<%BANNER%>

Changing Standards: Adjusting the Definition of an Acceptable Outcome


PAGE 1

CHANGING STANDARDS: ADJUSTING THE DEFINITION OF AN ACCEPTABLE OUTCOME By MEREDITH LEA TERRY A THESIS PRESETNED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

ACKNOWLEDGMENTS I am indebted to many people for providing support and encouragement throughout my work on this thesis. First, I want to thank my advisor, Dr. James Shepperd, who gave so much of his time, ideas, and energy to this project. Additionally, the thoughtful comments of my other committee members, Dr. Benjamin Karney and Dr. Mark Fondacaro, have contributed to this thesis. I am grateful to my parents who have always supported and fostered my interest in learning. The support and encouragement provided by my parents and sister helped me complete this project. Finally, I have had the opportunity to work with some exceptional graduate students in my time at the University of Florida, especially those in my lab, Patrick Carroll, Kate Dockery, and Jodi Grace. I appreciate all of the feedback, advice, and friendship that they have provided. ii

PAGE 3

TABLE OF CONTENTS page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES...............................................................................................................v CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW....................................................1 Changing Standards after Feedback Is Known.............................................................2 Changing Standards......................................................................................................3 Bracing..........................................................................................................................5 2 STUDY 1......................................................................................................................8 Overview and Hypothesis.............................................................................................8 Method..........................................................................................................................9 Participants............................................................................................................9 Procedures...........................................................................................................10 Results.........................................................................................................................13 Comparing Standards across Time......................................................................13 Comparing Estimates across Time......................................................................14 Comparing Estimates to Standards......................................................................14 Comparing Changing Standards to Bracing........................................................15 Assessing the Accuracy of Participants Standards and Expectations................16 Comparisons to Participants in Hypothetical Interactions..................................18 Discussion...................................................................................................................19 3 STUDY 2....................................................................................................................35 Overview.....................................................................................................................35 Method........................................................................................................................35 Participants..........................................................................................................35 Procedures...........................................................................................................35 Hypothesis...........................................................................................................37 Results.........................................................................................................................37 Differences between standards and expectations................................................37 iii

PAGE 4

Comparing Expectations across Time.................................................................38 Comparing Standards across Time......................................................................39 Comparing Changing Standards to Bracing........................................................40 Assessing the Accuracy of Participants Standards and Expectations................41 Ideals....................................................................................................................42 4 GENERAL DISCUSSION.........................................................................................55 APPENDIX DEPENDENT VARIABLES.............................................................................................62 LIST OF REFERENCES...................................................................................................65 BIOGRAPHICAL SKETCH.............................................................................................67 iv

PAGE 5

LIST OF TABLES Table page 1 Study 1: Average estimate of what determines a good outcome for each of the dimensions................................................................................................................24 2 Study 1: Standards and expectations from Time 1 to Time 2..................................25 3 Study 1: Correlations between standards and expectations at Time 1 and Time 2......................................................................................................................26 4 Study 1: Standards versus expectations at Time 1 and at Time 2............................27 5 Study 1: Number of participants who set standards higher, equal to or lower than expectations..............................................................................................................28 6 Study 1: Correlations between the change in standards and the change in expectations..............................................................................................................29 7 Study 1: Standard and received outcomes at Time 1 & Time 2...............................30 8 Study 1: Mean expected and received outcomes at Time 1 & Time 2.....................31 9 Study 1: Number of participants who were pessimistic, accurate and optimistic....32 10 Study 1: Mean expected outcome at Time 1 and 2 and hypothetical standard (minimum acceptable outcome for survey)..............................................................33 11 Study 1: Mean desired outcome (hypothetical survey) and mean standard (minimum acceptable outcome) at Time 1 & Time 2..............................................34 12 Study 2: Standards versus expectations at Time 1 and at Time 2 when standards and expectations are counterbalanced......................................................................43 13 Study 2: Expectations at Time 1 and Time 2 by condition.....................................44 14 Study 2: Standards at Time 1 to Time 2 by condition..............................................45 15 Study 2: Correlations between standards and expectations at Time 1 and at Time 2......................................................................................................................46 v

PAGE 6

16 Study 2: Correlations between the change in standards and the change in expectations..............................................................................................................47 17 Correlations between the change in standards and the change in expectation across both Study 1 and Study 2 (Total N=81)........................................................48 18 Study 2: Number of participants changed their standards the same amount that they change their expectations.................................................................................49 19 Study 2: Correlations between anxiety and change in standards & anxiety and change in expectations.............................................................................................50 20 Study 2: Standard and received outcomes at Time 1 & Time 2...............................51 21 Study 2: Mean expected and received outcomes at Time 1 & Time 2.....................52 22 Study 2: Standard and ideal outcomes at Time 1 & Time 2....................................53 23 Study 2: Mean expected and ideal outcomes at Time 1 & Time 2..........................54 vi

PAGE 7

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science CHANGING STANDARDS: ADJUSTING THE DEFINITION OF AN ACCEPTABLE OUTCOME By Meredith Lea Terry May 2004 Chair: James Shepperd Major Department: Psychology After receiving an unsatisfactory outcome, people often redefine what they consider an acceptable outcome. In addition, in anticipation of feedback people often become pessimistic in their predictions for the future, bracing for bad news. The current studies examined a) whether people would redefine what they consider an acceptable outcome prior to receiving feedback and not simply in response to bad news, and b) how personal standards compare with expectations and actual outcomes. In Study 1, participants interacted individually with 10 peers and then rated each peer on a variety of dimensions. Participants also reported what outcome would be acceptable (standard) and what outcome they expected (expectation) from their peers. Participants provided these reports twice: immediately after the interaction (Time 1), and again 30 minutes later, when participants expected their peer evaluations (Time 2). Participants reported lower standards and expectations at Time 2 than at Time 1. In Study 2, half the participants were assigned to a no feedback condition, believing they would not receive the vii

PAGE 8

evaluations from their peers. Also in Study 2, the order of the items assessing standards and the items assessing expectations were counterbalanced. In Study 2, for each of the five evaluative domains, participants also reported what they viewed as the ideal outcome. In Study 2, only participants expecting to receive peer evaluations reported lower standards and expectations at Time 2 than Time 1. In addition, participants ideals exceeded their standards and expectations. Analyses reveal changes in expectations are distinct from changes in standards. viii

PAGE 9

CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW There is nothing either good or bad, but thinking makes it so. William Shakespeare, Hamlet Prince of Denmark (Hamlet at II, ii) At first blush, distinguishing between what constitutes a good versus a bad outcome would seem to be a relatively easy task. A good outcome is getting a promotion, getting accepted into a competitive graduate program, or learning that a suspicious lump is benign. A bad outcome is being passed over for a promotion, getting rejected from a competitive graduate program, or learning that a suspicious lump is cancer. When outcomes fall on a continuum, however, what people define as good versus bad, satisfactory versus unsatisfactory, or success versus failure, is not always so clear and often is a matter of perspective. As Hamlet wisely noted, in many instances outcomes are neither inherently good nor inherently bad. Rather, perceivers impose these evaluative meanings. Hamlets insights emerge from recent upheavals in his family life and reflect his new regard for Denmark as a prison. This paper acknowledges Hamlets insights but moves beyond merely recognizing that peoples evaluations are a product of their current circumstances. This paper explores instead how evaluations change in anticipation of outcomes. Specifically, this research examines whether people change their definition of what constitutes an acceptable outcome over time as information bearing on the outcome draws near. 1

PAGE 10

2 Changing Standards after Feedback Is Known Research from a variety of theoretical traditions reveals that, after learning an outcome, people sometimes change their standards of what constitutes a satisfactory outcome to make the outcome more acceptable. Most notably, cognitive dissonance theory posits that when people experience inconsistency between their thoughts and actions, they are motivated to reduce the inconsistency (Festinger, 1957). The theory states that when people think they behaved in a way that is contrary to their self-views, they experience an uncomfortable feeling called dissonance. One way to remove the dissonance is to change how the outcome or behavior is evaluated. Specifically, if people can rationalize that the behavior is not contrary to their self-views, then the dissonance is alleviated (Festinger, 1957). In other words, people may change what they define as an acceptable behavior or outcome as a means of reducing dissonance. Likewise, when faced with a feeling of regret for a past action, the regret may prompt dissonance reducing behaviors (e.g. a reevaluation of the past behavior) to reduce the pain associated with regret (Gilovich & Medvec, 1995). The idea that people will change their standards of what constitutes a satisfactory outcome has received support in research on management. For example, one way that managers can contend with criticism is to change the criteria of what is versus is not important (Zemke, 1981). Evaluating the criticism as unimportant makes the feedback closer to what the managers had anticipated. For example, Zemke notes that managers will sometimes acknowledge the validity of criticism they receive yet nevertheless insist that their management style is the preferred way to run the company (Zemke, 1981). Similar evidence comes from research on response shift. Research on response shift was inspired in part by the observation that people with severe chronic illnesses

PAGE 11

3 report no less quality of life than do people without such illnesses (Breetvelt & Van Dam, 1991). Moreover, the quality of life they report is stable (Bach & Tilton, 1994). Apparently, chronically ill patients shift their definition of what it means to be satisfied with their health when they experience a change in their health. This research reveals that patients can maintain consistent satisfaction with their health in the face of actual declines in health a) by changing the standards they use for comparison, b) by changing their personal values, and c) by changing how they define the outcome (Sprangers & Schwartz, 2000). Sprangers and Schwartz (2000) provide the example of a woman with bone cancer who responds to her deteriorating health by changing what it means to be satisfied with life. When she becomes confined to a wheelchair she reports that she will be satisfied with her health as long as she is not bedridden. When she becomes bedridden, she reports she will be satisfied with her health as long as she can still socialize with friends. Changing Standards Prior research demonstrates that people may reevaluate their outcomes after receiving feedback (i.e., the managers receiving negative feedback, feelings of guilt or dissonance associated with past behaviors) or in response to new information (i.e., changing health status). We explored the possibility that people will change what constitutes an acceptable outcome prior to receiving feedback bearing on that outcome and not simply in response to feedback that is threatening. We introduce the term changing standards to refer to adjusting the definition of what constitutes an acceptable or satisfactory outcome in anticipation of information bearing on that outcome. We believe that changing standards allows a person the benefit of receiving an outcome or feedback that they view as acceptable. By changing standards, people can preemptively avoid

PAGE 12

4 uncomfortable feelings associated with unacceptable feedback. For example, a student about to take a test may only be satisfied by receiving an A. Later, with impending feedback, the student may adjust what is satisfactory from an A to a C+. By changing standards, there is an increase the likelihood that the person will receive an acceptable outcome. People may opt to proactively change their standards of what constitutes an acceptable outcome instead of readjusting their evaluations after receiving feedback for two reasons. First, William James described self-esteem as the ratio of success to pretensions. That is, self-evaluation reflects a comparison of outcomes to their wants and goals (James, 1890). Decreasing 'pretensions' (the wants and goals) is one way to increase self-esteem, and people may preemptively change their standard to ensure that their outcomes match their pretensions. By lowering the standard of what defines an acceptable outcome, people avoid the negative affect and negative self-evaluations that arise from receiving outcomes that are below their aspirations. A second reason people may be motivated to change their standards is to accurately predict and control their world (Pervin, 1963). As such, people may preemptively lower their standard of what constitutes an acceptable outcome so they can predict their world. Failing to change the standard and then receiving an outcome that is viewed as unacceptable may be interpreted as an inability to control and predict future outcomes. Of course, distinct from these reasons is the possibility that people may also change their standards in response to new information. With new information, people may realize that previous standards were unrealistic. Our interest in this paper is in changing standards in anticipation of receiving feedback.

PAGE 13

5 Bracing Changing standards in anticipation of information bearing on the outcome shares some commonality with bracing for bad news. With bracing, people lower their expectations of what is likely to happen in anticipation of feedback (Shepperd, Ouellette, & Fernandez, 1996; Taylor & Shepperd, 1998). One reasons people may lower their expectations is that bracing reduces the likelihood of disappointment should they receive an unexpected outcome. Evidence for bracing comes from a study in which students estimated their exam score prior to receiving it. Although participants were optimistic about the score they would receive three weeks prior the exam, they were realistic immediately after the exam. Moreover, this realism persisted three days later at the beginning of the class in which they would receive their grades. At the end of class, however, as the professor called students names to return their exams, students became pessimistic, estimating a score significantly lower on average than they actually received (Shepperd et al., 1996). Additional evidence for bracing comes from a second study in which college sophomores, juniors and seniors estimated their salary on their first job after college. Whereas seniors estimated a lower starting salary in April (two weeks prior to graduation for the seniors) than in January (three months before graduation), the sophomores and juniors, for whom graduation was over a year away, did not (Shepperd et al., 1996). Changing standards is similar to bracing in many ways. First, they both can have implications for how people feel about future outcomes. When people brace, they are better prepared for a previously unexpected outcome. When people change their standard, they make an unacceptable or negative outcome more acceptable or less negative. Second, both bracing and changing standards involve changing a previously set

PAGE 14

6 benchmark for evaluating outcomes. Third, they both occur in response to potential feedback that may challenge the persons self-concept. However, changing standards is distinct from bracing in that it reflects a shift in what people define as acceptable rather than a shift it what people expect to occur. Although changes in standards and expectations may covary, it is possible for people to shift one but not the other. That is, people may change their expectations but not their standards, or change their standards but not their expectations. For example, a person awaiting test results for a sexually transmitted disease may brace for possible bad news, yet remain steadfast in the belief that testing positive is an undesirable outcome. That is, the expectation may change while the standard remains constant. Conversely, athletes may expect to win a competition, but in the moments prior to the match, regard losing as an acceptable outcome provided they play as hard as they can. That is, the standard may change while the expectation remains constant. Of course, in some instances the two will covary, and may even be causally related. A shift in expectations of what will happen may prompt a shift in standards of what is acceptable. For example, students expecting a low grade in a class may, in response, lower the standard of what is an acceptable grade. Research on bracing shows that people will lower their expectations to prepare for future feedback. Other research shows that after people receive an outcome, they sometimes change their previous evaluations to reduce dissonance or make the new feedback appear more pleasant. But, research has yet to address whether people may alter the standard at which they define an outcome as good or acceptable before they receive feedback. Although people are likely to adjust what they define as an acceptable outcome after receiving feedback (to reduce dissonance, reduce regret, or in response to

PAGE 15

7 new information), the proposed research focuses specifically on the effect of anticipated feedback on peoples definitions of what constitutes an acceptable versus unacceptable outcome.

PAGE 16

CHAPTER 2 STUDY 1 Overview and Hypothesis Study 1 was designed to provide a preliminary test of whether the mere anticipation of outcome feedback prompts people to change their standards of what they define as an acceptable outcome. To this end, we needed an outcome that was important to people, yet also continuous and for which there was no clear cutoff or preexisting definition of what constituted an acceptable outcome. We also needed an outcome that was uncertain, allowing room for people to shift their standards. We opted against adopting the TAA paradigm used in past research paradigm. Although participants clearly cared about whether they tested positive for TAA deficiency, testing positive vs. negative for TAA deficiency is a dichotomous outcome. Likewise, we opted against examining performance on an intelligence test because people often hold pre-existing ideas about what constitutes an acceptable outcome. In the end we developed a new experimental paradigm in which people interacted with and evaluated several peers, and then reported their standards of what would be an acceptable evaluation from their peers. Our paradigm was modeled after a popular format for single people to meet potential dating partners called speed-dating. Participants took part in one of two groups. The first group of participants imagined they met 10 people and then reported what they would regard as an acceptable outcome on a number of dimensions. Our purpose was to determine the baseline for what constitutes an acceptable outcome in a social situation. We were solely interested in the 8

PAGE 17

9 average definitions of an acceptable outcome for each of the dimensions and how those estimates compared with the responses of participants who actually engaged in the interactions. The second group of participants interacted individually with 10 peers then rated each peer on a number of dimensions. For each dimension participants also reported what they regarded as an acceptable evaluation from their peers (standard) and how they expected to be evaluated by their peers (expectation). Participants provided these estimates twice: immediately after interacting with all peers (Time 1), and again just before receiving their peer evaluations (Time 2). We hypothesized that participants would report both a lower expectation and a lower minimum standard for the outcome at Time 2 than at Time 1. We also explored whether people lower their standards or criteria of what defines a satisfactory outcome in anticipation of information bearing on the outcome. Finally, we explored when and how the estimates made by participants who engaged in the interactions compared with the estimates of participants who merely imagined engaging in the interactions. Method Participants The first group of participants comprised Introductory and Upper Level Psychology students (N =50) who completed the hypothetical survey as part of an unrelated study or an in-class project. The second group of participants comprised Introductory psychology students (27 females, 12 males) participated in one of four sessions in partial fulfillment of a research participation requirement and were run in groups of up to 11 people. When fewer than 11

PAGE 18

10 participants arrived for the experiment, one or more research assistants posed as a participant to bring the number to 11. Procedures The first group of participants, the hypothetical survey group, completed a survey in which they imagined that they had just met 10 people. They also imagined that each of the ten people rated them on a number of dimensions. Participants reported what they would define as an acceptable outcome for each of the dimensions. On arriving for the experiment, the second group of participants learned that the purpose of the study was to examine first impressions and that they would interact with and then evaluate the other 10 participants in the study. The second group of participants randomly received nametags each printed with the name of a different color. Each nametag had the name of a color on it. We provided participants with the names of colors rather than their own names for several reasons. First, the colors allowed us to establish in advance a systematic order in which participants would interact and evaluate each other. Second, assigning participants colors preserved the anonymity of participants ratings. Rather than evaluating each person according to name, participants evaluated each person by color. In addition, when making their evaluations, participants identified themselves only by color. After drawing a nametag, each participant received a manila envelope containing the order in which they would interact with other participants and a first impressions form. The experimenter explained that during the course of the experiment, participants would engage in 10 interactions, one with each of the other 10 participants in the study. When the experimenter signaled, participants were to find their first interaction partner and engage in a 2 min conversation.

PAGE 19

11 The experimenter explained that participants could talk about anything that arises naturally in conversation such as hobbies, hometowns, and year in school. However, they were not to talk about the experiment, the people in the experiment, or other experiments in which they had participated. The experimenter further stated that after each interaction period, participants would have 30 s to jot down on the first impressions form notes that would help them remember the person they just met. The experimenter would then prompt them to move to the next interaction partner. The experimenter explained that the notes were to help them evaluate each of their fellow participants when all interactions were complete. Built into the schedule were three breaks for each participant when they could write down additional notes. When it was clear that all participants understood the procedures, the experimenter initiated interaction phase of the experiment. After completing the 10 interactions, participants were seated and received an evaluation questionnaire from the experimenter. The questionnaire first asked participants to rank their fellow participants (using their color names) starting with the person they liked the most and ending with the person they liked the least. Second, participants indicated (yes vs. no) in response to each color name whether they a) found the person friendly, b) would rank the person in their top three in terms of liking, c) would want to spend time with this person in the future, d) would want to learn more about the person, and e) found the person interesting. The experimenter asked participants to respond honestly and assured them that their responses were anonymous. On completion of the questionnaire, participants returned it and their first impressions form to the manila envelope, which was then collected by the experimenter and given to a

PAGE 20

12 second experimenter with instructions to enter the data in the computer. That second experimenter left the room. Participants next received a reflection questionnaire comprising items that examined how participants felt about the interactions and the impressions they made. In the first part, participants responded to five items asking them to indicate the minimum number of participants you would want to report that you a) were friendly, b) were ranked in the person s top three in terms of liking, c) would want to spend time with you in the future, d) would want to learn more about you, and e) were interesting. The instructions directed participants to circle a number from 0 to 10 indicating the number that they defined as acceptable for each of the previous items. This number served as the standard. In the second part, participants responded to same five items. However, rather than reporting what they felt would be acceptable, participants reported what they expected was likely to happen (e.g., How many participants do you predict will want to spend time with you in the future?). Again, participants could circle anywhere form 0 to 10 participants in response to each item. This number served as the expectation. Soon after all participants had completed filler questionnaires, the second experimenter returned and announced, after a brief discussion with the first experimenter, that the evaluation questionnaire data were entered into the computer and that they would learn after all how they were evaluated by their fellow participants. The experimenter then distributed to each participant a sealed manila envelope with his or her color name printed across the top. The experimenter explained that, before participants opened the envelope they would need to complete two more questionnaires. The first questionnaire consisted of six items measuring state anxiety (nervous, anxious, relaxed, at ease, and

PAGE 21

13 worried) and asked participants to respond according to how they felt at this moment (e.g., 1 = not at all nervous, 9 = very nervous). The second questionnaire was identical to the reflection questionnaire that participants completed earlier with the exception that the prediction items appeared in a different order. Likewise, the minimum number items appeared in a different order. The experimenter explained that, because participants were now receiving feedback and the feedback might influence their thinking, the experimenter would discard the reflection questionnaire that the participants completed earlier and only use in data analysis the responses on the questionnaire participants were now completing. When all participants had completed these final questionnaires, they were debriefed. Results We averaged across all participants who imagined they had engaged in the interactions to create an average standard for each of the dimensions. The average standards ranged from 4.7 (the number of people to rate you in their top three) to 7.8 (the number of people to find you friendly). The average across all dimensions was 5.9. We performed no tests to test for differences between these standards because we were not interested in how standards differ across domains. Our interest was solely in establishing a baseline for the standard Table 1 presents the means for each dimension. Comparing Standards across Time Did participants who engaged in the interactions change their standard of what constituted an acceptable outcome in anticipation of evaluations from their fellow participants? Table 2 presents the mean responses for all five items and the average of the five items. As predicted, averaging across the five items, the outcome that participants reported would be acceptable at Time 2 was significantly lower than the

PAGE 22

14 outcome that participants reported would be acceptable at Time 1. Analysis of the individual items revealed that participants reported a significantly lower standard at Time 2 than at Time 1 for all items except for the item asking how many people they wanted to find them friendly. Comparing Estimates across Time Did participants who engaged in the interactions lower their expectations of how they would be evaluated in the moments prior to receiving feedback? The bottom half of Table 2 displays the mean expectation at Time 1 and Time 2 for each of the five items and for the average across the five items. Consistent with prior research on bracing (Shepperd et al., 1996), averaging across the five items, participants reported a significantly lower expectation at Time 2 than at Time 1. Analysis of each of the five individual items revealed that participants reported a significantly lower expectation at Time 2 than at Time 1 for all items except for the item asking participants to estimate the number of people that would want to spend time with them in the future. Comparing Estimates to Standards Table 3 provides the correlations between the standards and expectations for Time 1 and Time 2 for each dimension. We were not surprised that standards and expectations were correlated. They are both benchmarks by which people evaluate their outcomes. Thus, we would anticipate that the participants who reported high standards would also report high expectations and vice versa. However, the correlation coefficients do not speak to whether the standards and expectations are the same thing. We were more interested in how the average standard across participants compared to the average expectation.

PAGE 23

15 We next compared participants expectations of how they would be evaluated by their peers to their standards of what was an acceptable evaluation. As evident in Table 4, for all items at both Time 1 and Time 2, the evaluation participants expected from their peers fell below the evaluation they viewed as acceptable. In sum, participants systematically report a lower standard than expectation at both Time 1 and Time 2. In addition, for participants, we also subtracted their expectation from their standard. Table 6 presents the distribution of participants who set lower standards than expectations, the same standard as expectation, and higher standards than expectations for each dimension both at Time 1 and Time 2. As is evident in Table 5, the majority of the participants set a higher standard than expectation at Time 1. Although less pronounced at Time 2, there is still a tendency for participants to report a higher standard than expectation. In sum, participants on average set higher standards than expectations at both Time 1 and Time 2. We entertain several possible explanations for the difference between participants standards and their expectations in the discussion. Comparing Changing Standards to Bracing We were not only interested in showing that standards and expectations were different, but also, that they measured different constructs. Table 6 displays the correlations between the change in standards the change in expectations for each dimension. In only one instance were the change in standard and the change in expectation significantly correlated. The absence of a significant correlation reveals that whether or participants changed their standard was unrelated to whether or not they changed their expectations.

PAGE 24

16 Assessing the Accuracy of Participants Standards and Expectations Because participants evaluated each other, we could examine how participants standards compared with the evaluations they received from their peers. Table 7 compares the evaluations participants received from their peers with the standard participants reported would constitute an acceptable outcome for each of the five items. At Time 1, the standard that participants reported that would be an acceptable did not significantly differ from the actual feedback their peers provided for all items except the number of people who ranked the participant in their top three favorite people. Because participants significantly lowered their standard of an acceptable outcome from Time 1 to Time 2, the minimum standard participants reported at Time 2 was significantly lower than the actual feedback they received for all five dimensions. We also examined how participants expectations of what would happen compared with the evaluations they received from their peers. Table 8 compares the actual evaluations that participants received from their peers with participants expectations of what was likely to happen for each of the five items. Consistent with prior research on bracing (Shepperd et al., 1996), averaging across the five items, participants expectations of how they would be evaluated by their peers was significantly lower than the actual evaluation participants received from their peers at both Time 1 and Time 2. Analysis of each of the five individual items revealed that expectations fell below outcomes both at Time 1 and Time 2 for all items except for the item asking participants to estimate the number of people that would rank the participant in their top three favorite people. The analyses just described explore differences in mean responses. Mean responses, of course, can be influenced by extreme responses and do not reveal whether

PAGE 25

17 or how many individual participants were accurate or inaccurate in their expectations of what would happen. Table 9 presents the number of participants who were accurate and inaccurate when expectations were compared with actual outcomes. To provide a strong test of our hypothesis, we were liberal in how we defined accuracy. Specifically, we defined an estimate as accurate if the estimate fell within + or one place from the actual outcome. Thus, for example, if a participant estimated 7 people would find them friendly, we classified the estimate as accurate if 5, 6 or 7 people rated them as friendly (7 +/1). If the expectation fell more than one place below the outcome, we classified the estimate as pessimistic. If the expectation fell more than one place above the outcome, we classified the estimate as optimistic. We repeated this system of classification when we compared participants standards with their outcomes. The distribution of participants in Table 9 indicates support for accuracy in their standards at Time 1. Taking a closer look at the average standard at Time 1 suggests that the pattern of responses is fairly evenly dispersed around accuracy. The pattern also suggests more participants were accurate with their average standard at Time 1 than were optimistic or pessimistic. In sum, there appears to be less systematic favoring of either optimism or pessimism. On the other hand, the frequencies for average expectation at Time 1, average standard at Time 2, and average expectation at Time 2 suggest a tendency for more participants to report pessimism. Also, in many of the specific dimensions, as feedback approached (Time 2), there was an increase in people making estimates that fell under the column indicating a pessimistic view, and a decrease in the number of people who subscribed to an overly optimistic view. Recall that the mean responses among participants showed that the standards at Time 1 were not significantly

PAGE 26

18 different from the actual peer evaluations. Also recall that the standards at Time 2 and the expectations at Time 1 and Time 2 were significantly lower than the actual peer evaluations. In sum, mean responses indicated accuracy for the standards set at Time 1. In addition, the distribution of individual participant's responses also suggest more of a tendency for accuracy at the standard at Time 1 and more of a pessimistic tendency at the standard at Time 2 and the expectations at both Time 1 and Time 2. Comparisons to Participants in Hypothetical Interactions It is possible that participants actually have standards much higher than the feedback they actually received and that the pressure of the experimental situation or the interaction task itself created a feeling of anxiety. The anxiety may have led these participants to report lower standards at Time 1, prompting greater accuracy in the standards reported at Time 1. Although, we did not assess anxiety at Time 1, we view this explanation for participants' accuracy at Time 1 unlikely. Participants estimates at Time 1 were identical to the estimates of participants in the hypothetical condition. Table 10 compares the expectations for the participants who engaged in the interaction to the standards provided by the participants in the hypothetical survey. For all five items, the expectations participants set at Time 1 and Time 2 were significantly lower than the independent sample estimates of an acceptable outcome. Table 11 compares the standards at Time 1 and Time 2 from participant who engaged in the interaction to independent sample from the hypothetical survey. As is evident, we found no significant difference between the standards set at Time 1 by participants who engaged in the interactions and the standards of participants in the independent sample.

PAGE 27

19 Discussion As predicted, participants estimated a lower evaluation at Time 2, when feedback was imminent, than at Time 1, when feedback was unexpected. More importantly, participants reported a lower standard at Time 2 than at Time 1. Also, participants average standard was consistently higher than the average expectation. Although the standards were correlated with the expectations, the change in standard was rarely correlated with the change in expectations. In terms of when the benchmarks reported by the participants were consistent with their actual peer evaluations, participants' standards at Time 1 were most consistent with their peers evaluations, whereas participants' standards at Time 2 and expectations at both Time 1 and Time 2 were inconsistent with their actual peer evaluations. What participants first estimated to be an acceptable outcome foreshadowed the peer evaluations they actually received. Our participants first estimate of the outcome they desired was actually consistent with how they were evaluated by their peers. Also, the standard at Time 1 was not significantly different than the standard provided the hypothetical interaction participants. It may seem that because participants lowered both their standard and their expectations, standards are not distinct from expectations. As we previously mentioned, often what people want is different from what people think is likely to happen. We are not the first to regard standards and expectations as distinct. Several other researchers have also argued that the two constructs differ (Battle, 1965; Battle, 1966; Locke, 1967). Our findings, however, provide some evidence that the two constructs are distinct. First, at Time 1 and Time 2, recall that, on average, participants set significantly higher standards than expectations. Simply put, each time participants were given the

PAGE 28

20 opportunity, they reported standards that were significantly different from their expectations and this direction favored a higher standard than expectation. In addition to group averages, the pattern of frequencies in Table 9 suggests that the majority of participants set higher standards than expectations. Second, and most important, we believe a distinction between standards and expectations results from the fact that participants change in standards was not correlated with their change in expectation. The lack of correlation indicates the extent to which the participants changed their standards differs quantitatively from how they changed their expectations. A person who lowers their standards a lot may not lower his or her expectation as much, or vice versa. Although more research assessing the functional differences between standards and expectations needs to be done, we, with caution, take the lack of correlation between the change in standards and the change expectation as some preliminary support that standards and expectations are distinct constructs. As mentioned earlier, participants set significantly higher standards than expectations at both Time 1 and Time 2. This is an intriguing finding because it suggests that participants desired an outcome that they believed they would not obtain. There are several reasons that could account for why participants standards exceeded their expectations. We should note at this time, that all participants completed expectation items shortly after completing the standard items. First, it is possible that as time elapsed between completing the standard items and the expectation items, participants became increasingly anxious. The anxiety, in turn, led participants to display lower estimates for the expectation items than the standard items. We view this explanation unlikely because the amount of time that elapsed between completing the two sets of items was small, less

PAGE 29

21 than a minute. In addition, the reported standards were higher than the expectations at both Time 1 and Time 2. The discrepancy exists at Time 1 even though there was little to no reason to feel anxious, as they were not expecting feedback. A second possibility is that participants were displaying modesty in their expectations so as not to appear presumptuous or self-aggrandizing. After first reporting what they believed to be an acceptable outcome, participants felt moved to then report lower expectations to appear modest or even self-deprecating. We also view this explanation unlikely because participants were not reporting their estimates out loud, but rather they were privately reporting them on paper and therefore had less pressure to keep up a modest appearance because they had no public audience for which they could promote a modest appearance. Also, participants were instructed to turn in the forms in the envelopes they received so there is little reason to think that the participants would believe the experimenters would have access to their responses at that time. Therefore, participants were thus under little pressure to appear modest. In addition, the time between estimates, less than a minute, would have allowed little time to consider how their reported standards would be perceived. A third possibility is that people may want to be liked by everyone and so they report unrealistically high standards when asked to indicate their minimum acceptable number. In contrast, when asked to indicate their expectations, participants may have felt constrained by what they felt was likely to happen, which prompted them to report expectations below their standards. We find two problems with this explanation. First, we did not get a ceiling effect in participants response to the items asking about their standards (what they wanted to happen). Second, participants standards at Time 1

PAGE 30

22 generally did not differ from the evaluations they received from their peers. Both findings suggest that participants were not idealistic in their standards. The final, and most likely possibility is that participants had high standards for what they feel is acceptable and felt at both Time 1 and Time 2 that they that were unlikely to receive feedback that supported this standard. This explanation relies on the distinction between what people want and what they expect they are likely to receive. This explanation deserves further investigation. One other finding that emerged in participants standards and expectations deserves attention. Specifically, the mean response to the items asking participants how many people they wanted to rank them in their top 3 suggests a degree of unrealistic optimism. Out of ten people, statistical odds would say people should expect on average three people to rank them in their top 3. However, the participants in our study reported an average standard of 4.3 at Time 1. This statistical and logical failure of the participants may account for why the only standard at Time 1 that was significantly different from what actually happened was the standard for the number of people to rank you in their top 3 (see Table 8). Also in Table 10, many more people were optimistic (across all times and both standards and expectations) in response to this item than in response to other items. There are several reasons for this anomaly in the data. First, participants may have a poor understanding of statistics and may have failed to see that their number exceeded what, on average, was statistically likely. Second, it is likely that the higher responses to other items (the number of people to find you friendly, interesting) anchored the participants and they adjusted poorly from that higher point. For all of the other

PAGE 31

23 dimensions, participants reported minimum standards that were higher so perhaps those relatively high standards influenced the standard for the number of people to rank you in their top three favorite people. A third possibility is based on the idea that the top three dimension is the one dimension that actually provides a clear answer to what is an objectively good or acceptable outcome. Statistically, to exceed the statistically average of 3 is good, to have less than three people rank you in their top 3 is bad. A common finding in western cultures is the better-than-average effect (Alicke, Klotz, Breitenbecher, Yurak, & Vredenburg 1995, Dunning, Meyerowitz, & Holzberg, 1989). Research finds that many people in western cultures (often well over the majority) rate themselves as better than average on a variety of dimensions. It seems people are reluctant to think that they may be below average and even being average is viewed as negative (Alicke et al., 1995, Dunning et al., 1989). Perhaps the better-than-average effect extends to setting standards for an acceptable outcome, resulting in our participants, on average, being unwilling to claim that they have an average standard in terms of being well liked by others. Perhaps, not only do the majority of people believe they are better than average drivers, they also (by self-report), set standards that indicate they are better than average in terms of popularity.

PAGE 32

24 Table 1: Study 1: Average estimate of what determines a good outcome for each of the dimensions Item Average number of people out of ten Find you friendly 7.8 Find you interesting 6.3 Rank you in their top three 4.7 Learn more about you 5.5 Spend time with you in the future 5.0 Mean Across All Items 5.9

PAGE 33

25 Table 2: Study 1: Standards and expectations from Time 1 to Time 2. Standard Number of people you would like to... Time 1 M Time 2 M t p Find you friendly 7.4 7.1 1.18 ns Find you interesting 6.3 5.2 3.45 .002 Rank you in their top three 4.3 3.7 3.27 .003 Learn more about you 5.5 5.1 2.11 .05 Spend time with you in the future 5.5 4.6 4.33 .0001 Mean Across All Items 5.8 5.1 4.29 .0001 Expectation Number of people you expect to... Time 1 M Time 2 M t p Find you friendly 6.8 6.2 2.40 .03 Find you interesting 5.0 4.5 2.66 .02 Rank you in their top three 3.5 3.0 2.86 .007 Learn more about you 4.8 4.5 2.16 .04 Spend time with you in the future 4.3 4.1 1.24 ns Mean Across All Items 4.9 4.5 3.75 .0006

PAGE 34

26 Table 3: Study 1: Correlations between standards and expectations at Time 1 and Time 2. Item Time 1 Time 2 Find You Friendly .63 .65 Find You Interesting .42 .46 Rate You in Their Top 3 .58 .61 Learn More About You .57 .63 Spend Time in Future .68 .62 Note: All correlations significant at p < .01.

PAGE 35

27 Table 4: Study 1: Standards versus expectations at Time 1 and at Time 2. Time 1 Number of people you would like to... Standard M Expectation M t p Find you friendly 7.4 6.8 1.79 .09 Find you interesting 6.3 5.0 3.79 .002 Rank you in their top three 4.3 3.5 2.30 .03 Learn more about you 5.5 4.8 2.25 .04 Spend time with you in the future 5.5 4.3 3.79 .001 Mean Across All Items 5.8 4.9 3.26 .003 Time 2 Find you friendly 7.1 6.2 2.23 .03 Find you interesting 5.2 4.5 1.93 .07 Rank you in their top three 3.7 3.0 2.30 .03 Learn more about you 5.1 4.5 2.21 .04 Spend time with you in the future 4.6 4.1 1.69 .10 Mean Across All Items 5.1 4.5 2.34 .03

PAGE 36

28 Table 5: Study 1: Number of participants who set standards higher, equal to or lower than expectations. Standard < Expectation Standard = Expectation Standard > Expectation Time 1 Find you friendly 8 11 20 Find you interesting 7 6 26 Rank you in their top three 6 10 23 Learn more about you 7 11 21 Spend time with you in future 6 5 28 Time 2 Find you friendly 5 15 19 Find you interesting 8 13 18 Rank you in their top three 5 19 15 Learn more about you 6 15 18 Spend time with you in future 8 15 16 All numbers in the table refer to the number of participants out of 39 who fall in each category.

PAGE 37

29 Table 6: Study 1: Correlations between the change in standards and the change in expectations. Dimension Correlation Find You Friendly *.44 Find You Interesting -.06 Rate You in Their Top 3 .21 Learn More About You .24 Spend Time in Future .24 p < .01

PAGE 38

30 Table 7: Study 1: Standard and received outcomes at Time 1 & Time 2. Time 1 Number of people you would like to... Standard M Outcome M t p Find you friendly 7.4 8.1 1.71 .10 Find you interesting 6.3 6.9 1.23 ns Rank you in their top three 4.3 2.8 2.86 .001 Learn more about you 5.5 6.0 1.21 ns Spend time with you in the future 5.5 5.2 <1 ns Mean Across All Items 5.8 5.8 <1 ns Time 2 Find you friendly 7.1 8.1 2.23 .04 Find you interesting 5.2 6.9 4.52 .001 Rank you in their top three 3.7 2.8 1.84 .08 Learn more about you 5.1 6.0 2.08 .05 Spend time with you in the future 4.6 5.2 1.43 ns Mean Across All Items 5.1 5.8 1.97 .06

PAGE 39

31 Table 8: Study 1: Mean expected and received outcomes at Time 1 & Time 2. Time 1 Number of people you expect to... Expected Outcome M Received Outcome M t p Find you friendly 6.8 8.1 3.11 .004 Find you interesting 5.0 6.9 4.35 .0001 Rank you in their top three 3.5 2.8 1.62 ns Learn more about you 4.8 6.0 2.70 .02 Spend time with you in the future 4.3 5.2 1.98 .06 Mean Across All Items 4.9 5.8 2.54 .02 Time 2 Find you friendly 6.2 8.1 4.1 .0002 Find you interesting 4.5 6.9 6.0 .0001 Rank you in their top three 3.0 2.8 <1 ns Learn more about you 4.8 6.0 3.39 .002 Spend time with you in the future 4.1 5.2 2.36 .03 Mean Across All Items 4.5 5.8 3.84 .0004

PAGE 40

32 Table 9: Study 1: Number of participants who were pessimistic, accurate and optimistic. Pessimistic Accurate Optimistic Average Standard at Time 1 11 18 10 Average Standard at Time 2 14 18 7 Average Expectation at Time 1 13 20 6 Average Expectation at Time 2 16 23 1 Find you friendly Standard at Time 1 14 20 5 Standard at Time 2 16 15 8 Expectation at Time 1 16 19 4 Expectation at Time 2 20 17 2 Find you interesting Standard at Time 1 17 11 11 Standard at Time 2 20 27 3 Expectation at Time 1 24 11 4 Expectation at Time 2 23 15 1 Put you in their top 3 Standard at Time 1 5 14 20 Standard at Time 2 5 21 13 Expectation at Time 1 6 20 13 Expectation at Time 2 7 19 13 Want to learn more about you Standard at Time 1 12 17 10 Standard at Time 2 15 17 7 Expectation at Time 1 20 11 8 Expectation at Time 2 20 13 6 Want to spend time with you Standard at Time 1 11 12 16 Standard at Time 2 14 16 9 Expectation at Time 1 13 18 8 Expectation at Time 2 15 17 7 All numbers in the table refer to the number of participants out of 39 who fall in each category. The pessimistic label indicates lower expectation or standard than what actually happened. The accurate label indicates near the same expectation or standard as what actually happened. The optimistic label indicates higher expectation or standard than what actually happened. The rows indicating average count the number of people whose average standard or expectation (across all five dimensions) is lower, equal to, or higher than the average outcome they received (across all five dimensions).

PAGE 41

33 Table 10: Study 1: Mean expected outcome at Time 1 and 2 and hypothetical standard (minimum acceptable outcome for survey). Expected Outcome M Hypothetical Standard M T p Time 1 Find you friendly 6.8 7.8 3.10 .004 Find you interesting 5.0 6.3 4.02 .001 Rank you in their top three 3.5 4.7 3.47 .002 Learn more about you 4.8 5.6 2.39 .03 Spend time with you in the future 4.3 5.0 2.06 .007 Time 2 Find you friendly 6.2 7.8 4.26 .0001 Find you interesting 4.5 6.3 6.21 .0001 Rank you in their top three 3.0 4.7 5.69 .0001 Learn more about you 4.5 5.6 3.25 .003 Spend time with you in future 4.1 5.0 2.86 .007

PAGE 42

34 Table 11: Study 1: Mean desired outcome (hypothetical survey) and mean standard (minimum acceptable outcome) at Time 1 & Time 2. Acceptable Outcome M Hypothetical Standard M T p Time 1 Find you friendly 7.4 7.8 1.09 ns Find you interesting 6.3 6.3 <1 ns Rank you in their top three 4.3 4.7 <1 ns Learn more about you 5.5 5.5 <1 ns Spend time with you in the future 5.5 5.0 1.13 ns Mean 5.8 5.7 .29 ns Time 2 Find you friendly 7.1 7.8 1.75 .09 Find you interesting 5.2 6.3 3.40 .002 Rank you in their top three 3.7 4.7 2.76 .009 Learn more about you 5.1 5.6 1.15 ns Spend time with you in future 4.6 5.0 1.22 ns Mean 5.1 5.9 1.74 .09

PAGE 43

CHAPTER 3 STUDY 2 Overview The primary purpose of Study 2 was to replicate the findings of Study 1, but also to show that the change in standard occurs in response to impending feedback. Study 2 thus included a no feedback (control condition) in which participants believed they would not receive evaluations from their peers. A secondary purpose of Study 2 was to examine whether the discrepancy between standards and expectations was due to the order of the questionnaires. In Study 1 participants always completed the standard items before the expectation items. In Study 2 we counterbalanced the order of these items. Method Participants. Introductory psychology students (N = 53) participated in partial fulfillment of a research participation requirement and were run in groups of 11 people with the exception of one session in which there were only 9 participants. Procedures The procedure was identical to Study 1 with the following exceptions. First, after participants interacted and evaluated each other, the experimenter randomly separated participants into two rooms under the pretext of providing them more room to complete the next set of questionnaires. All participants then provided their standards of what would be an acceptable outcome and their outcome expectations on each of the five measures (Time 1). 35

PAGE 44

36 Second, the Time 1 evaluation materials provided additional information about the terms standards and expectations. Specifically, participants read: a standard is what you consider the minimally acceptable outcome. In other words, a standard is the lowest outcome that you would find acceptable in a given situation. Participants also read: an expectation is the outcome that you predict would occur. In other words, an expectation is the outcome that you think is likely to happen in a given situation. Third, we counterbalanced the items assessing standards with the items assessing expectations. Fourth, for each of the five evaluative domains, reported what they viewed as the ideal outcome. Fifth, after we collected the responses and participants had completed several filler questionnaires, participants assigned to the feedback room learned, as in Study 1, that they would receive the evaluations of the peers. Participants in the no feedback room kept all questionnaires they completed until the end of the experiment and continued to believe that they would not receive the evaluations of their peers. All participants then completed the same expectation, standard and ideal items they completed at Time 1 with the items assessing participants expectation once again counterbalanced with the items assessing participants standards. In addition, as in Study 1, participants completed items measuring state anxiety. When all participants had completed these final questionnaires, they were reunited, debriefed to assess their perceptions of the experiment, and thanked for their participation.

PAGE 45

37 Hypothesis We tested the following hypotheses: 1. Standards vs. Expectations: Replicating the pattern in Study 1, participants standards would be significantly higher than participants expectations. 2. Expectations: Participants in the feedback condition would report a significantly lower expectation at Time 2 than at Time 1, whereas participants in the no feedback condition would report no difference in expectations from Time 1 to Time 2. 3. Standards: Participants in the feedback condition would report a lower standard at Time 2 than at Time 1, whereas participants in the no feedback condition would report no difference in standards from Time 1 to Time 2. 4. Change in Standards vs. Bracing: Participants change in standards would be uncorrelated with their change in expectations. In addition, the majority of participants would report a change in standards that was quantitatively different from their change in expectation. 5. Anxiety Reports: Anxiety reports would be higher among participants in the feedback condition than among participants in the no feedback condition. Also, anxiety reports would correlate positively with bracing. In addition, anxiety reports would correlate positively with the magnitude of change in standard. 6. Comparisons to actual outcomes: Replicating the findings of Study 1, the evaluations participants receive from their peers, would be significantly higher than their expectations at Time 1 and Time 2 and significantly higher than their standards at Time 2. 7. Ideals: Participants ideal outcomes would be significantly higher than both the standards and expectations that they reported. Results Differences between standards and expectations Recall that in Study 1, participants reported lower expectations than standards at both Times 1 and 2. In Study 2, the items assessing standards and expectations were counterbalanced at both Time 1 and Time 2 to eliminate order of the questionnaires as an explanation for the differences in responses. Table 12 displays the difference between standards and expectations when the items are counterbalanced. Contrary to our

PAGE 46

38 hypothesis and the results of Study 1, in no case were participants standards significantly higher than their expectations. In fact, the generally finding was that expectations were significantly higher than the standards. We revisit the issue of the differences between standards and expectations in the discussion. Comparing Expectations across Time Did participants in the feedback condition lower their expectations of how they would be evaluated in the moments prior to receiving feedback (Hypothesis 2)? The top half of Table 13 displays the mean expectation at Time 1 and Time 2 for each of the five items and for the average across the five items for the feedback condition. Consistent with prior research on bracing (Shepperd et al., 1996), participants anticipating evaluative feedback from their peers rated reported a significantly lower average expectation at Time 2 than at Time 1. Analysis of each of the five individual items revealed that participants expecting feedback reported a significantly lower expectation at Time 2 than at Time 1 for all items except for the item asking participants to estimate the number of people that would find them interesting. The bottom half of Table 13 displays the mean expectations at Time 1 and Time 2 for each of the five items and the average across the five items for the no feedback condition. We hypothesized that the expectations of participants anticipating no feedback would not differ from Time 1 to Time 2. Consistent with this prediction, analysis of the average expectation across all five dimensions and all five of the dimensions individually yielded no significant differences between Time 1 and Time 2.

PAGE 47

39 Comparing Standards across Time Did participants who were anticipating feedback change their standard of what constituted an acceptable outcome from Time 1 to Time 2 (Hypothesis 3)? The top half of Table 14 presents the mean responses for all five items and the average of the five items for the feedback condition. As predicted, averaging across the five items, participants expecting to receive evaluations from their peers reported a lower standard at Time 2 than at Time 1. Analysis of the individual items revealed that participants reported a significantly lower standard at Time 2 than at Time 1 for all items except for the item asking how many people they wanted to find them friendly and how many people they wanted to want to learn more about them. The bottom half of Table 14 displays the mean standards at Time 1 and Time 2 for each of the five items and the average across the five items for the no feedback condition. We hypothesized that standards of participants who anticipated no feedback would not differ from Time 1 to Time 2. Consistent with our hypothesis, we found no difference across time in the standards of no feedback participants except for the item asking participants to estimate the number of people that would find them interesting. The results of Study 2 provide further support that as feedback draws near, people will alter their estimates for the future. As predicted, only participants in the feedback condition changed their estimates and their standards from Time 1 to Time 2. An unexpected pattern of means did appear in the data. The difference between the conditions appears to occur at Time 1. We explore possible reasons for this finding in the discussion.

PAGE 48

40 Comparing Changing Standards to Bracing Are standards and expectations the same thing, or are they different? As in Study 1, the correlations between standards and expectations at Time 1 and at Time 2 were rather high (see Table 15). As noted earlier, the correlations provide a somewhat misleading picture of the correspondence between standards and expectations. Recall that participants expectations differed from their standards (see Table 12). We next compared the magnitude of change in standard from Time 1 to Time 2 with the magnitude of change in expectation from Time 1 to Time 2. As evident in Table 16, in no instance was the change in expectation significantly correlated with the change in standards, suggesting that participants were not supplying the same response to the items assessing standards and the items assessing expectations. Changes in Standards and Changes in Expectations were uncorrelated in both Studies 1 and 2. However, it can be argued that neither study had adequate power to test the correlation, and that a larger sample size is necessary to uncover significant effects. It is noteworthy, however, that the correlation between change in standard and change in expectation is negative in three of five instances in Study 2. Nevertheless, to test the possibility that the absence of a correlation between change in standard and change in expectation was due to an inadequate sample size, we combined the data from Studies 1 and 2 and computed the correlations again. As evident in Table 17, combining the data from Studies 1 and 2 had little effect. In only one instance (number of people who will find you friendly) was the correlation significant and this correlation was notably small. In sum, the extent to which participants changed their standards was uncorrelated with the extent to which they changed their expectations, suggesting that the two constructs are not identical.

PAGE 49

41 An additional way to examine whether changes in standards and changes in expectations represent the same construct is to compare how many participants displayed the same amount of change in standards as change in expectations. Recall the dependent variable was the number of people the participants expect or desire to find them friendly or interesting, etc. For example, if a person lowered the standard for a good outcome from 8 people to 6 people, would the expectation also drop 2 units? Table 18 presents the number of participants out of 53 that changed their standard the same quantitative amount as they changed their expectation. As is evident, for only a minority of the participants was the change in standards identical to the change in expectations. Anxiety Reports We hypothesized that participants will report greater anxiety in the feedback condition than the no feedback condition (Hypothesis 5). Although the means were in the predicted direction, statistical analyses revealed no difference in anxiety for participants in the feedback condition (M = 3.23) and participants in the no feedback condition (M = 2.56), t (51) = 1.56, p < .13. Not surprisingly given that participants in the two conditions did not differ in their anxiety reports, analyses revealed that anxiety and the amount of change in standards and the amount of change in expectations were largely uncorrelated (see Table 19). Assessing the Accuracy of Participants Standards and Expectations As in Study 1, we compared participants standards and expectations with the actual evaluations they received from their peers. We predicted that the evaluations participants receive from their peers would be significantly higher than their expectations at Time 1 and Time 2 and significantly higher than their standards at Time 2 (see Hypothesis 6). Tables 20 and 21 compare the evaluation participants received from their

PAGE 50

42 peers with the standard participants reported as an acceptable outcome and the expectations they reported. As we hypothesized (and consistent with Study 1), participants estimates at Time 1, and their estimates and standards at Time 2, were significantly lower than the rating made by peers for all items except the item asking participants whether they would be rated in the top three by their peers. Contrary to our hypothesis, however, participants standards at Time 1 were also significantly lower than their peer evaluations. Ideals As noted earlier, it is possible that participants were not supplying a standard in Study 1, but rather were supplying an ideal outcome; the outcome they would like to occur in the best possible world. We argued that this seemed unlikely because we did not get a ceiling effect in participants standards in Study 1. Nevertheless, to test whether the standard was distinct from the ideal outcome, we asked participants to indicate what they regarded as the ideal outcome. We predicted that they ideal outcome what be significantly higher than the standard participants reported (Hypothesis 7). Table 22 presents the standards and ideals at Times 1 and 2. As predicted across all five dimensions at both Time 1 and Time 2, participants ideals exceeded their standards. Table 23 presents the expectations and ideals at Times 1 and 2. As predicted across all five dimensions at both Time 1 and Time 2, participants ideals exceeded their expectations. Ideals were always assessed after standards and expectations to ensure that responding to the ideal items did not influence the ratings for standards or expectations.

PAGE 51

43 Table 12: Study 2: Standards versus expectations at Time 1 and at Time 2 when standards and expectations are counterbalanced. Time 1 Number of people you would like to... Standard M Expectation M t p Find you friendly 6.9 7.7 2.65 .01 Find you interesting 5.4 5.6 .70 ns Rank you in their top three 3.1 3.6 1.79 .08 Learn more about you 4.1 4.9 3.65 .001 Spend time with you in the future 4.3 4.6 1.27 ns Mean Across All Items 4.7 5.3 2.48 .02 Time 2 Find you friendly 6.8 7.4 2.43 .02 Find you interesting 4.2 5.3 3.69 .001 Rank you in their top three 2.7 3.5 3.40 .002 Learn more about you 3.9 4.5 2.42 .02 Spend time with you in the future 3.4 4.2 4.42 .001 Mean Across All Items 4.2 5.0 3.98 .0002

PAGE 52

44 Table 13: Study 2: Expectations at Time 1 and Time 2 by condition. Feedback Condition Number of people you expect to... Time 1 M Time 2 M t P Find you friendly 8.0 7.6 1.73 .096 Find you interesting 5.6 5.3 .71 ns Rank you in their top three 3.8 3.3 2.05 .05 Learn more about you 4.9 4.2 2.96 .007 Spend time with you in the future 4.7 4.0 1.85 .08 Mean Across All Items 5.4 4.9 2.62 .02 No Feedback Condition Number of people you expect to... Time 1 M Time 2 M t p Find you friendly 7.4 7.4 0 ns Find you interesting 5.6 5.3 .86 ns Rank you in their top three 3.3 3.6 1.37 ns Learn more about you 4.9 4.8 .63 ns Spend time with you in the future 4.7 4.4 1.14 ns Mean Across All Items 5.2 5.1 .49 ns

PAGE 53

45 Table 14: Study 2: Standards at Time 1 to Time 2 by condition. Feedback Condition Number of people you would like to... Time 1 M Time 2 M t p Find you friendly 7.3 6.8 1.19 ns Find you interesting 5.8 4.4 3.19 .004 Rank you in their top three 3.4 2.9 1.79 .09 Learn more about you 4.2 3.8 1.22 ns Spend time with you in the future 4.7 3.4 3.68 .002 Mean Across All Items 5.1 4.3 3.34 .003 No Feedback Condition Number of people you would like to... Time 1 M Time 2 M t p Find you friendly 6.4 6.7 .68 ns Find you interesting 4.8 4.0 2.88 .008 Rank you in their top three 2.9 2.5 1.09 ns Learn more about you 3.9 4.0 .35 ns Spend time with you in the future 3.9 3.4 1.55 ns Mean Across All Items 4.4 4.1 .97 ns

PAGE 54

46 Table 15: Study 2: Correlations between standards and expectations at Time 1 and at Time 2. Time 1 Time 2 Find You Friendly .51 .46 Find You Interesting .43 .51 Rate You in Their Top 3 .59 .64 Learn More About You .70 .66 Spend Time in Future .57 .62 Note: All correlations significant at p < .01.

PAGE 55

47 Table 16: Study 2: Correlations between the change in standards and the change in expectations. Dimension Correlation Find You Friendly .05 Find You Interesting -.10 Rate You in Their Top 3 .18 Learn More About You -.01 Spend Time in Future .11 p < .05

PAGE 56

48 Table 17: Correlations between the change in standards and the change in expectation across both Study 1 and Study 2 (Total N=81). Dimension Correlation Find You Friendly *.23 Find You Interesting -.06 Rate You in Their Top 3 .03 Learn More About You .10 Spend Time in Future .20 p < .05

PAGE 57

49 Table 18: Study 2: Number of participants changed their standards the same amount that they change their expectations. Changed Expectations and Standards the same amount N % Find You Friendly 9 17% Find You Interesting 10 19% Rate You in Their Top 3 19 36% Learn More About You 16 30% Spend Time in Future 17 32% The numbers refer to the participants out of 53.

PAGE 58

50 Table 19: Study 2: Correlations between anxiety and change in standards & anxiety and change in expectations. Change in Standard Change in Expectation Feedback Condition (M = 3.3) Find You Friendly .03 -.06 Find You Interesting -.11 -.27 Rate You in Their Top 3 .08 .20 Learn More About You -.04 .29 Spend Time in Future .31 .32 No Feedback Condition (M =2.5) Find You Friendly **.49 .30 Find You Interesting .12 .06 Rate You in Their Top 3 .18 .11 Learn More About You .48 .10 Spend Time in Future .19 -.08 Overall (M =2.8) Find You Friendly *.28 .14 Find You Interesting .00 -.13 Rate You in Their Top 3 .14 .22 Learn More About You .08 .24 Spend Time in Future *.29 .17 The means in parenthesis correspond to the average reported anxiety for the group. p < .05 p < .01 **

PAGE 59

51 Table 20: Study 2: Standard and received outcomes at Time 1 & Time 2. Time 1 Number of people you would like to... Standard M Received Outcome M t p Find you friendly 6.9 8.9 5.68 .0001 Find you interesting 5.4 7.4 5.84 .0001 Rank you in their top three 3.1 3.0 .41 ns Learn more about you 4.1 6.4 6.88 .0001 Spend time with you in the future 4.3 6.1 4.54 .0001 Mean Across All Items 4.7 6.4 5.61 .0001 Time 2 Find you friendly 6.8 8.9 6.13 .0001 Find you interesting 4.2 7.4 8.72 .0001 Rank you in their top three 2.7 3.0 .75 ns Learn more about you 3.9 6.4 6.75 .0001 Spend time with you in the future 3.4 6.1 7.04 .0001 Mean Across All Items 4.2 6.4 7.82 .0001

PAGE 60

52 Table 21: Study 2: Mean expected and received outcomes at Time 1 & Time 2. Time 1 Number of people you expect to... Expected Outcome M Received Outcome M t p Find you friendly 7.7 8.9 4.32 .0001 Find you interesting 5.6 7.4 5.24 .0001 Rank you in their top three 3.6 3.0 1.73 .09 Learn more about you 4.9 6.4 4.35 .0001 Spend time with you in the future 4.6 6.1 3.93 .0003 Mean Across All Items 5.3 6.4 3.98 .0002 Time 2 Find you friendly 7.4 8.9 5.92 .0001 Find you interesting 5.3 7.4 6.61 .0001 Rank you in their top three 3.5 3.0 1.42 ns Learn more about you 4.5 6.4 5.39 .0001 Spend time with you in the future 4.2 6.1 4.94 .0001 Mean Across All Items 5.0 6.4 5.48 .0001

PAGE 61

53 Table 22: Study 2: Standard and ideal outcomes at Time 1 & Time 2. Time 1 Number of people you would like to... Standard Outcome M Ideal Outcome M t p Find you friendly 6.9 8.8 5.46 <.0001 Find you interesting 5.4 8.1 8.14 <.0001 Rank you in their top three 3.1 6.4 7.68 <.0001 Learn more about you 4.1 7.2 8.37 <.0001 Spend time with you in the future 4.3 6.8 7.03 <.0001 Mean Across All Items 4.7 7.5 8.66 <.0001 Time 2 Find you friendly 6.8 8.6 5.92 <.0001 Find you interesting 4.2 7.1 7.46 <.0001 Rank you in their top three 2.7 5.5 6.96 <.0001 Learn more about you 3.9 6.4 5.94 <.0001 Spend time with you in the future 3.4 6.1 7.27 <.0001 Mean Across All Items 4.2 6.7 7.81 <.0001

PAGE 62

54 Table 23: Study 2: Mean expected and ideal outcomes at Time 1 & Time 2. Time 1 Number of people you expect to... Expected Outcome M Ideal Outcome M t p Find you friendly 7.7 8.8 3.74 .0005 Find you interesting 5.6 8.1 7.62 <.0001 Rank you in their top three 3.6 6.4 8.37 <.0001 Learn more about you 4.9 7.2 7.12 <.0001 Spend time with you in the future 4.6 6.8 6.62 <.0001 Mean Across All Items 5.3 7.5 8.65 <.0001 Time 2 Find you friendly 7.4 8.6 5.17 <.0001 Find you interesting 5.3 7.1 4.91 <.0001 Rank you in their top three 3.5 5.5 4.85 <.0001 Learn more about you 4.5 6.4 4.94 <.0001 Spend time with you in the future 4.2 6.1 5.35 <.0001 Mean Across All Items 5.0 6.7 5.59 <.0001

PAGE 63

CHAPTER 4 GENERAL DISCUSSION Often events do not come complete with objective indicators of what is and is not an acceptable outcome. With a lack of objective anchors, people must personally define their standard of what is acceptable. We proposed that people would adjust their standards not just in response to new information, but also in response to the anticipation of new information. We tested this latter possibility by leading participants to believe they would soon learn how their peers evaluated them and them asking them to report their standards across a variety of social dimensions. The results from two studies revealed that people reported lower standards of what constituted acceptable social outcomes when they expected feedback and did not adjust standards when they did not expect feedback. The results of the two studies also reveal several other important findings. First, people lower their standards of what defines an acceptable outcome when they believe they will soon receive outcome feedback. Yet, changing standards was not a chance fluctuation; participants expecting no feedback did not change their standards. We propose that lowering the definition of what constitutes an acceptable outcome represents a means of avoiding unfavorable feedback. Second, standards are not synonymous with expectations. In both studies participants reported standards that were significantly different from their expectations. Recall in Study 1, the participants completed items assessing standards prior to items assessing expectations. In Study 1 participants reported standards that were significantly 55

PAGE 64

56 higher than their expectations. In Study 2, when the standard and expectation items were counterbalanced, the opposite pattern emerged, with participants reporting expectations higher than standards. The contradictory pattern of differences between standards and expectations deserves future exploration. A resolution to this may arise from treating the estimates of standards and expectations as a between-subjects variable. Third, the results from both studies demonstrated that changing standards is distinct from bracing for bad news. In Study 1, for four of the five dimensions, the change in standard was not correlated with the change in expectation. This result was replicated in Study 2, which revealed that changes in standards and changes in expectations were uncorrelated for all five dimensions. To address the possibility that low statistical power was responsible for the nonsignificant correlations, we re-examined the correlations after combined the data from Studies 1 and 2. The combined analyses again revealed the changes in standards were uncorrelated with changes in expectations. Finally, we tabulated the amount of change for each participant on each dimension over time. Specifically, for each dimension, we tabulated the number of units of change in standards from Time 1 to Time 2 and the number of units of change in expectations from Time 1 to Time 2. Only for a fraction of participants did the amount of change in standards equal the amount of change in expectations. In sum, it appears that changing standards is distinct from bracing for bad news. Recent research depicts bracing as an attempt to avoid disappointment where disappointment is specific negative emotional response to an outcome (Carroll, Dockery & Shepperd, 2003). We argue that people change their standards for a different reason. In the introduction we proposed a variety of reasons that people may proactively change

PAGE 65

57 their standards. The present results reveal that the extent to which people brace is not predictive of how they change their standards. We believe that, compared with changes in expectations, people may change their standards for different reasons and with different consequences. Although, in a broad sense, bracing and changing standards may serve a self-enhancement function, a closer look may reveal differences in how these processes alter mood or perceptions. Bracing alleviates disappointment about a future outcome and as such, is an alteration of the likelihood of the occurrence of a negative event. Changing standards alleviates the negativity of a future outcome and as such, is an alteration of the perception of the outcome. We believe that the consequences of changing standards and bracing are different. Bracing speaks to a persons evaluation of the subjective likelihood of a certain outcome (e.g., testing positive for a sexually transmitted disease or receiving a failing grade) and it implies acceptance by the person that the undesired outcome is a possibility (i.e., they may be a person who contracts a sexually transmitted disease or fails a course in college). On the other hand, changing standards focuses on the evaluation of the outcome. To change the evaluation of what determines an acceptable physical screening or grade in a college course does not carry implications for the person or necessitate a change in the persons self-concept. A task for future research is to explore the extent to which bracing and changes in standards may have different purpose, result, and consequence for the individual. Fifth, in Study 2, the change in standards was often uncorrelated with reported anxiety. Over both conditions together, there was a significant correlation between change in standards and reported anxiety for two of the five dimensions. We expected

PAGE 66

58 that participants self-reported anxiety would positively and strongly correlate with the change in their standards across all dimensions. Perhaps our measure of anxiety did not fully tap into the feelings and emotions associated with changing standards. That is, some other emotional or psychological process is responsible for the decline in standards among participants anticipating imminent feedback. A second possibility is that the presence of the other participants in the feedback condition caused the participants to desire to project an image of nonchalance at the prospect of receiving feedback. Thus, because participants wanted to appear unruffled by the prospect of receiving feedback, they reported lower anxiety. A closer examination of Table 14 reveals an unexpected pattern of results. Although we found that participants significantly lowered their standards in the feedback condition but not in the no-feedback condition, participants appeared to differ in their standards at Time 1 rather than Time 2. Further exploring of the data revealed that the order of the expectations and standards items could not account for this finding. The Time difference is troubling and may have occurred for one of two reasons. First, participants were divided into the different rooms shortly before completing the Time 1 items. Thus, the design of Study 2 created a slight difference in when and where participants completed Time 1. The moving of the no feedback participants to the other room prior to Time 1 may have affected the ratings at Time 1. Replications that assess Time 1 data prior to separation of participants can eliminate this effect. A second possible explanation is, despite our attempts to make the conditions equal, there was an error in random assignment. Future replications using this paradigm with increased subject number and power may eliminate this explanation.

PAGE 67

59 There are still unanswered questions about changing standards. Are there individual differences in the extent to which people will change their standards? The importance of the dimension for the person may affect the extent to which they change their standards. Also, the necessity of receiving a specified outcome may moderate changing standards. For example, at the end of the semester, to receive a passing grade in a class, students have a grade they must receive or exceed on the final exam or assignment. This need to obtain a specified grade may moderate their likelihood of changing standards. Also, some people have received a wide range of outcomes both experiencing great success and at times receiving failure feedback. For example, students who have received passing and failing grades in their lifetime have a wide range of experience with feedback on that dimension. Other people have received more stable, consistent feedback on a dimension throughout their lives. For example, some students who have always received As with little variability or students who have always received Ds with little variability have little experience with a range of possible outcomes. People with more experience with a range of outcomes may change their standards more than people who have always received consistent feedback because they have more flexibility in what they define as an acceptable outcome. Another unanswered questions is how do statistically likely standards (having three people rate you in their top three) or established standards (grading systems in most school settings) influence the personal standards that people set? The findings from Study 1 suggest that people may give inadequate weight to base rate information when setting their personal standards. Established standards, such as grading scales may provide a starting point for the personal standards that people set. The grading criteria may vary

PAGE 68

60 depending on the teacher, class or school district, but the student may still use this criteria as a starting point for determining what is personally acceptable to them. For example, students may establish a preliminary standard at the grade that is considered passing in a given class or students may set their standard relative to the mean score of the class on a test on a particular exam. Another interesting questions is whether changes in standards represent a conscious process or an automatic response? An examination of participants reports during the debriefing demonstrated the change in standards was conscious for some participants but not for others. Some participants noted they changed their standards to make themselves feel better about their feedback. Other participants were unaware that they had changed their standards. In addition, it would seem important to know how permanent are peoples standard changes? Shifts in expectations are characterized and momentary and in response to environmental pressures. If changes in standards are not as directly related to environmental pressures or are more driven by accuracy concerns, changes in standards may be relatively permanent (Carroll, Dockery, & Shepperd, 2003). Still another question alluded to earlier in the discussion is what are the consequences of changing standards and how do the consequences of changing standards differ from the consequences of bracing. Perhaps changing standards has less of an impact on a persons self-concept because it is an alteration of the evaluation of the outcome. To change the evaluation of what determines an acceptable outcome does not reflect on the person. A student who redefines an acceptable grade in a class from an A to a C+ may do so with little change to the self-concept. A student who braces and expects to receive a C+ on an exam,

PAGE 69

61 may be forced to adjust his or her view of being a good student to a more mediocre student. Another questions surrounding changing standards is how is the process of changing standards affected by actual feedback? In the introduction we suggested people change their standard in response to information or the anticipation of information. What if the feedback people receive is discrepant from the feedback they defined acceptable? Do people use the feedback they receive, even when it contradicts what they believe a good outcome is? The present studies suggest that, although people may brace for bad news to avoid disappointment, they may also adjust what they consider an acceptable outcome before receiving feedback. Changing standards may work hand in hand with bracing. Shepperd, Ouellette and Fernandez (1996) note that although bracing may serve to avoid disappointment it does not alter the implications of that feedback for the person. Changing standards, making a previously unacceptable outcome acceptable, may be a way that people can alter the implications of the feedback one the persons self-concept.

PAGE 70

APPENDIX DEPENDENT VARIABLES The items on this page ask you to provide a standard for a variety of outcomes. A standard is what you consider the minimally acceptable outcome. In other words, a standard is the lowest outcome that you would find acceptable in a given situation. In this experiment, we will ask you to report the minimum number of people (out of 10) you would like to respond positively to each of the following questions. For the following questions, think about how you would feel if one person evaluated you positively and nine did not, or five people evaluated you positively and five did not, or nine people evaluated you positively and one did not. With this in mind, what number would you would be happy with? That is, what is the minimum number of people in this study you would want to evaluate you positively on each of the dimensions below? What is the minimum number of participants you would like to report that they (circle the single response that represents an acceptable number of participants) want to spend time with you in the future? 0 1 2 3 4 5 6 7 8 9 10 participants found you interesting? 0 1 2 3 4 5 6 7 8 9 10 participants found you friendly? 0 1 2 3 4 5 6 7 8 9 10 participants ranked you as one of the three people they liked best? 0 1 2 3 4 5 6 7 8 9 10 participants want to learn more about you? 0 1 2 3 4 5 6 7 8 9 10 participants 62

PAGE 71

63 The items on this page ask you to provide your expectation for a variety of outcomes. An expectation is the outcome that you predict would occur. In other words, an expectation is the outcome that you think is likely to happen in a given situation. In this experiment, we will ask you to report the number of people (out of 10) you expect to respond positively to each of the following questions. Now, think about what is likely to happen. How many participants do you predict will report that they (circle the single response that represents how many participants you expect to...) want to spend time with you in the future? 0 1 2 3 4 5 6 7 8 9 10 participants found you interesting? 0 1 2 3 4 5 6 7 8 9 10 participants found you friendly? 0 1 2 3 4 5 6 7 8 9 10 participants ranked you as one of the three people they liked best? 0 1 2 3 4 5 6 7 8 9 10 participants want to learn more about you? 0 1 2 3 4 5 6 7 8 9 10 participants

PAGE 72

64 The items on this page ask you to provide your ideal for a variety of outcomes. An ideal is the outcome highest outcome that you would want to occur. In other words, the ideal outcome in a given situation. In this experiment, we will ask you to report the ideal number of people (out of 10) you would want to respond positively to each of the following questions. Now, think about the ideal outcome. How many participants want to report that they (circle the single response that represents the ideal, or highest, number of participants you would want to...) want to spend time with you in the future? 0 1 2 3 4 5 6 7 8 9 10 participants found you interesting? 0 1 2 3 4 5 6 7 8 9 10 participants found you friendly? 0 1 2 3 4 5 6 7 8 9 10 participants ranked you as one of the three people they liked best? 0 1 2 3 4 5 6 7 8 9 10 participants want to learn more about you? 0 1 2 3 4 5 6 7 8 9 10 participants

PAGE 73

LIST OF REFERENCES Alicke, M. D., Klotz, M. L., Breitenbecher, D. L., Yurak, T. J., & Vredenburg, D. S. (1995). Personal contact, individuation, and the better-than-average effect. Journal of Personality and Social Psychology, 68, 804-825. Bach, J. R., & Tilton, M. C. (1994). Life satisfaction and well-being measures in ventilator assisted individuals with traumatic tetraplegia. Archives of Physical Medicine and Rehabilitation, 75, 626-632. Battle, E. S. (1968). Motivational determinants of academic competence. Journal of Personality and Social Psychology, 4, 634-642. Battle, E. S. (1966). Motivational determinants of academic task persistence. Journal of Personality and Social Psychology, 2, 209-218. Breetvelt I. S., & Van Dam, F. S. A. M. (1991). Underreporting by cancer patients: The case of response-shift. Social Science and Medicine, 32, 981-987. Carroll, P., Dockery, K., & Shepperd, J. A. (2003) Exchanging optimism for pessimism in personal predictions. Unpublished manuscript, University of Florida. Dunning, D., Meyerowitz, J. A., & Holzberg, A. D. (1989). Ambiguity and self-evaluation: The role of idiosyncratic trait definition in self-serving assessments of ability. Journal of Personality and Social Psychology, 57, 1082-1090. Festinger, L. (1957). A theory of cognitive dissonance. Stanford: Stanford University Press. Gilovich, T., & Medvec, V. H. (1995). The experience of regret: What, when, and why. Psychological Review, 102, 379-395. James, W. (1890). The principles of psychology (vl. 1). New York: Holt. Locke, E. A. (1967). Relationship of success and expectation to affect on goal-seeking tasks. Journal of Personality and Social Psychology, 7, 125-134 Pervin, L. A. (1963). The need to predict and control under conditions of threat. Journal of Personality, 31, 570-587. 65

PAGE 74

66 Sprangers, M. A. G., & Schwartz, C. E. (2000). Integrating response shift into health-related quality-of-life research: A theoretical model. In C. E. Schwartz & M. A. G. Sprangers (Eds.), Adaptation to changing health: Response shift in quality of life research (pp. 11-23). Washington D.C: American Psychological Association. Shepperd, J. A., Ouellete, J. A., & Fernandez, J. K. (1996). Abandoning unrealistic optimism: Performance estimates and the temporal proximity of self-relevant feedback. Journal of Personality and Social Psychology, 70, 844-855. Taylor, K. M., & Shepperd, J. A. (1998). Bracing for the worst: Severity, testing, and feedback timing as moderators of the optimistic bias. Personality and Social Psychology Bulletin, 24, 915-925. Zemke, R. (1981). Breaking the news to management. In J. Gordon, R. Zemke, P. Jones (Eds.), The best of Training Magazine: Designing and delivering cost-effective training and measuring the results (2nd ed.)(733-736). Minneapolis, MN: Lakewood Publishing.

PAGE 75

BIOGRAPHICAL SKETCH Meredith Terry was born June 26, 1979, in Honolulu, Hawaii, to Michael and Debby Terry. Meredith spent most of early years living in Orange Park, Florida. She graduated from the University of Florida with a bachelors degree in psychology in May 2001. She continued her education at the University of Florida where she completed a Master of Science in psychology in May 2004. 67


Permanent Link: http://ufdc.ufl.edu/UFE0004834/00001

Material Information

Title: Changing Standards: Adjusting the Definition of an Acceptable Outcome
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004834:00001

Permanent Link: http://ufdc.ufl.edu/UFE0004834/00001

Material Information

Title: Changing Standards: Adjusting the Definition of an Acceptable Outcome
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004834:00001


This item has the following downloads:


Full Text











CHANGING STANDARDS:
ADJUSTING THE DEFINITION OF AN ACCEPTABLE OUTCOME














By

MEREDITH LEA TERRY


A THESIS PRESETNED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF
FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF MASTER OF SCIENCE


UNIVERSITY OF FLORIDA


2004
















ACKNOWLEDGMENTS

I am indebted to many people for providing support and encouragement throughout

my work on this thesis. First, I want to thank my advisor, Dr. James Shepperd, who gave

so much of his time, ideas, and energy to this project. Additionally, the thoughtful

comments of my other committee members, Dr. Benj amin Karney and Dr. Mark

Fondacaro, have contributed to this thesis.

I am grateful to my parents who have always supported and fostered my interest in

learning. The support and encouragement provided by my parents and sister helped me

complete this proj ect. Finally, I have had the opportunity to work with some exceptional

graduate students in my time at the University of Florida, especially those in my lab,

Patrick Carroll, Kate Dockery, and Jodi Grace. I appreciate all of the feedback, advice,

and friendship that they have provided.





















TABLE OF CONTENTS


page


ACKNOWLEDGMENT S ................. ................. ii.............


LIST OF TABLES ................ ...............v............ ....


CHAPTER


1 INTRODUCTION AND LITERATURE REVIEW .............. .....................


Changing Standards after Feedback Is Known ................. ..............................2
Changing Standards ................. ...............3.................
Bracing ................. ...............5............ ....



2 STUDY 1 .............. ...............8.....


Overview and Hypothesis............... ...............
M ethod ................. ...............9.................

Participants .............. ...............9.....
Procedures .............. ...............10....
Re sults .................. ............ ... ...............13.....

Comparing Standards across Time ................. ...............13................
Comparing Estimates across Time ................. ...............14................
Comparing Estimates to Standards............... ...............1
Comparing Changing Standards to Bracing ................. .. .. .......... .................1 5
Assessing the Accuracy of Participants' Standards and Expectations ................16

Comparisons to Participants in Hypothetical Interactions .............. ..................18
Discussion ................. ...............19.................



3 STUDY 2 ................. ...............3.. 5......... ...


Overview ................. ...............35.................
M ethod ................ ...............3.. 5..............

Participants. ............. ...............3 5....
Procedures .............. ...............35....

Hypothesis ................ ...............37.................
R e sults................... .... ....... .. .. .... ...............37.

Differences between standards and expectations .............. .....................3












Comparing Expectations across Time ................. ...............38................
Comparing Standards across Time ................. ...............39................
Comparing Changing Standards to Bracing ................. .. .. .......... ................ .40
Assessing the Accuracy of Participants' Standards and Expectations ................41
Ideals s....._._................. ........_ ..........4



4 GENERAL DISCUS SION ................. ...............55.......... ....


APPENDIX


DEPENDENT VARIABLES ................. ...............62.................


LIST OF REFERENCE S ................. ...............65................


BIOGRAPHICAL SKETCH .............. ...............67....


















LIST OF TABLES


Table pg

1 Study 1: Average estimate of what determines a good outcome for each of the
dimensions ........... ..... .._._. ...............24.....

2 Study 1: Standards and expectations from Time 1 to Time 2. ............. .................25

3 Study 1: Correlations between standards and expectations at Time 1 and
Tim e 2. ............. ...............26.....

4 Study 1: Standards versus expectations at Time 1 and at Time 2. ...........................27

5 Study 1: Number of participants who set standards higher, equal to or lower than
expectations. .............. ...............28....

6 Study 1: Correlations between the change in standards and the change in
expectations. .............. ...............29....

7 Study 1: Standard and received outcomes at Time 1 & Time 2 ............... ...............30

8 Study 1: Mean expected and received outcomes at Time 1 & Time 2.....................3 1

9 Study 1: Number of participants who were pessimistic, accurate and optimistic....32

10 Study 1: Mean expected outcome at Time 1 and 2 and hypothetical standard
(minimum acceptable outcome for survey) ................. ..............................33

11 Study 1: Mean desired outcome (hypothetical survey) and mean standard
(minimum acceptable outcome) at Time 1 & Time 2. ............. .....................3

12 Study 2: Standards versus expectations at Time 1 and at Time 2 when standards
and expectations are counterbalanced. ............. ...............43.....

13 Study 2: Expectations at Time 1 and Time 2 by condition. ............. ...................44

14 Study 2: Standards at Time 1 to Time 2 by condition ................. ............. .......45

15 Study 2: Correlations between standards and expectations at Time 1 and at
Tim e 2. ............. ...............46.....










16 Study 2: Correlations between the change in standards and the change in
expectations. .............. ...............47....

17 Correlations between the change in standards and the change in expectation
across both Study 1 and Study 2 (Total N=81). ................ ................ ..........48

18 Study 2: Number of participants changed their standards the same amount that
they change their expectations. ............. ...............49.....

19 Study 2: Correlations between anxiety and change in standards & anxiety and
change in expectations. ............. ...............50.....

20 Study 2: Standard and received outcomes at Time 1 & Time 2............... ................51

21 Study 2: Mean expected and received outcomes at Time 1 & Time 2.....................52

22 Study 2: Standard and ideal outcomes at Time 1 & Time 2. ............. ..................53

23 Study 2: Mean expected and ideal outcomes at Time 1 & Time 2. .........................54
















Abstract of Thesis Presented to the Graduate School of the University of Florida in
Partial Fulfillment of the Requirements for the Degree of Master of Science

CHANGING STANDARDS:
ADJUSTING THE DEFINITION OF AN ACCEPTABLE OUTCOME

By

Meredith Lea Terry


May 2004

Chair: James Shepperd
Maj or Department: Psychology

After receiving an unsatisfactory outcome, people often redefine what they

consider an acceptable outcome. In addition, in anticipation of feedback people often

become pessimistic in their predictions for the future, bracing for bad news. The current

studies examined a) whether people would redefine what they consider an acceptable

outcome prior to receiving feedback and not simply in response to bad news, and b) how

personal standards compare with expectations and actual outcomes. In Study 1,

participants interacted individually with 10 peers and then rated each peer on a variety of

dimensions. Participants also reported what outcome would be acceptable (standard) and

what outcome they expected (expectation) from their peers. Participants provided these

reports twice: immediately after the interaction (Time 1), and again 30 minutes later,

when participants expected their peer evaluations (Time 2). Participants reported lower

standards and expectations at Time 2 than at Time 1. In Study 2, half the participants

were assigned to a no feedback condition, believing they would not receive the









evaluations from their peers. Also in Study 2, the order of the items assessing standards

and the items assessing expectations were counterbalanced. In Study 2, for each of the

five evaluative domains, participants also reported what they viewed as the ideal

outcome. In Study 2, only participants expecting to receive peer evaluations reported

lower standards and expectations at Time 2 than Time 1. In addition, participants' ideals

exceeded their standards and expectations. Analyses reveal changes in expectations are

distinct from changes in standards.















CHAPTER 1
INTRODUCTION AND LITERATURE REVIEW

There is nothing either good or bad, but thinking makes it so.
William Shakespeare, Hamlet Prince of Denmark (Hamlet at II, ii)

At first blush, distinguishing between what constitutes a good versus a bad

outcome would seem to be a relatively easy task. A good outcome is getting a

promotion, getting accepted into a competitive graduate program, or learning that a

suspicious lump is benign. A bad outcome is being passed over for a promotion, getting

rej ected from a competitive graduate program, or learning that a suspicious lump is

cancer. When outcomes fall on a continuum, however, what people define as good

versus bad, satisfactory versus unsatisfactory, or success versus failure, is not always so

clear and often is a matter of perspective. As Hamlet wisely noted, in many instances

outcomes are neither inherently good nor inherently bad. Rather, perceivers impose these

evaluative meanings. Hamlet' s insights emerge from recent upheavals in his family life

and reflect his new regard for Denmark as a prison. This paper acknowledges Hamlet' s

insights but moves beyond merely recognizing that people's evaluations are a product of

their current circumstances. This paper explores instead how evaluations change in

anticipation of outcomes. Specifically, this research examines whether people change

their definition of what constitutes an acceptable outcome over time as information

bearing on the outcome draws near.










Changing Standards after Feedback Is Known

Research from a variety of theoretical traditions reveals that, after learning an

outcome, people sometimes change their standards of what constitutes a satisfactory

outcome to make the outcome more acceptable. Most notably, cognitive dissonance

theory posits that when people experience inconsistency between their thoughts and

actions, they are motivated to reduce the inconsistency (Festinger, 1957). The theory

states that when people think they behaved in a way that is contrary to their self-views,

they experience an uncomfortable feeling called dissonance. One way to remove the

dissonance is to change how the outcome or behavior is evaluated. Specifically, if people

can rationalize that the behavior is not contrary to their self-views, then the dissonance is

alleviated (Festinger, 1957). In other words, people may change what they define as an

acceptable behavior or outcome as a means of reducing dissonance. Likewise, when

faced with a feeling of regret for a past action, the regret may prompt dissonance

reducing behaviors (e.g. a reevaluation of the past behavior) to reduce the pain associated

with regret (Gilovich & Medvec, 1995).

The idea that people will change their standards of what constitutes a satisfactory

outcome has received support in research on management. For example, one way that

managers can contend with criticism is to change the criteria of what is versus is not

important (Zemke, 1981). Evaluating the criticism as unimportant makes the feedback

closer to what the managers had anticipated. For example, Zemke notes that managers

will sometimes acknowledge the validity of criticism they receive yet nevertheless insist

that their management style is the preferred way to run the company (Zemke, 1981).

Similar evidence comes from research on response shift. Research on response

shift was inspired in part by the observation that people with severe chronic illnesses









report no less quality of life than do people without such illnesses (Breetvelt & Van Dam,

1991). Moreover, the quality of life they report is stable (Bach & Tilton, 1994).

Apparently, chronically ill patients shift their definition of what it means to be satisfied

with their health when they experience a change in their health. This research reveals that

patients can maintain consistent satisfaction with their health in the face of actual

declines in health a) by changing the standards they use for comparison, b) by changing

their personal values, and c) by changing how they define the outcome (Sprangers &

Schwartz, 2000). Sprangers and Schwartz (2000) provide the example of a woman with

bone cancer who responds to her deteriorating health by changing what it means to be

satisfied with life. When she becomes confined to a wheelchair she reports that she will

be satisfied with her health as long as she is not bedridden. When she becomes

bedridden, she reports she will be satisfied with her health as long as she can still

socialize with friends.

Changing Standards

Prior research demonstrates that people may reevaluate their outcomes after

receiving feedback (i.e., the managers receiving negative feedback, feelings of guilt or

dissonance associated with past behaviors) or in response to new information (i.e.,

changing health status). We explored the possibility that people will change what

constitutes an acceptable outcome prior to receiving feedback bearing on that outcome

and not simply in response to feedback that is threatening. We introduce the term

changing standards to refer to adjusting the definition of what constitutes an acceptable or

satisfactory outcome in anticipation of information bearing on that outcome. We believe

that changing standards allows a person the benefit of receiving an outcome or feedback

that they view as acceptable. By changing standards, people can preemptively avoid









uncomfortable feelings associated with unacceptable feedback. For example, a student

about to take a test may only be satisfied by receiving an "A." Later, with impending

feedback, the student may adjust what is satisfactory from an "A" to a "C+". By

changing standards, there is an increase the likelihood that the person will receive an

acceptable outcome.

People may opt to proactively change their standards of what constitutes an

acceptable outcome instead of readjusting their evaluations after receiving feedback for

two reasons. First, William James described self-esteem as the ratio of success to

pretensions. That is, self-evaluation reflects a comparison of outcomes to their wants and

goals (James, 1890). Decreasing 'pretensions' (the wants and goals) is one way to increase

self-esteem, and people may preemptively change their standard to ensure that their

outcomes match their pretensions. By lowering the standard of what defines an

acceptable outcome, people avoid the negative affect and negative self-evaluations that

arise from receiving outcomes that are below their aspirations.

A second reason people may be motivated to change their standards is to

accurately predict and control their world (Pervin, 1963). As such, people may

preemptively lower their standard of what constitutes an acceptable outcome so they can

predict their world. Failing to change the standard and then receiving an outcome that is

viewed as unacceptable may be interpreted as an inability to control and predict future

outcomes. Of course, distinct from these reasons is the possibility that people may also

change their standards in response to new information. With new information, people

may realize that previous standards were unrealistic. Our interest in this paper is in

changing standards in anticipation of receiving feedback.









Bracing

Changing standards in anticipation of information bearing on the outcome shares

some commonality with bracing for bad news. With bracing, people lower their

expectations of what is likely to happen in anticipation of feedback (Shepperd, Ouellette,

& Fernandez, 1996; Taylor & Shepperd, 1998). One reasons people may lower their

expectations is that bracing reduces the likelihood of disappointment should they receive

an unexpected outcome. Evidence for bracing comes from a study in which students

estimated their exam score prior to receiving it. Although participants were optimistic

about the score they would receive three weeks prior the exam, they were realistic

immediately after the exam. Moreover, this realism persisted three days later at the

beginning of the class in which they would receive their grades. At the end of class,

however, as the professor called students' names to return their exams, students became

pessimistic, estimating a score significantly lower on average than they actually received

(Shepperd et al., 1996). Additional evidence for bracing comes from a second study in

which college sophomores, juniors and seniors estimated their salary on their first j ob

after college. Whereas seniors estimated a lower starting salary in April (two weeks prior

to graduation for the seniors) than in January (three months before graduation), the

sophomores and juniors, for whom graduation was over a year away, did not (Shepperd et

al., 1996).

Changing standards is similar to bracing in many ways. First, they both can have

implications for how people feel about future outcomes. When people brace, they are

better prepared for a previously unexpected outcome. When people change their standard,

they make an unacceptable or negative outcome more acceptable or less negative.

Second, both bracing and changing standards involve changing a previously set









benchmark for evaluating outcomes. Third, they both occur in response to potential

feedback that may challenge the person's self-concept.

However, changing standards is distinct from bracing in that it reflects a shift in

what people define as acceptable rather than a shift it what people expect to occur.

Although changes in standards and expectations may covary, it is possible for people to

shift one but not the other. That is, people may change their expectations but not their

standards, or change their standards but not their expectations. For example, a person

awaiting test results for a sexually transmitted disease may brace for possible bad news,

yet remain steadfast in the belief that testing positive is an undesirable outcome. That is,

the expectation may change while the standard remains constant. Conversely, athletes

may expect to win a competition, but in the moments prior to the match, regard losing as

an acceptable outcome provided they play as hard as they can. That is, the standard may

change while the expectation remains constant. Of course, in some instances the two will

covary, and may even be causally related. A shift in expectations of what will happen

may prompt a shift in standards of what is acceptable. For example, students expecting a

low grade in a class may, in response, lower the standard of what is an acceptable grade.

Research on bracing shows that people will lower their expectations to prepare for

future feedback. Other research shows that after people receive an outcome, they

sometimes change their previous evaluations to reduce dissonance or make the new

feedback appear more pleasant. But, research has yet to address whether people may

alter the standard at which they define an outcome as good or acceptable before they

receive feedback. Although people are likely to adjust what they define as an acceptable

outcome after receiving feedback (to reduce dissonance, reduce regret, or in response to










new information), the proposed research focuses specifically on the effect of anticipated

feedback on people' s definitions of what constitutes an acceptable versus unacceptable

outcome.















CHAPTER 2
STUDY 1

Overview and Hypothesis

Study 1 was designed to provide a preliminary test of whether the mere

anticipation of outcome feedback prompts people to change their standards of what they

define as an acceptable outcome. To this end, we needed an outcome that was important

to people, yet also continuous and for which there was no clear cutoff or preexisting

definition of what constituted an acceptable outcome. We also needed an outcome that

was uncertain, allowing room for people to shift their standards. We opted against

adopting the TAA paradigm used in past research paradigm. Although participants

clearly cared about whether they tested positive for TAA deficiency, testing positive vs.

negative for TAA deficiency is a dichotomous outcome. Likewise, we opted against

examining performance on an intelligence test because people often hold pre-existing

ideas about what constitutes an acceptable outcome. In the end we developed a new

experimental paradigm in which people interacted with and evaluated several peers, and

then reported their standards of what would be an acceptable evaluation from their peers.

Our paradigm was modeled after a popular format for single people to meet potential

dating partners called "speed-dating."

Participants took part in one of two groups. The first group of participants

imagined they met 10 people and then reported what they would regard as an acceptable

outcome on a number of dimensions. Our purpose was to determine the baseline for what

constitutes an acceptable outcome in a social situation. We were solely interested in the









average definitions of an acceptable outcome for each of the dimensions and how those

estimates compared with the responses of participants who actually engaged in the

interactions.

The second group of participants interacted individually with 10 peers then rated

each peer on a number of dimensions. For each dimension participants also reported what

they regarded as an acceptable evaluation from their peers (standard) and how they

expected to be evaluated by their peers (expectation). Participants provided these

estimates twice: immediately after interacting with all peers (Time 1), and again just

before receiving their peer evaluations (Time 2).

We hypothesized that participants would report both a lower expectation and a

lower minimum standard for the outcome at Time 2 than at Time 1. We also explored

whether people lower their standards or criteria of what defines a satisfactory outcome in

anticipation of information bearing on the outcome. Finally, we explored when and how

the estimates made by participants who engaged in the interactions compared with the

estimates of participants who merely imagined engaging in the interactions.

Method

Participants

The first group of participants comprised Introductory and Upper Level

Psychology students (N =50) who completed the hypothetical survey as part of an

unrelated study or an in-class proj ect.

The second group of participants comprised Introductory psychology students (27

females, 12 males) participated in one of four sessions in partial fulfillment of a research

participation requirement and were run in groups of up to 11 people. When fewer than 1 1










participants arrived for the experiment, one or more research assistants posed as a

participant to bring the number to 11.

Procedures

The first group of participants, the hypothetical survey group, completed a survey

in which they imagined that they had just met 10 people. They also imagined that each of

the ten people rated them on a number of dimensions. Participants reported what they

would define as an acceptable outcome for each of the dimensions.

On arriving for the experiment, the second group of participants learned that the

purpose of the study was to examine first impressions and that they would interact with

and then evaluate the other 10 participants in the study. The second group of participants

randomly received nametags each printed with the name of a different color. Each

nametag had the name of a color on it. We provided participants with the names of

colors rather than their own names for several reasons. First, the colors allowed us to

establish in advance a systematic order in which participants would interact and evaluate

each other. Second, assigning participants colors preserved the anonymity of

participants' ratings. Rather than evaluating each person according to name, participants

evaluated each person by color. In addition, when making their evaluations, participants

identified themselves only by color.

After drawing a nametag, each participant received a manila envelope

containing the order in which they would interact with other participants and a "first

impressions" form. The experimenter explained that during the course of the experiment,

participants would engage in 10 interactions, one with each of the other 10 participants in

the study. When the experimenter signaled, participants were to find their first interaction

partner and engage in a 2 min conversation.









The experimenter explained that participants could talk about anything that arises

naturally in conversation such as hobbies, hometowns, and year in school. However,

they were not to talk about the experiment, the people in the experiment, or other

experiments in which they had participated. The experimenter further stated that after

each interaction period, participants would have 30 s to jot down on the "first

impressions" form notes that would help them remember the person they just met. The

experimenter would then prompt them to move to the next interaction partner. The

experimenter explained that the notes were to help them evaluate each of their fellow

participants when all interactions were complete. Built into the schedule were three

"breaks" for each participant when they could write down additional notes. When it was

clear that all participants understood the procedures, the experimenter initiated interaction

phase of the experiment.

After completing the 10 interactions, participants were seated and received an

evaluation questionnaire from the experimenter. The questionnaire first asked

participants to rank their fellow participants (using their "color" names) starting with the

person they liked the most and ending with the person they liked the least. Second,

participants indicated (yes vs. no) in response to each color name whether they a) found

the person friendly, b) would rank the person in their top three in terms of liking, c)

would want to spend time with this person in the future, d) would want to learn more

about the person, and e) found the person interesting. The experimenter asked

participants to respond honestly and assured them that their responses were anonymous.

On completion of the questionnaire, participants returned it and their first impressions

form to the manila envelope, which was then collected by the experimenter and given to a









second experimenter with instructions to enter the data in the computer. That second

experimenter left the room.

Participants next received a "reflection" questionnaire comprising items that

examined how participants felt about the interactions and the impressions they made. In

the first part, participants responded to five items asking them to indicate "the minimum

number of participants you would want to report that you" a) were friendly, b) were

ranked in the person 's top three in terms of liking, c) would want to spend time with you

in the future, d) would want to learn more about you, and e) were interesting. The

instructions directed participants to circle a number from 0 to 10 indicating the number

that they defined as acceptable for each of the previous items. This number served as the

standard. In the second part, participants responded to same five items. However, rather

than reporting what they felt would be acceptable, participants reported what they

expected was likely to happen (e.g., "How many participants do you predict will want to

spend time with you in the future?"). Again, participants could circle anywhere form 0 to

10 participants in response to each item. This number served as the expectation.

Soon after all participants had completed filler questionnaires, the second

experimenter returned and announced, after a brief discussion with the first experimenter,

that the evaluation questionnaire data were entered into the computer and that they would

learn after all how they were evaluated by their fellow participants. The experimenter

then distributed to each participant a sealed manila envelope with his or her color name

printed across the top. The experimenter explained that, before participants opened the

envelope they would need to complete two more questionnaires. The first questionnaire

consisted of six items measuring state anxiety (nervous, anxious, relaxed, at ease, and









worried) and asked participants to respond according to how they felt at this moment

(e.g., 1 = not at all nervous, 9 = very nervous). The second questionnaire was identical to

the "reflection" questionnaire that participants completed earlier with the exception that

the "prediction" items appeared in a different order. Likewise, the "minimum number"

items appeared in a different order. The experimenter explained that, because

participants were now receiving feedback and the feedback might influence their

thinking, the experimenter would discard the "reflection" questionnaire that the

participants completed earlier and only use in data analysis the responses on the

questionnaire participants were now completing. When all participants had completed

these Einal questionnaires, they were debriefed.

Results

We averaged across all participants who imagined they had engaged in the

interactions to create an average standard for each of the dimensions. The average

standards ranged from 4.7 (the number of people to rate you in their top three) to 7.8 (the

number of people to Eind you friendly). The average across all dimensions was 5.9. We

performed no tests to test for differences between these standards because we were not

interested in how standards differ across domains. Our interest was solely in establishing

a baseline for the standard Table 1 presents the means for each dimension.

Comparing Standards across Time

Did participants who engaged in the interactions change their standard of what

constituted an acceptable outcome in anticipation of evaluations from their fellow

participants? Table 2 presents the mean responses for all Hyve items and the average of

the five items. As predicted, averaging across the Hyve items, the outcome that

participants reported would be acceptable at Time 2 was significantly lower than the









outcome that participants reported would be acceptable at Time 1. Analysis of the

individual items revealed that participants reported a significantly lower standard at Time

2 than at Time 1 for all items except for the item asking how many people they wanted to

Eind them friendly.

Comparing Estimates across Time

Did participants who engaged in the interactions lower their expectations of how

they would be evaluated in the moments prior to receiving feedback? The bottom half of

Table 2 displays the mean expectation at Time 1 and Time 2 for each of the Hyve items

and for the average across the Hyve items. Consistent with prior research on bracing

(Shepperd et al., 1996), averaging across the five items, participants reported a

significantly lower expectation at Time 2 than at Time 1. Analysis of each of the five

individual items revealed that participants reported a significantly lower expectation at

Time 2 than at Time 1 for all items except for the item asking participants to estimate the

number of people that would want to spend time with them in the future.

Comparing Estimates to Standards

Table 3 provides the correlations between the standards and expectations for Time

1 and Time 2 for each dimension. We were not surprised that standards and expectations

were correlated. They are both benchmarks by which people evaluate their outcomes.

Thus, we would anticipate that the participants who reported high standards would also

report high expectations and vice versa. However, the correlation coefficients do not

speak to whether the standards and expectations are the same thing. We were more

interested in how the average standard across participants compared to the average

expectation.










We next compared participants' expectations of how they would be evaluated by

their peers to their standards of what was an acceptable evaluation. As evident in Table

4, for all items at both Time 1 and Time 2, the evaluation participants' expected from

their peers fell below the evaluation they viewed as acceptable. In sum, participants

systematically report a lower standard than expectation at both Time 1 and Time 2.

In addition, for participants, we also subtracted their expectation from their

standard. Table 6 presents the distribution of participants who set lower standards than

expectations, the same standard as expectation, and higher standards than expectations

for each dimension both at Time 1 and Time 2. As is evident in Table 5, the majority of

the participants set a higher standard than expectation at Time 1. Although less

pronounced at Time 2, there is still a tendency for participants to report a higher standard

than expectation. In sum, participants on average set higher standards than expectations

at both Time 1 and Time 2. We entertain several possible explanations for the difference

between participants' standards and their expectations in the discussion.

Comparing Changing Standards to Bracing

We were not only interested in showing that standards and expectations were

different, but also, that they measured different constructs. Table 6 displays the

correlations between the change in standards the change in expectations for each

dimension. In only one instance were the change in standard and the change in

expectation significantly correlated. The absence of a significant correlation reveals that

whether or participants changed their standard was unrelated to whether or not they

changed their expectations.









Assessing the Accuracy of Participants' Standards and Expectations

Because participants evaluated each other, we could examine how participants'

standards compared with the evaluations they received from their peers. Table 7

compares the evaluations participants received from their peers with the standard

participants reported would constitute an acceptable outcome for each of the Hyve items.

At Time 1, the standard that participants reported that would be an acceptable did not

significantly differ from the actual feedback their peers provided for all items except the

number of people who ranked the participant in their top three favorite people. Because

participants significantly lowered their standard of an acceptable outcome from Time 1 to

Time 2, the minimum standard participants reported at Time 2 was significantly lower

than the actual feedback they received for all fiye dimensions.

We also examined how participants' expectations of what would happen

compared with the evaluations they received from their peers. Table 8 compares the

actual evaluations that participants received from their peers with participants'

expectations of what was likely to happen for each of the Hyve items. Consistent with prior

research on bracing (Shepperd et al., 1996), averaging across the Hyve items, participants'

expectations of how they would be evaluated by their peers was significantly lower than

the actual evaluation participants received from their peers at both Time 1 and Time 2.

Analysis of each of the five individual items revealed that expectations fell below

outcomes both at Time 1 and Time 2 for all items except for the item asking participants

to estimate the number of people that would rank the participant in their top three favorite

people.

The analyses just described explore differences in mean responses. Mean

responses, of course, can be influenced by extreme responses and do not reveal whether









or how many individual participants were accurate or inaccurate in their expectations of

what would happen. Table 9 presents the number of participants who were accurate and

inaccurate when expectations were compared with actual outcomes. To provide a strong

test of our hypothesis, we were liberal in how we defined accuracy. Specifically, we

defined an estimate as accurate if the estimate fell within + or one place from the actual

outcome. Thus, for example, if a participant estimated 7 people would Eind them

friendly, we classified the estimate as accurate if 5, 6 or 7 people rated them as friendly

(7 +/- 1). If the expectation fell more than one place below the outcome, we classified the

estimate as pessimistic. If the expectation fell more than one place above the outcome,

we classified the estimate as optimistic. We repeated this system of classification when

we compared participants' standards with their outcomes.

The distribution of participants in Table 9 indicates support for accuracy in their

standards at Time 1. Taking a closer look at the average standard at Time 1 suggests that

the pattern of responses is fairly evenly dispersed around accuracy. The pattern also

suggests more participants were accurate with their average standard at Time 1 than were

optimistic or pessimistic. In sum, there appears to be less systematic favoring of either

optimism or pessimism. On the other hand, the frequencies for average expectation at

Time 1, average standard at Time 2, and average expectation at Time 2 suggest a

tendency for more participants to report pessimism. Also, in many of the specific

dimensions, as feedback approached (Time 2), there was an increase in people making

estimates that fell under the column indicating a pessimistic view, and a decrease in the

number of people who subscribed to an overly optimistic view. Recall that the mean

responses among participants showed that the standards at Time I were not significantly









different from the actual peer evaluations. Also recall that the standards at Time 2 and

the expectations at Time 1 and Time 2 were significantly lower than the actual peer

evaluations. In sum, mean responses indicated accuracy for the standards set at Time 1.

In addition, the distribution of individual participant's responses also suggest more of a

tendency for accuracy at the standard at Time 1 and more of a pessimistic tendency at the

standard at Time 2 and the expectations at both Time 1 and Time 2.

Comparisons to Participants in Hypothetical Interactions

It is possible that participants actually have standards much higher than the

feedback they actually received and that the pressure of the experimental situation or the

interaction task itself created a feeling of anxiety. The anxiety may have led these

participants to report lower standards at Time 1, prompting greater accuracy in the

standards reported at Time 1. Although, we did not assess anxiety at Time 1, we view

this explanation for participants' accuracy at Time 1 unlikely. Participants' estimates at

Time I were identical to the estimates of participants in the hypothetical condition.

Table 10 compares the expectations for the participants who engaged in the

interaction to the standards provided by the participants in the hypothetical survey. For

all five items, the expectations participants set at Time 1 and Time 2 were significantly

lower than the independent sample estimates of an acceptable outcome. Table 1 1

compares the standards at Time 1 and Time 2 from participant who engaged in the

interaction to independent sample from the hypothetical survey. As is evident, we found

no significant difference between the standards set at Time 1 by participants who

engaged in the interactions and the standards of participants in the independent sample.









Discussion

As predicted, participants estimated a lower evaluation at Time 2, when feedback

was imminent, than at Time 1, when feedback was unexpected. More importantly,

participants reported a lower standard at Time 2 than at Time 1. Also, participants'

average standard was consistently higher than the average expectation. Although the

standards were correlated with the expectations, the change in standard was rarely

correlated with the change in expectations.

In terms of when the benchmarks reported by the participants were consistent

with their actual peer evaluations, participants' standards at Time I were most consistent

with their peers' evaluations, whereas participants' standards at Time 2 and expectations

at both Time 1 and Time 2 were inconsistent with their actual peer evaluations. What

participants first estimated to be an acceptable outcome foreshadowed the peer

evaluations they actually received. Our participants' first estimate of the outcome they

desired was actually consistent with how they were evaluated by their peers. Also, the

standard at Time I was not significantly different than the standard provided the

hypothetical interaction participants.

It may seem that because participants lowered both their standard and their

expectations, standards are not distinct from expectations. As we previously mentioned,

often what people want is different from what people think is likely to happen. We are

not the first to regard standards and expectations as distinct. Several other researchers

have also argued that the two constructs differ (Battle, 1965; Battle, 1966; Locke, 1967).

Our findings, however, provide some evidence that the two constructs are distinct. First,

at Time 1 and Time 2, recall that, on average, participants set significantly higher

standards than expectations. Simply put, each time participants were given the










opportunity, they reported standards that were significantly different from their

expectations and this direction favored a higher standard than expectation. In addition to

group averages, the pattern of frequencies in Table 9 suggests that the maj ority of

participants set higher standards than expectations.

Second, and most important, we believe a distinction between standards and

expectations results from the fact that participants' change in standards was not correlated

with their change in expectation. The lack of correlation indicates the extent to which

the participants changed their standards differs quantitatively from how they changed

their expectations. A person who lowers their standards a lot may not lower his or her

expectation as much, or vice versa. Although more research assessing the functional

differences between standards and expectations needs to be done, we, with caution, take

the lack of correlation between the change in standards and the change expectation as

some preliminary support that standards and expectations are distinct constructs.

As mentioned earlier, participants set significantly higher standards than

expectations at both Time 1 and Time 2. This is an intriguing finding because it suggests

that participants desired an outcome that they believed they would not obtain. There are

several reasons that could account for why participants' standards exceeded their

expectations. We should note at this time, that all participants completed expectation

items shortly after completing the standard items. First, it is possible that as time elapsed

between completing the standard items and the expectation items, participants became

increasingly anxious. The anxiety, in turn, led participants to display lower estimates for

the expectation items than the standard items. We view this explanation unlikely because

the amount of time that elapsed between completing the two sets of items was small, less









than a minute. In addition, the reported standards were higher than the expectations at

both Time 1 and Time 2. The discrepancy exists at Time 1 even though there was little to

no reason to feel anxious, as they were not expecting feedback.

A second possibility is that participants were displaying modesty in their

expectations so as not to appear presumptuous or self-aggrandizing. After first reporting

what they believed to be an acceptable outcome, participants felt moved to then report

lower expectations to appear modest or even self-deprecating. We also view this

explanation unlikely because participants were not reporting their estimates out loud, but

rather they were privately reporting them on paper and therefore had less pressure to keep

up a modest appearance because they had no public audience for which they could

promote a modest appearance. Also, participants were instructed to turn in the forms in

the envelopes they received so there is little reason to think that the participants would

believe the experimenters would have access to their responses at that time. Therefore,

participants were thus under little pressure to appear modest. In addition, the time

between estimates, less than a minute, would have allowed little time to consider how

their reported standards would be perceived.

A third possibility is that people may want to be liked by everyone and so they

report unrealistically high standards when asked to indicate their minimum acceptable

number. In contrast, when asked to indicate their expectations, participants may have felt

constrained by what they felt was likely to happen, which prompted them to report

expectations below their standards. We find two problems with this explanation. First,

we did not get a ceiling effect in participants' response to the items asking about their

standards (what they wanted to happen). Second, participants' standards at Time 1










generally did not differ from the evaluations they received from their peers. Both

Endings suggest that participants were not idealistic in their standards.

The Einal, and most likely possibility is that participants had high standards for

what they feel is acceptable and felt at both Time 1 and Time 2 that they that were

unlikely to receive feedback that supported this standard. This explanation relies on the

distinction between what people want and what they expect they are likely to receive.

This explanation deserves further investigation.

One other Einding that emerged in participants' standards and expectations

deserves attention. Specifically, the mean response to the items asking participants how

many people they wanted to rank them in their top 3 suggests a degree of unrealistic

optimism. Out of ten people, statistical odds would say people should expect on average

three people to rank them in their top 3. However, the participants in our study reported

an average standard of 4.3 at Time 1.

This statistical and logical failure of the participants may account for why the

only standard at Time 1 that was significantly different from what actually happened was

the standard for the number of people to rank you in their top 3 (see Table 8). Also in

Table 10, many more people were optimistic (across all times and both standards and

expectations) in response to this item than in response to other items.

There are several reasons for this anomaly in the data. First, participants may have

a poor understanding of statistics and may have failed to see that their number exceeded

what, on average, was statistically likely. Second, it is likely that the higher responses to

other items (the number of people to Eind you friendly, interesting) anchored the

participants and they adjusted poorly from that higher point. For all of the other









dimensions, participants reported minimum standards that were higher so perhaps those

relatively high standards influenced the standard for the number of people to rank you in

their top three favorite people.

A third possibility is based on the idea that the top three dimension is the one

dimension that actually provides a clear answer to what is an objectively good or

acceptable outcome. Statistically, to exceed the statistically average of 3 is good, to have

less than three people rank you in their top 3 is bad. A common finding in western

cultures is the better-than-average effect (Alicke, Klotz, Breitenbecher, Yurak, &

Vredenburg 1995, Dunning, Meyerowitz, & Holzberg, 1989). Research finds that many

people in western cultures (often well over the majority) rate themselves as better than

average on a variety of dimensions. It seems people are reluctant to think that they may

be below average and even being average is viewed as negative (Alicke et al., 1995,

Dunning et al., 1989). Perhaps the better-than-average effect extends to setting standards

for an acceptable outcome, resulting in our participants, on average, being unwilling to

claim that they have an average standard in terms of being well liked by others. Perhaps,

not only do the maj ority of people believe they are better than average drivers, they also

(by self-report), set standards that indicate they are better than average in terms of

popul arity.







24


Table 1: Study 1: Average estimate of what determines a good outcome for each of the
dimensions


Item

Find you friendly
Find you interesting
Rank you in their top three
Learn more about you
Spend time with you in the future
Mean Across All Items


Average number of
people out of ten
7.8
6.3
4.7
5.5
5.0
5.9










Table 2: Study 1: Standards and expectations from Time 1 to Time 2.
Standard Time 1 Time 2
Number of people you would like to... M M t p
Find you friendly 7.4 7.1 1.18 ns
Find you interesting 6.3 5.2 3.45 .002
Rank you in their top three 4.3 3.7 3.27 .003
Learn more about you 5.5 5.1 2.11 .05
Spend time with you in the future 5.5 4.6 4.33 .0001
Mean Across All Items 5.8 5.1 4.29 .0001
Expectation Time 1 Time 2
Number of people you expect to... M M t p
Find you friendly 6.8 6.2 2.40 .03
Find you interesting 5.0 4.5 2.66 .02
Rank you in their top three 3.5 3.0 2.86 .007
Learn more about you 4.8 4.5 2.16 .04
Spend time with you in the future 4.3 4.1 1.24 ns
Mean Across All Items 4.9 4.5 3.75 .0006






26


Table 3: Study 1: Correlations between standards and expectations at Time 1 and Time 2.

Item Time 1 Time 2
Find You Friendly .63 .65
Find You Interesting .42 .46
Rate You in Their Top 3 .58 .61
Learn More About You .57 .63
Spend Time in Future .68 .62
Note: All correlations significant at p < .01.












t
1.79
3.79
2.30
2.25
3.79


.09
.002
.03
.04
.001


4.9 3.26 .003

6.2 2.23 .03
4.5 1.93 .07
3.0 2.30 .03
4.5 2.21 .04
4.1 1.69 .10
4.5 2.34 .03


Table 4: Study 1: Standards versus expectations at Time 1
Time 1 Standard
Number of people you would like to... M
Find you friendly 7.4
Find you interesting 6.3
Rank you in their top three 4.3
Learn more about you 5.5
Spend time with you in the future 5.5
Mean Across All Items 5.8
Time 2
Find you friendly 7.1
Find you interesting 5.2
Rank you in their top three 3.7
Learn more about you 5.1
Spend time with you in the future 4.6
Mean Across All Items 5.1


and at Time 2.
Expectation
M
6.8
5.0
3.5
4.8
4.3











Table 5: Study 1: Number of participants who set standards higher, equal to or lower than
expectations.
Standard < Standard = Standard >
Expectation Expectation Expectation
Time 1
Find you friendly 8 11 20
Find you interesting 7 6 26
Rank you in their top three 6 10 23
Learn more about you 7 11 21
Spend time with you in future 6 5 28
Time 2
Find you friendly 5 15 19
Find you interesting 8 13 18
Rank you in their top three 5 19 15
Learn more about you 6 15 18
Spend time with you in future 8 15 16
All numbers in the table refer to the number of participants out of 39 who fall in each
category.







29


Table 6: Study 1: Correlations between the change in standards and the change in
expectations.


Dimension
Find You Friendly
Find You Interesting
Rate You in Their Top 3
Learn More About You
Spend Time in Future
p<.01*


Correlation
*.44
-.06
.21
.24
.24














p
.10
ns
.001
ns
ns


<1 ns

2.23 .04
4.52 .001
1.84 .08
2.08 .05
1.43 ns
1.97 .06


Table 7: Study 1: Standard and received outcomes at Time 1 & Time 2.
Time 1 Standard Outcome
Number of people you would like to... M M
Find you friendly 7.4 8.1
Find you interesting 6.3 6.9
Rank you in their top three 4.3 2.8
Learn more about you 5.5 6.0
Spend time with you in the future 5.5 5.2
Mean Across All Items 5.8 5.8
Time 2
Find you friendly 7.1 8.1
Find you interesting 5.2 6.9
Rank you in their top three 3.7 2.8
Learn more about you 5.1 6.0
Spend time with you in the future 4.6 5.2
Mean Across All Items 5.1 5.8


t
1.71
1.23
2.86
1.21
<1










Table 8: Study 1: Mean expected and received outcomes at Time 1 & Time 2.
Expected Received
Time 1 Outcome Outcome
Number of people you expect to... M M t p
Find you friendly 6.8 8.1 3.11 .004
Find you interesting 5.0 6.9 4.35 .0001
Rank you in their top three 3.5 2.8 1.62 ns
Learn more about you 4.8 6.0 2.70 .02
Spend time with you in the future 4.3 5.2 1.98 .06
Mean Across All Items 4.9 5.8 2.54 .02
Time 2
Find you friendly 6.2 8.1 4.1 .0002
Find you interesting 4.5 6.9 6.0 .0001
Rank you in their top three 3.0 2.8 <1 ns
Learn more about you 4.8 6.0 3.39 .002
Spend time with you in the future 4.1 5.2 2.36 .03
Mean Across All Items 4.5 5.8 3.84 .0004









Table 9: Study 1: Number of participants who were pessimistic, accurate and optimistic.
Pessimi stic Accurate Optimi sti c
Average Standard at Time 1 11 18 10
Average Standard at Time 2 14 18 7
Average Expectation at Time 1 13 20 6
Average Expectation at Time 2 16 23 1
Find you friendly
Standard at Time 1 14 20 5
Standard at Time 2 16 15 8
Expectation at Time 1 16 19 4
Expectation at Time 2 20 17 2
Find you interesting
Standard at Time 1 17 11 11
Standard at Time 2 20 27 3
Expectation at Time 1 24 11 4
Expectation at Time 2 23 15 1
Put you in their top 3
Standard at Time 1 5 14 20
Standard at Time 2 5 21 13
Expectation at Time 1 6 20 13
Expectation at Time 2 7 19 13
Want to learn more about you
Standard at Time 1 12 17 10
Standard at Time 2 15 17 7
Expectation at Time 1 20 11 8
Expectation at Time 2 20 13 6
Want to spend time with you
Standard at Time 1 11 12 16
Standard at Time 2 14 16 9
Expectation at Time 1 13 18 8
Expectation at Time 2 15 17 7
All numbers in the table refer to the number of participants out of 39 who fall in each
category.
The 'pessimistic' label indicates lower expectation or standard than what actually
happened.
The 'accurate' label indicates near the same expectation or standard as what actually
happened.
The 'optimistic' label indicates higher expectation or standard than what actually
happened.
The rows indicating average count the number of people whose average standard or
expectation (across all five dimensions) is lower, equal to, or higher than the average
outcome they received (across all five dimensions).









Table 10: Study 1: Mean expected outcome at Time 1 and 2 and hypothetical standard
(minimum acceptable outcome for survey).
Expected Hypothetical
Outcome Standard
M M T p
Time 1
Find you friendly 6.8 7.8 3.10 .004
Find you interesting 5.0 6.3 4.02 .001
Rank you in their top three 3.5 4.7 3.47 .002
Learn more about you 4.8 5.6 2.39 .03
Spend time with you in the future 4.3 5.0 2.06 .007
Time 2
Find you friendly 6.2 7.8 4.26 .0001
Find you interesting 4.5 6.3 6.21 .0001
Rank you in their top three 3.0 4.7 5.69 .0001
Learn more about you 4.5 5.6 3.25 .003
Spend time with you in future 4.1 5.0 2.86 .007










Table 11: Study 1: Mean desired outcome (hypothetical survey) and mean standard
(minimum acceptable outcome) at Time 1 & Time 2.
Acceptable Hypothetical
Outcome Standard
M M T p
Time 1
Find you friendly 7.4 7.8 1.09 ns
Find you interesting 6.3 6.3 <1 ns
Rank you in their top three 4.3 4.7 <1 ns
Learn more about you 5.5 5.5 <1 ns
Spend time with you in the future 5.5 5.0 1.13 ns
Mean 5.8 5.7 .29 ns
Time 2
Find you friendly 7.1 7.8 1.75 .09
Find you interesting 5.2 6.3 3.40 .002
Rank you in their top three 3.7 4.7 2.76 .009
Learn more about you 5.1 5.6 1.15 ns
Spend time with you in future 4.6 5.0 1.22 ns
Mean 5.1 5.9 1.74 .09















CHAPTER 3
STUDY 2

Overview

The primary purpose of Study 2 was to replicate the findings of Study 1, but also

to show that the change in standard occurs in response to impending feedback. Study 2

thus included a no feedback (control condition) in which participants believed they would

not receive evaluations from their peers. A secondary purpose of Study 2 was to examine

whether the discrepancy between standards and expectations was due to the order of the

questionnaires. In Study 1 participants always completed the standard items before the

expectation items. In Study 2 we counterbalanced the order of these items.

Method

Participants.

Introductory psychology students (N= 53) participated in partial fulfillment of a

research participation requirement and were run in groups of 11 people with the

exception of one session in which there were only 9 participants.

Procedures

The procedure was identical to Study 1 with the following exceptions. First, after

participants interacted and evaluated each other, the experimenter randomly separated

participants into two rooms under the pretext of providing them more room to complete

the next set of questionnaires. All participants then provided their standards of what

would be an acceptable outcome and their outcome expectations on each of the five

measures (Time 1).










Second, the Time 1 evaluation materials provided additional information about the terms

"standards" and "expectations". Specifically, participants read: "a standard is what you

consider the minimally acceptable outcome. In other words, a standard is the lowest

outcome that you would find acceptable in a given situation." Participants also read: "an

expectation is the outcome that you predict would occur. In other words, an expectation is

the outcome that you think is likely to happen in a given situation." Third, we

counterbalanced the items assessing standards with the items assessing expectations.

Fourth, for each of the five evaluative domains, reported what they viewed as the ideal

outcome.

Fifth, after we collected the responses and participants had completed several

filler questionnaires, participants assigned to the feedback room learned, as in Study 1,

that they would receive the evaluations of the peers. Participants in the no feedback room

kept all questionnaires they completed until the end of the experiment and continued to

believe that they would not receive the evaluations of their peers. All participants then

completed the same expectation, standard and ideal items they completed at Time I with

the items assessing participants' expectation once again counterbalanced with the items

assessing participants' standards. In addition, as in Study 1, participants completed items

measuring state anxiety. When all participants had completed these final questionnaires,

they were reunited, debriefed to assess their perceptions of the experiment, and thanked

for their participation.










Hypothesis

We tested the following hypotheses:

1. Standards vs. Expectations: Replicating the pattern in Study 1, participants' standards
would be significantly higher than participants' expectations.

2. Expectations: Participants in the feedback condition would report a significantly
lower expectation at Time 2 than at Time 1, whereas participants in the no feedback
condition would report no difference in expectations from Time 1 to Time 2.

3. Standards: Participants in the feedback condition would report a lower standard at
Time 2 than at Time 1, whereas participants in the no feedback condition would
report no difference in standards from Time 1 to Time 2.

4. Change in Standards vs. Bracing: Participants change in standards would be
uncorrelated with their change in expectations. In addition, the majority of
participants would report a change in standards that was quantitatively different from
their change in expectation.

5. Anxiety Reports: Anxiety reports would be higher among participants in the feedback
condition than among participants in the no feedback condition. Also, anxiety reports
would correlate positively with bracing. In addition, anxiety reports would correlate
positively with the magnitude of change in standard.

6. Comparisons to actual outcomes: Replicating the Eindings of Study 1, the evaluations
participants receive from their peers, would be significantly higher than their
expectations at Time 1 and Time 2 and significantly higher than their standards at
Time 2.

7. Ideals: Participants' ideal outcomes would be significantly higher than both the
standards and expectations that they reported.


Results

Differences between standards and expectations

Recall that in Study 1, participants reported lower expectations than standards at

both Times 1 and 2. In Study 2, the items assessing standards and expectations were

counterbalanced at both Time 1 and Time 2 to eliminate order of the questionnaires as an

explanation for the differences in responses. Table 12 displays the difference between

standards and expectations when the items are counterbalanced. Contrary to our










hypothesis and the results of Study 1, in no case were participants' standards significantly

higher than their expectations. In fact, the generally finding was that expectations were

significantly higher than the standards. We revisit the issue of the differences between

standards and expectations in the discussion.

Comparing Expectations across Time

Did participants in the feedback condition lower their expectations of how they

would be evaluated in the moments prior to receiving feedback (Hypothesis 2)? The top

half of Table 13 displays the mean expectation at Time 1 and Time 2 for each of the Hyve

items and for the average across the Hyve items for the feedback condition. Consistent

with prior research on bracing (Shepperd et al., 1996), participants anticipating evaluative

feedback from their peers rated reported a significantly lower average expectation at

Time 2 than at Time 1. Analysis of each of the Hyve individual items revealed that

participants expecting feedback reported a significantly lower expectation at Time 2 than

at Time 1 for all items except for the item asking participants to estimate the number of

people that would Eind them interesting.

The bottom half of Table 13 displays the mean expectations at Time 1 and Time 2

for each of the Hyve items and the average across the Hyve items for the no feedback

condition. We hypothesized that the expectations of participants anticipating no feedback

would not differ from Time 1 to Time 2. Consistent with this prediction, analysis of the

average expectation across all Eive dimensions and all Hyve of the dimensions individually

yielded no significant differences between Time 1 and Time 2.










Comparing Standards across Time

Did participants who were anticipating feedback change their standard of what

constituted an acceptable outcome from Time 1 to Time 2 (Hypothesis 3)? The top half

of Table 14 presents the mean responses for all Hyve items and the average of the Hyve

items for the feedback condition. As predicted, averaging across the Hyve items,

participants expecting to receive evaluations from their peers reported a lower standard at

Time 2 than at Time 1. Analysis of the individual items revealed that participants

reported a significantly lower standard at Time 2 than at Time 1 for all items except for

the item asking how many people they wanted to Eind them friendly and how many

people they wanted to want to learn more about them.

The bottom half of Table 14 displays the mean standards at Time 1 and Time 2

for each of the Hyve items and the average across the Hyve items for the no feedback

condition. We hypothesized that standards of participants who anticipated no feedback

would not differ from Time 1 to Time 2. Consistent with our hypothesis, we found no

difference across time in the standards of no feedback participants except for the item

asking participants to estimate the number of people that would Eind them interesting.

The results of Study 2 provide further support that as feedback draws near, people

will alter their estimates for the future. As predicted, only participants in the feedback

condition changed their estimates and their standards from Time 1 to Time 2. An

unexpected pattern of means did appear in the data. The difference between the

conditions appears to occur at Time 1. We explore possible reasons for this finding in the

discussion.










Comparing Changing Standards to Bracing

Are standards and expectations the same thing, or are they different? As in Study

1, the correlations between standards and expectations at Time 1 and at Time 2 were

rather high (see Table 15). As noted earlier, the correlations provide a somewhat

misleading picture of the correspondence between standards and expectations. Recall

that participants' expectations differed from their standards (see Table 12). We next

compared the magnitude of change in standard from Time 1 to Time 2 with the

magnitude of change in expectation from Time 1 to Time 2. As evident in Table 16, in

no instance was the change in expectation significantly correlated with the change in

standards, suggesting that participants were not supplying the same response to the items

assessing standards and the items assessing expectations.

Changes in Standards and Changes in Expectations were uncorrelated in both

Studies 1 and 2. However, it can be argued that neither study had adequate power to test

the correlation, and that a larger sample size is necessary to uncover significant effects. It

is noteworthy, however, that the correlation between change in standard and change in

expectation is negative in three of five instances in Study 2. Nevertheless, to test the

possibility that the absence of a correlation between change in standard and change in

expectation was due to an inadequate sample size, we combined the data from Studies 1

and 2 and computed the correlations again. As evident in Table 17, combining the data

from Studies 1 and 2 had little effect. In only one instance (number of people who will

find you friendly) was the correlation significant and this correlation was notably small.

In sum, the extent to which participants changed their standards was uncorrelated with

the extent to which they changed their expectations, suggesting that the two constructs

are not identical.









An additional way to examine whether changes in standards and changes in

expectations represent the same construct is to compare how many participants displayed

the same amount of change in standards as change in expectations. Recall the dependent

variable was the number of people the participants expect or desire to find them friendly

or interesting, etc. For example, if a person lowered the standard for a good outcome

from 8 people to 6 people, would the expectation also drop 2 units? Table 18 presents

the number of participants out of 53 that changed their standard the same quantitative

amount as they changed their expectation. As is evident, for only a minority of the

participants was the change in standards identical to the change in expectations.

Anxiety Reports
We hypothesized that participants will report greater anxiety in the feedback

condition than the no feedback condition (Hypothesis 5). Although the means were in

the predicted direction, statistical analyses revealed no difference in anxiety for

participants in the feedback condition (M~= 3.23) and participants in the no feedback

condition (M~= 2.56), t (51) = 1.56, p < .13. Not surprisingly given that participants in

the two conditions did not differ in their anxiety reports, analyses revealed that anxiety

and the amount of change in standards and the amount of change in expectations were

largely uncorrelated (see Table 19).

Assessing the Accuracy of Participants' Standards and Expectations

As in Study 1, we compared participants' standards and expectations with the

actual evaluations they received from their peers. We predicted that the evaluations

participants receive from their peers would be significantly higher than their expectations

at Time 1 and Time 2 and significantly higher than their standards at Time 2 (see

Hypothesis 6). Tables 20 and 21 compare the evaluation participants received from their










peers with the standard participants reported as an acceptable outcome and the

expectations they reported. As we hypothesized (and consistent with Study 1),

participants' estimates at Time 1, and their estimates and standards at Time 2, were

significantly lower than the rating made by peers for all items except the item asking

participants whether they would be rated in the "top three" by their peers. Contrary to

our hypothesis, however, participants' standards at Time I were also significantly lower

than their peer evaluations.

Ideals

As noted earlier, it is possible that participants were not supplying a standard in

Study 1, but rather were supplying an "ideal" outcome; the outcome they would like to

occur in the best possible world. We argued that this seemed unlikely because we did not

get a ceiling effect in participants' standards in Study 1. Nevertheless, to test whether the

standard was distinct from the ideal outcome, we asked participants to indicate what they

regarded as the ideal outcome. We predicted that they ideal outcome what be

significantly higher than the standard participants reported (Hypothesis 7). Table 22

presents the standards and ideals at Times 1 and 2. As predicted across all five

dimensions at both Time 1 and Time 2, participants' ideals exceeded their standards.

Table 23 presents the expectations and ideals at Times 1 and 2. As predicted across all

Hyve dimensions at both Time 1 and Time 2, participants' ideals exceeded their

expectations. Ideals were always assessed after standards and expectations to ensure that

responding to the ideal items did not influence the ratings for standards or expectations.





















Mean Across All Items 4.7 5.3 2.48 .02
Time 2
Find you friendly 6.8 7.4 2.43 .02
Find you interesting 4.2 5.3 3.69 .001
Rank you in their top three 2.7 3.5 3.40 .002
Learn more about you 3.9 4.5 2.42 .02
Spend time with you in the future 3.4 4.2 4.42 .001
Mean Across All Items 4.2 5.0 3.98 .0002


Table 12: Study 2: Standards versus expectations at Time
and expectations are counterbalanced.
Time 1 Standard
Number of people you would like to... M
Find you friendly 6.9
Find you interesting 5.4
Rank you in their top three 3.1
Learn more about you 4.1
Spend time with you in the future 4.3


1 and at Time 2 when standards


Expectation
M
7.7
5.6
3.6
4.9
4.6


t
2.65
.70
1.79
3.65
1.27


p
.01
ns
.08
.001
ns





















Mean Across All Items 5.4 4.9 2.62 .02
No Feedback Condition Time 1 Time 2
Number of people you expect to... M M t p
Find you friendly 7.4 7.4 0 ns
Find you interesting 5.6 5.3 .86 ns
Rank you in their top three 3.3 3.6 1.37 ns
Learn more about you 4.9 4.8 .63 ns
Spend time with you in the future 4.7 4.4 1.14 ns
Mean Across All Items 5.2 5.1 .49 ns


Table 13: Study 2: Expectations at Time
Feedback Condition
Number of people you expect to...
Find you friendly
Find you interesting
Rank you in their top three
Learn more about you
Spend time with you in the future


1 and Time 2 by condition.
Time 1 Time 2
M M
8.0 7.6
5.6 5.3
3.8 3.3
4.9 4.2
4.7 4.0


P
.096
ns
.05
.007
.08


t
1.73
.71
2.05
2.96
1.85










Table 14: Study 2: Standards at Time 1 to Time 2 by condition.
Feedback Condition Time 1 Time 2
Number of people you would like to... M M t p
Find you friendly 7.3 6.8 1.19 ns
Find you interesting 5.8 4.4 3.19 .004
Rank you in their top three 3.4 2.9 1.79 .09
Learn more about you 4.2 3.8 1.22 ns
Spend time with you in the future 4.7 3.4 3.68 .002
Mean Across All Items 5.1 4.3 3.34 .003
No Feedback Condition Time 1 Time 2
Number of people you would like to... M M t p
Find you friendly 6.4 6.7 .68 ns
Find you interesting 4.8 4.0 2.88 .008
Rank you in their top three 2.9 2.5 1.09 ns
Learn more about you 3.9 4.0 .35 ns
Spend time with you in the future 3.9 3.4 1.55 ns
Mean Across All Items 4.4 4.1 .97 ns






46


Table 15: Study 2: Correlations between standards and expectations at Time 1 and at
Time 2.
Time 1 Time 2
Find You Friendly .51 .46
Find You Interesting .43 .51
Rate You in Their Top 3 .59 .64
Learn More About You .70 .66
Spend Time in Future .57 .62
Note: All correlations significant at p < .01.







47


Table 16: Study 2: Correlations between the change in standards and the change in
expectations.


Dimension
Find You Friendly
Find You Interesting
Rate You in Their Top 3
Learn More About You
Spend Time in Future
p <.05 *


Correlation
.05
-.10
.18
-.01
.11






48


Table 17: Correlations between the change in standards and the change in expectation
across both Study 1 and Study 2 (Total N=81).
Dimension Correlation
Find You Friendly *.23
Find You Interesting -.06
Rate You in Their Top 3 .03
Learn More About You .10
Spend Time in Future .20
p <.05 *






49


Table 18: Study 2: Number of participants changed their standards the same amount that
they change their expectations.
Changed Expectations and Standards the same amount
N %
Find You Friendly 9 17%
Find You Interesting 10 19%
Rate You in Their Top 3 19 36%
Learn More About You 16 30%
Spend Time in Future 17 32%
The numbers refer to the participants out of 53.










Table 19: Study 2: Correlations between anxiety and change in standards & anxiety and
change in expectations.
Change in Change in
Standard Expectation
Feedback Condition (M = 3.3)
Find You Friendly .03 -.06
Find You Interesting -. 11 -.27
Rate You in Their Top 3 .08 .20
Learn More About You -.04 .29
Spend Time in Future .31 .32
No Feedback Condition (M =2.5)
Find You Friendly **.49 .30
Find You Interesting .12 .06
Rate You in Their Top 3 .18 .11
Learn More About You .48 .10
Spend Time in Future .19 -.08
Overall (M =2.8)
Find You Friendly *.28 .14
Find You Interesting .00 -.13
Rate You in Their Top 3 .14 .22
Learn More About You .08 .24
Spend Time in Future *.29 .17
The means in parenthesis correspond to the average reported anxiety for the group.
p <.05 *
p <.01 **










Table 20: Study 2: Standard and received outcomes at Time 1 & Time 2.
Time 1 Received
Number of people you would like to... Standard Outcome
M M t p
Find you friendly 6.9 8.9 5.68 .0001
Find you interesting 5.4 7.4 5.84 .0001
Rank you in their top three 3.1 3.0 .41 ns
Learn more about you 4.1 6.4 6.88 .0001
Spend time with you in the future 4.3 6.1 4.54 .0001
Mean Across All Items 4.7 6.4 5.61 .0001
Time 2
Find you friendly 6.8 8.9 6.13 .0001
Find you interesting 4.2 7.4 8.72 .0001
Rank you in their top three 2.7 3.0 .75 ns
Learn more about you 3.9 6.4 6.75 .0001
Spend time with you in the future 3.4 6.1 7.04 .0001
Mean Across All Items 4.2 6.4 7.82 .0001










Table 21: Study 2: Mean expected and received outcomes at Time 1 & Time 2.
Time 1 Expected Received
Number of people you expect to... Outcome Outcome
M M t p
Find you friendly 7.7 8.9 4.32 .0001
Find you interesting 5.6 7.4 5.24 .0001
Rank you in their top three 3.6 3.0 1.73 .09
Learn more about you 4.9 6.4 4.35 .0001
Spend time with you in the future 4.6 6.1 3.93 .0003
Mean Across All Items 5.3 6.4 3.98 .0002
Time 2
Find you friendly 7.4 8.9 5.92 .0001
Find you interesting 5.3 7.4 6.61 .0001
Rank you in their top three 3.5 3.0 1.42 ns
Learn more about you 4.5 6.4 5.39 .0001
Spend time with you in the future 4.2 6.1 4.94 .0001
Mean Across All Items 5.0 6.4 5.48 .0001











Table 22: Study 2: Standard and ideal outcomes at Time
Time 1 Standard
Number of people you would like to... Outcome

Find you friendly 6.9
Find you interesting 5.4
Rank you in their top three 3.1
Learn more about you 4.1
Spend time with you in the future 4.3
Mean Across All Items 4.7
Time 2
Find you friendly 6.8
Find you interesting 4.2
Rank you in their top three 2.7
Learn more about you 3.9
Spend time with you in the future 3.4
Mean Across All Items 4.2


1 & Time 2.
Ideal
Outcome
M
8.8
8.1
6.4
7.2
6.8
7.5

8.6
7.1
5.5
6.4
6.1
6.7


t p


5.46
8.14
7.68
8.37
7.03
8.66

5.92
7.46
6.96
5.94
7.27
7.81


<.0001
<.0001
<.0001
<.0001
<.0001
<.0001

<.0001
<.0001
<.0001
<.0001
<.0001
<.0001










Table 23: Study 2: Mean expected and ideal outcomes at Time 1 & Time 2.
Time 1 Expected Ideal
Number of people you expect to... Outcome Outcome t p
M M
Find you friendly 7.7 8.8 3.74 .0005
Find you interesting 5.6 8.1 7.62 <.0001
Rank you in their top three 3.6 6.4 8.37 <.0001
Learn more about you 4.9 7.2 7.12 <.0001
Spend time with you in the future 4.6 6.8 6.62 <.0001
Mean Across All Items 5.3 7.5 8.65 <.0001
Time 2
Find you friendly 7.4 8.6 5.17 <.0001
Find you interesting 5.3 7.1 4.91 <.0001
Rank you in their top three 3.5 5.5 4.85 <.0001
Learn more about you 4.5 6.4 4.94 <.0001
Spend time with you in the future 4.2 6.1 5.35 <.0001
Mean Across All Items 5.0 6.7 5.59 <.0001















CHAPTER 4
GENERAL DISCUSSION

Often events do not come complete with obj ective indicators of what is and is not

an acceptable outcome. With a lack of objective anchors, people must personally define

their standard of what is acceptable. We proposed that people would adjust their

standards not just in response to new information, but also in response to the anticipation

of new information. We tested this latter possibility by leading participants to believe

they would soon learn how their peers evaluated them and them asking them to report

their standards across a variety of social dimensions. The results from two studies

revealed that people reported lower standards of what constituted acceptable social

outcomes when they expected feedback and did not adjust standards when they did not

expect feedback.

The results of the two studies also reveal several other important findings. First,

people lower their standards of what defines an acceptable outcome when they believe

they will soon receive outcome feedback. Yet, changing standards was not a chance

fluctuation; participants expecting no feedback did not change their standards. We

propose that lowering the definition of what constitutes an acceptable outcome represents

a means of avoiding unfavorable feedback.

Second, standards are not synonymous with expectations. In both studies

participants reported standards that were significantly different from their expectations.

Recall in Study 1, the participants completed items assessing standards prior to items

assessing expectations. In Study 1 participants reported standards that were significantly










higher than their expectations. In Study 2, when the standard and expectation items were

counterbalanced, the opposite pattern emerged, with participants reporting expectations

higher than standards. The contradictory pattern of differences between standards and

expectations deserves future exploration. A resolution to this may arise from treating the

estimates of standards and expectations as a between-subj ects variable.

Third, the results from both studies demonstrated that changing standards is

distinct from bracing for bad news. In Study 1, for four of the Hyve dimensions, the

change in standard was not correlated with the change in expectation. This result was

replicated in Study 2, which revealed that changes in standards and changes in

expectations were uncorrelated for all fiye dimensions. To address the possibility that low

statistical power was responsible for the nonsignifieant correlations, we re-examined the

correlations after combined the data from Studies 1 and 2. The combined analyses again

revealed the changes in standards were uncorrelated with changes in expectations.

Finally, we tabulated the amount of change for each participant on each dimension over

time. Specifically, for each dimension, we tabulated the number of units of change in

standards from Time 1 to Time 2 and the number of units of change in expectations from

Time 1 to Time 2. Only for a fraction of participants did the amount of change in

standards equal the amount of change in expectations. In sum, it appears that changing

standards is distinct from bracing for bad news.

Recent research depicts bracing as an attempt to avoid disappointment where

disappointment is specific negative emotional response to an outcome (Carroll, Dockery

& Shepperd, 2003). We argue that people change their standards for a different reason.

In the introduction we proposed a variety of reasons that people may proactively change










their standards. The present results reveal that the extent to which people brace is not

predictive of how they change their standards. We believe that, compared with changes

in expectations, people may change their standards for different reasons and with

different consequences. Although, in a broad sense, bracing and changing standards may

serve a self-enhancement function, a closer look may reveal differences in how these

processes alter mood or perceptions. Bracing alleviates disappointment about a future

outcome and as such, is an alteration of the likelihood of the occurrence of a negative

event. Changing standards alleviates the negativity of a future outcome and as such, is an

alteration of the perception of the outcome.

We believe that the consequences of changing standards and bracing are different.

Bracing speaks to a person' s evaluation of the subj ective likelihood of a certain outcome

(e.g., testing positive for a sexually transmitted disease or receiving a failing grade) and it

implies acceptance by the person that the undesired outcome is a possibility (i.e., they

may be a person who contracts a sexually transmitted disease or fails a course in college).

On the other hand, changing standards focuses on the evaluation of the outcome. To

change the evaluation of what determines an acceptable physical screening or grade in a

college course does not carry implications for the person or necessitate a change in the

person's self-concept. A task for future research is to explore the extent to which bracing

and changes in standards may have different purpose, result, and consequence for the

individual .

Fifth, in Study 2, the change in standards was often uncorrelated with reported

anxiety. Over both conditions together, there was a significant correlation between

change in standards and reported anxiety for two of the five dimensions. We expected









that participants' self-reported anxiety would positively and strongly correlate with the

change in their standards across all dimensions. Perhaps our measure of anxiety did not

fully tap into the feelings and emotions associated with changing standards. That is, some

other emotional or psychological process is responsible for the decline in standards

among participants anticipating imminent feedback. A second possibility is that the

presence of the other participants in the feedback condition caused the participants to

desire to proj ect an image of nonchalance at the prospect of receiving feedback. Thus,

because participants wanted to appear unruffled by the prospect of receiving feedback,

they reported lower anxiety.

A closer examination of Table 14 reveals an unexpected pattern of results.

Although we found that participants significantly lowered their standards in the feedback

condition but not in the no-feedback condition, participants appeared to differ in their

standards at Time 1 rather than Time 2. Further exploring of the data revealed that the

order of the expectations and standards items could not account for this finding. The

Time difference is troubling and may have occurred for one of two reasons. First,

participants were divided into the different rooms shortly before completing the Time 1

items. Thus, the design of Study 2 created a slight difference in when and where

participants completed Time 1. The moving of the no feedback participants to the other

room prior to Time 1 may have affected the ratings at Time 1. Replications that assess

Time 1 data prior to separation of participants can eliminate this effect. A second

possible explanation is, despite our attempts to make the conditions equal, there was an

error in random assignment. Future replications using this paradigm with increased

subject number and power may eliminate this explanation.









There are still unanswered questions about changing standards. Are there

individual differences in the extent to which people will change their standards? The

importance of the dimension for the person may affect the extent to which they change

their standards. Also, the necessity of receiving a specified outcome may moderate

changing standards. For example, at the end of the semester, to receive a passing grade

in a class, students have a grade they must receive or exceed on the final exam or

assignment. This need to obtain a specified grade may moderate their likelihood of

changing standards. Also, some people have received a wide range of outcomes both

experiencing great success and at times receiving failure feedback. For example, students

who have received passing and failing grades in their lifetime have a wide range of

experience with feedback on that dimension. Other people have received more stable,

consistent feedback on a dimension throughout their lives. For example, some students

who have always received A's with little variability or students who have always

received D's with little variability have little experience with a range of possible

outcomes. People with more experience with a range of outcomes may change their

standards more than people who have always received consistent feedback because they

have more flexibility in what they define as an acceptable outcome.

Another unanswered questions is how do statistically likely standards (having

three people rate you in their top three) or established standards (grading systems in most

school settings) influence the personal standards that people set? The findings from Study

1 suggest that people may give inadequate weight to base rate information when setting

their personal standards. Established standards, such as grading scales may provide a

starting point for the personal standards that people set. The grading criteria may vary










depending on the teacher, class or school district, but the student may still use this criteria

as a starting point for determining what is personally acceptable to them. For example,

students may establish a preliminary standard at the grade that is considered passing in a

given class or students may set their standard relative to the mean score of the class on a

test on a particular exam.

Another interesting questions is whether changes in standards represent a

conscious process or an automatic response? An examination of participants' reports

during the debriefing demonstrated the change in standards was conscious for some

participants but not for others. Some participants noted they changed their standards to

make themselves feel better about their feedback. Other participants were unaware that

they had changed their standards.

In addition, it would seem important to know how permanent are people's

standard changes? Shifts in expectations are characterized and momentary and in

response to environmental pressures. If changes in standards are not as directly related to

environmental pressures or are more driven by accuracy concerns, changes in standards

may be relatively permanent (Carroll, Dockery, & Shepperd, 2003). Still another question

alluded to earlier in the discussion is what are the consequences of changing standards

and how do the consequences of changing standards differ from the consequences of

bracing. Perhaps changing standards has less of an impact on a person's self-concept

because it is an alteration of the evaluation of the outcome. To change the evaluation of

what determines an acceptable outcome does not reflect on the person. A student who

redefines an acceptable grade in a class from an "A" to a "C+" may do so with little

change to the self-concept. A student who braces and expects to receive a C+ on an exam,










may be forced to adjust his or her view of being a good student to a more mediocre

student.

Another questions surrounding changing standards is how is the process of

changing standards affected by actual feedback? In the introduction we suggested people

change their standard in response to information or the anticipation of information. What

if the feedback people receive is discrepant from the feedback they defined acceptable?

Do people use the feedback they receive, even when it contradicts what they believe a

good outcome is?

The present studies suggest that, although people may brace for bad news to avoid

disappointment, they may also adjust what they consider an acceptable outcome before

receiving feedback. Changing standards may work hand in hand with bracing. Shepperd,

Ouellette and Fernandez (1996) note that although bracing may serve to avoid

disappointment it does not alter the implications of that feedback for the person.

Changing standards, making a previously unacceptable outcome acceptable, may be a

way that people can alter the implications of the feedback one the person' s self-concept.















APPENDIX
DEPENDENT VARIABLES

The items on this page ask you to provide a standarddddd~~~~~~dddddd for a variety of outcomes. A
standard is what you consider the minimally acceptable outcome. In other words, a
standard is the lowest outcome that you would find acceptable in a given situation. In
this experiment, we will ask you to report the minimum number of people (out of 10) you
would like to respond positively to each of the following questions.

For the following questions, think about how you would feel if one person evaluated you
positively and nine did not, or five people evaluated you positively and five did not, or
nine people evaluated you positively and one did not. With this in mind, what number
would you would be happy with? That is, what is the minimum number of people in this
study you would want to evaluate you positively on each of the dimensions below?

What is the minimum number of participants you would like to report that they... (circle
the single response that represents an acceptable number of participants)

want to spend time with you in the future?


0 1 23 45


found you interesting?


6 7 8 9 10 participants


0 12 34 5 6


7 8 9 10


participants


found you friendly?

0 1 23 45


6 7 8 9 10 participants


ranked you as one of the three people they liked best?


0 12 34 5 6


want to learn more about you?

0 12 34 5 6


7 8 9 10 participants


7 8 9 10 participants










The items on this page ask you to provide your expzectation for a variety of outcomes. An
expectation is the outcome that you predict would occur. In other words, an expectation is
the outcome that you think is likely to happen in a given situation. In this experiment, we
will ask you to report the number of people (out of 10) you expect to respond positively
to each of the following questions.

Now, think about what is likely to happen. How many participants do you predict will
report that they... (circle the single response that represents how many participants you
expect to...)



want to spend time with you in the future?


0 1 23 45


found you interesting?


6 7 8 9 10 participants


0 12 34 5 6


7 8 9 10


participants


found you friendly?

0 1 23 45


6 7 8 9 10 participants


ranked you as one of the three people they liked best?


0 12 34 5 6


want to learn more about you?

0 12 34 5 6


7 8 9 10 participants


7 8 9 10 participants









The items on this page ask you to provide your ideal for a variety of outcomes. An ideal
is the outcome highest outcome that you would want to occur. In other words, the ideal
outcome in a given situation. In this experiment, we will ask you to report the ideal
number of people (out of 10) you would want to respond positively to each of the
following questions.

Now, think about the ideal outcome. How many participants want to report that they...
(circle the single response that represents the ideal, or highest, number of participants you
would want to...)



want to spend time with you in the future?


0 1 23 45


found you interesting?


6 7 8 9 10 participants


0 12 34 56 7 8 910


participants


found you friendly?

0 1 23 45


6 7 8 9 10 participants


ranked you as one of the three people they liked best?


0 12 34 5 6


7 8 9 10 participants


want to learn more about you?

0 1 2 3 4 5 6 7 8 9 10 participants
















LIST OF REFERENCES


Alicke, M. D., Klotz, M. L., Breitenbecher, D. L., Yurak, T. J., & Vredenburg, D. S.
(1995). Personal contact, individuation, and the better-than-average effect. Journal
ofPersonality and' Social Psychology, 68, 804-825.

Bach, J. R., & Tilton, M. C. (1994). Life satisfaction and well-being measures in
ventilator assisted individuals with traumatic tetraplegia. Archives of Physical
Medicine and Rehabilitation, 75, 626-632.

Battle, E. S. (1968). Motivational determinants of academic competence. Journal of
Personality and' Social Psychology, 4, 634-642.

Battle, E. S. (1966). Motivational determinants of academic task persistence. Journal of
Personality and' Social Psychology, 2, 209-218.

Breetvelt I. S., & Van Dam, F. S. A. M. (1991). Underreporting by cancer patients: The
case of response-shift. Social Science and2~edicine, 32, 981-987.

Carroll, P., Dockery, K., & Shepperd, J. A. (2003) Exchanging optimism for pessimism
in personal predictions. Unpublished manuscript, University of Florida.

Dunning, D., Meyerowitz, J. A., & Holzberg, A. D. (1989). Ambiguity and self-
evaluation: The role of idiosyncratic trait definition in self-serving assessments of
ability. Journal ofPersonality and Social Psychology, 57, 1082-1090.

Festinger, L. (1957). A theory of cognitive dissonance. Stanford: Stanford University
Press.

Gilovich, T., & Medvec, V. H. (1995). The experience of regret: What, when, and why.
Psychological Review, 102, 379-395.

James, W. (1890). The principles ofpsychology (vd. 1). New York: Holt.

Locke, E. A. (1967). Relationship of success and expectation to affect on goal-seeking
tasks. Journal ofPersonality and Social Psychology, 7, 125-134

Pervin, L. A. (1963). The need to predict and control under conditions of threat. Journal
ofPersonality, 31, 570-587.










Sprangers, M. A. G., & Schwartz, C. E. (2000). Integrating response shift into health-
related quality-of-life research: A theoretical model. In C. E. Schwartz & M. A. G.
Sprangers (Eds.), Ad'aptation to changing health: Response shift in quality of hife
research (pp. 11-23). Washington D.C: American Psychological Association.

Shepperd, J. A., Ouellete, J. A., & Fernandez, J. K. (1996). Abandoning unrealistic
optimism: Performance estimates and the temporal proximity of self-relevant
feedback. Journal ofPersonality and Social Psychology, 70, 844-855.

Taylor, K. M., & Shepperd, J. A. (1998). Bracing for the worst: Severity, testing, and
feedback timing as moderators of the optimistic bias. Personality and' Social
Psychology Bulletin, 24, 915-925.

Zemke, R. (1981). Breaking the news to management. In J. Gordon, R. Zemke, P. Jones
(Eds.), The best of Training Ma'~gazine: Designing and' delivering~ cost-effective
training and' measuring the results (2nd ed.)(733-736). Minneapolis, MN:
Lakewood Publishing.
















BIOGRAPHICAL SKETCH

Meredith Terry was born June 26, 1979, in Honolulu, Hawaii, to Michael and

Debby Terry. Meredith spent most of early years living in Orange Park, Florida. She

graduated from the University of Florida with a bachelor' s degree in psychology in May

2001. She continued her education at the University of Florida where she completed a

Master of Science in psychology in May 2004.