process-outcome thought focus, strong-weak arguments conditions with a third factor: accuracy instructions (versus a control group). We find that the thought focus by argument strength interaction found in E1 and E2 replicates in the control group, but not under explicit instructions encouraging elaboration. Thus, we believe that the advantages of process-focused thought (when observed) result from relatively natural, spontaneous planning processes rather than resulting from an increase in analytical message processing. Mediation analyses using a variable reflecting "degree of planning" further support this account. Additional analyses indicate that the amount of thought does not mediate the effects of process- versus outcome-focused thoughts across any of our studies. Thus, this effect appears to be a function of the content, rather than the amount, of thought.


Taylor, Shelley E., Lien B. Pham, Inna D. Rivkin, and David A. Armor (1998), "Harnessing the Imagination: Mental Simulation, Self-Regulation, and Coping," American Psychologist, v. 53, n. 4, pp. 429-439.

Taylor, Shelley E. and Sherry K. Schneider (1989), "Coping and the Simulation of Events," Social Cognition, v. 7, n. 2, pp. 174-194.



Parthasarathy Krishnamurthy, University of Houston

Anuradha Sivaraman, University of Houston

Counterfactual thinking is a process of mentally simulating a past experience with a view to undoing the outcome of the experience. Roese and Olson (1993) introduced the idea of understanding counterfactuals by investigating their structure. They suggest that the structure of counterfactual mental simulations will be influenced by the valence of the outcome that is being undone. Roese and Olson (1993) find that additive counterfactuals, which involve undoing outcomes by changing inactions to actions, or more generally, adding new aspects to an experience, are more likely in response to failure. Subtractive counterfactuals, which involve undoing outcomes by changing actions to inactions, or more generally, removing aspects from an experience, are more likely in response to success.

Our research juxtaposes research on the structure of counterfactuals (Roese & Olson, 1993) with a classic dichotomy in social cognition, selfBother discrepancy. This discrepancy has been demonstrated to moderate the process of mental simulation (e.g., Anderson 1983). he principal research issue is as follows: What are the specific differences in the structure of simulated counterfactuals generated in response to self-focused versus other-focused experiences? There are two reasons why we undertook this research. First, counterfactuals generated in response to others’ experiences are likely just as pervasive as those involving one’s own experiences. Indeed, undoing others’ actions and inactions to speculate on how outcomes could have been different is an integral part of social discourse (a.k.a., Monday morning quarterbacking). Second, as consumers, we often learn from others’ experiences just as much as we do from our own. Given the findings that counterfactual thinking influences attitudes and behavioral intentions (Krishnamurthy & Sivaraman, 2002), it stands to reason that selfBother discrepancies in the nature of counterfactual thinking may differentially influence attitudes and intentions in regard to such experiences. Third, the selfBother contrast in counterfactual thinking is structurally similar to the actorBobserver contrast observed in attribution theory. An additional goal of our research is to assess the extent to which counterfactual thinking and attribution share the same underlying process. Specifically, we were interested in the actor observer differences in the fundamental attribution error in which actors make person focused attribution for positive outcomes and situation focused attributions for negative outcomes, while the reverse pattern is noted for observers.

Experiences involving one’s self versus those involving others can be thought of in terms of actorBobserver differences. Actors have access to inner thoughts and feelings, which often generate the rationale for a given decision. Observers do not have access to these thoughts. These thoughts and feelings constrain the actor to a smaller set of actions or inactions that can be undone. Thus, the barriers for generating counterfactuals will be in direct proportion to the amount of information one possesses about the reasons for the actions and inactions that preceded the outcome. Given that such information is available readily in regards to one’s own actions and inactions, it should be more difficult to undo one’s own actions and than to undo others’ actions.

Our expectation is that, the pattern of results observed by Roese and Olson (1993)Cmore subtractive counterfactuals following success and more additive counterfactuals when undoing failureCshould be less pronounced for self-focused episodes than for other-focused episodes.

In Study 1, 140 undergraduate students participated in a 2 (Focus: Self/Other) x 2 (Outcome Expectation: High/Low) x 2 (Outcome Valence: Positive/Negative) between subjects design similar to Roese and Olson (1993) with the additional manipulation of focus of reference. Participants were given one of eight hypothetical scenarios in which a student (oneself) with a good (poor) standing in a class goes through a series of events and eventually passes (fail) the exam. Consistent with prior research, outcome expectations were manipulated by varying the current standing in the imaginary class. Participants were instructed to indicate how the outcome could have been different. The principal dependent variable was the number of additive and subtractive counterfactual thoughts in the protocol, coded using procedures outlined by Roese and Olson (1993).

The principal finding was that the effects observed in the other-focus condition were either greatly reduced or absent in the self-focus condition, comporting with our expectation that self-focused episodes present greater barriers to counterfactual thinking than other-focused episodes.

In study 2, we employed the same design as study 1 with the following exceptions. First, participants were run in pairs, with one person narrating a consumption experience and the other person listening to the experience, thus instantiating a self versus other manipulation. The valence of the consumption episode, positive versus negative, was manipulated on a between subjects basis. Third, we assessed how consumers’ attitudes and behavioral intentions are influenced by counterfactuals generated in response to consumption episodes involving themselves versus that of others. Recent research indicates that counterfactual thinking in response to failures promotes intention to perform success facilitating behaviors and affects attitudes toward subsequently encountered messages (Krishnamurthy & Sivaraman, 2002).

Although study 1 indicated a moderating effect of focus of reference on structure of counterfactual simulation, the results of study 2 did not indicate such an effect. However, we found an effect of the focus of reference on the focus of counterfactual thoughts. Overall, the results for person versus situation-focused counterfactuals were different from the pattern suggested by the literature on actor-observer differences in the fundamental attribution error. This suggests that the process of counterfactual simulation may be different from the process that generates attributions.

The results for attitude and behavioral intention indicated that valence had a stronger effect on these measures in the self-focused conditions than in the other-focused conditions. This is inconsistent with our expectations that the effect of the valence will be stronger for other-focused experiences.

In summary, even if the structure of counterfactuals are unaffected by the focus of reference, the focus of counterfactuals are. Furthermore, the effect is different from what would be anticipated by the literature on attribution theory, indicating that counterfactual simulation and attribution are distinct processes.


Anderson, Craig A. (1983), "Imagination and Expectation: The Effect of Imagining Behavioral Scripts on Personal Intentions," Journal of Personality and Social Psychology, v. 45, n. 2, pp. 293-305.

Krishnamurthy, Parthasarathy and Anuradha Sivaraman (2002), "Counterfactual Thinking and Advertising Responses," Journal of Consumer Research, 28 (March), 650B658.

Roese, Neal J., and James M. Olson (1993), "The Structure of Counterfactual Thought," Personality and Social Psychology Bulletin, 19 (June), 312B319.



Baba Shiv, University of Iowa

Joel Huber, Duke University

Himanshu Mishra, University of Iowa

During shopping, consumers often arrive at their final choices with quite different decision goals in mind. For example, the consumer’s goal could be choice-oriented (e.g., "Which item do I pick?") or it could be value-oriented (e.g., "How much am I willing to pay for this item?") or it could be anticipated-satisfaction-oriented (e.g., "How satisfied am I going to be with the products I am considering?"). Previous research has shown that the type of goal the consumer uses can have a considerable effect on what s/he ends up choosing (e.g., Nowlis and Simonson 1997; Shiv and Huber 2000). These shifts in preferences have been shown to arise due to different psychological processes that are engendered with these different goals. For example, the findings in Shiv and Huber (2000) suggest that shifts in preferences between choice and anticipated-satisfaction goals arise because, compared to choice, anticipated satisfaction results in a processing strategy that involves mental simulations focused on imagining one or more options. Consequently, in the presence of an anticipated-satisfaction goal, attributes such as "leather trim" and "sunroof," that matter more for the long-term satisfaction one derives from owning a car, are likely to become focal and have a bigger impact on the construction of preferences compared to attributes such as "price." The consumer, therefore, is more likely to choose a car with these attributes than in the presence of a choice-oriented goal or a value goal.

We focus on a different research question in this research. Rather than focusing on the effects of decision goals on choice at the point of purchase, we ask the question, How are product returns affected by mental simulation generated in response to pre-purchase goals of anticipated satisfaction versus choice and value? In other words, Will these goals differentially affect the extent to which consumers change their minds after they have made their initial choices? By addressing this question, our goal is to provide retailers with a way of minimizing product returns, an issue that seems to have become a matter of considerable concern for retail stores in recent years.

To address our central research question, we carried out a lab experiment that used a single-factor (decision-goal: anticipated-satisfaction vs. choice vs. value) between-subjects design. The experiment was carried out in two different rooms. The procedure was designed to mimic the multi-phase decision making that often occurs in case of infrequently purchased items: The decision-phase (where the choices are tentative), the period of non-commitment (the duration of which is often determined by the store’s return policy), and the period of commitment (where the decision if final). In the first room, respondents went through the decision phase by visiting a mock Best-Buy web-site, where they received information on two handhelds, Sony Clie T615 with color screen and priced at $249, and Sony Clie SJ20 with monochrome screen and priced at $199. Except for differences on the nature of the screen and the prices, the two models were similar to one another. Respondents were told that they would first make a tentative choice between two options, the T615 priced at $249 and the SJ20 priced at $199, plus $50 in cash. They would then go over to a second room, one person at a time, and use their chosen option for as long as they wished. This was designed to mimic the period of non-commitment. They would finally make up their minds on which option to choose (period of commitment). Participants were also told that by taking part in the study, they would be entered into a lottery. The winner of the lottery will get a prize worth $249, which will be in the form of the T615 if that’s what s/he finally ends up choosing, or the SJ20 plus $20 in cash, if that’s what s/he finally ends up choosing. Decision-goal was manipulated in the first room as in Shiv and Huber (2000). One group of respondents chose immediately after acquiring information about the two handhelds. Another group rated their anticipated satisfaction with each of the two alternatives on three 1-7 scales anchored on dissatisfied/satisfied, unhappy/happy, and feel bad/feel good. They then made their choices. A third group rated the value associated with the two alternatives on three 1-7 scales anchored on a bad deal/a good deal, a bad value/a good value, and a bad buy/a good buy. They then made their choices. After making their choices, respondents indicated the extent to which they engaged in mental imagery and the extent to which they elaborated.

The extent to which participants changed their minds from the earlier choices was then examined. Choice of Sony Clie T615 was significantly higher in the anticipated-satisfaction condition (64.7%) than in the choice (38.4%, z=2.24, p<.05) and value conditions (35.3%, z=2.43, p<.05). In phase 2 of the experiment, choice of Sony Clie T615 remained unchanged in the anticipated-satisfaction condition (79.4% and 64.7% in phase 2 and phase 1, respectively, z=1.34, p>.20). But in the choice and value conditions, the shifts in choices between phase 2 and phase 1 were significant. Choice of Sony Clie T615 changed from 38.4% to 82% (z=3.92, p<.001) in the choice condition, and from 35.3 t 82.4% in the value condition (z=3.96, p<.001). We propose that the reason for the "change of minds" that occurred in the choice and value conditions is that choice-oriented and value-oriented goals cause people to become myopic and put more weight on attributes that matter for the short haul (e.g., price). After the initial decision, price becomes less salient and vivid attributes, such as the screen-type in a PDA tend to become more salient. This shift in salience causes vivid attributes to be weighted more heavily after the initial decision, leading to the "change of minds." On the other hand, an anticipated-satisfaction goal causes people to focus more on their long-term satisfaction while making their initial decision. After the initial decision, when they think more about their choices, they see no need to change their minds and, therefore, continue to stick with their initial choices. Thus, we have some evidence that when people change the focus of their thought to be more consistent consistent with long-term usage considerations, they may become better (or at least more consistent) decision makers. Supplementary analyses indicate that the process driving our effects involves the focus of thought (e.g. the nature of mental simulation), rather than simply the amount of thought (i.e, elaboration).


Nowlis, Stephen M. and Itamar Simonson (1997), "Attribute-Task Compatibility as a Determinant of Consumer Preference Reversals," Journal of Marketing Research, 34 (May), 205-18.

Shiv, Baba and Joel Huber (2000), "The Impact of Anticipating Satisfaction on Choice," Journal of Consumer Research, 27 (September), 202-216.