GREGORY NORTHCRAFT SUSAN ASHFORD - …webuser.bus.umich.edu/sja/pdf/PreservationEverydayLife.pdftors-performance expectations, self-esteem, ... confident self-presentation (Goffman,

ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES 41, 42-64 (1990)

The Preservation of Self in Everyday Life: The Effects of Performance Expectations and Feedback Context on

Feedback Inquiry

GREGORY B. NORTHCRAFT

Department of Management and Policy, University of Arizona

AND

SUSAN J. ASHFORD

Amos Tuck School of Business Administration, Dartmouth College

This paper examines the roles of performance expectations, feedback context, and self-esteem in feedback inquiry. Seventy-eight subjects participated in a stock market simulation. Results showed that performance expectations and the public context in which the subjects sought feedback significantly influenced the frequency of feedback inquiry. However, these effects de- pended on type of feedback (personal performance vs social comparison) being requested and additionally were influenced by self-esteem of the feedback seeker. 0 1990 Academic Press, Inc.

Feedback on individual performance serves several useful functions in organizations. It cues individuals as to what the organization considers proper performance, motivates performance, and enables individual learning (Vroom, 1964). Much research has been directed toward understanding when and how feedback should be given to maximize its effect (see Ilgen, Fisher, & Taylor, 1979, for a review). Ashford and Cummings (1983) noted, however, that beyond those messages carefully provided by the organization, feedback also exists more generally in the environment and is available for individuals to use. They argued that any action or lack of action by others can be interpreted as feedback by individuals. Thus individuals can at any time use their observations of occurrences in the environment to infer a feedback message (Ashford & Cummings, 1983). Additionally, individuals can seek out feedback of their own volition by directly inquiring as to how others perceive and evaluate their behavior. Ashford and Cummings (1983) labeled these strategies “monitoring” and “inquiry,” respectively. This paper examines the impact of three fac-

Address reprint requests to Gregory B. Northcraft, Department of Management and Policy, University of Arizona, Tucson, AZ 8572 I.

42 0749-5978190 $3.00 Copyright 0 1990 by Academic Press, Inc. AU rights of reproduction in any form reserved.

INFLUENCES ON FEEDBACK SEEKING 43

tors-performance expectations, self-esteem, and feedback context-on individuals’ willingness to seek feedback by inquiry.

In a world where people tend to avoid directly providing informal feedback to others (Blumberg, 1972; Felson, 1980; Schoeneman, 1981) and where supervisors tend to distort negative feedback when they provide it in formal performance appraisal sessions (Fisher, 1979; Ilgen & Knowl- ton, 1980), feedback inquiry and monitoring constitute important means of obtaining information (Herold & Parsons, 1985). Feedback inquiry and monitoring strategies may be particularly important because individuals directly control their use. Ilgen and Moore (1987) found that allowing individuals choice over when they would use feedback significantly en- hanced task performance. With inquiry and monitoring, individuals control the use of feedback-they receive feedback when they want and need it, rather than on some organizationally determined schedule. Finally, while individuals can generate feedback by either monitoring or inquiry, Ashford (1989) notes that monitoring requires a great deal of inference and is likely to lead to more biased assessments than those obtained through inquiry. Thus, feedback inquiry is an important process to understand, given that others are frequently reluctant to convey feedback messages, that having choice over when to obtain and use feedback may enhance performance, and that the feedback obtained by this strategy may be less biased than that inferred from monitoring.

Given the importance of feedback inquiry, those personal and contextual factors that inhibit its use are of particular importance. Two previous studies have examined this issue. First, Walsh, Ashford, and Hill (1985) identified several environmental characteristics that contribute to what they termed an obstructed information environment-an environment in which individuals’ seeking efforts are thwarted. These characteristics include time lag between collection and delivery of feedback, supervisor unavailability, and lack of co-workers. Second, Fedor, Eder, & Buckley (1989) found that subordinates’ perceptions of their supervisors’ intentions (i.e., constructive/nonconstructive) in giving them feedback affected subsequent feedback seeking from this source. The current study concerns two psychological and social factors that may deter feedback inquiry: (1) the natural tendency individuals have to protect their egos, and (2) the self-presentation costs of admitting uncertainty and insecurity.

Performance Expectations and Znquiry

Feedback is qualitatively diierent from other types of information. It is evaluative information. It directly references the self and consequently is inherently affective in nature (Ashford & Cummings, 1983). Therefore, individuals are likely to be conflicted about feedback. While they should see it as valuable because it allows them to correct errors and attain goals,

44 NORTHCRAFT AND ASHFORD

they also will be motivated to maintain favorable views of themselves and thus might avoid feedback if they believe it will hurt their self-concepts (Brickman & Bulman, 1977; Jones & Gerard, 1967). This tension will particularly affect how individuals approach feedback they believe will be negative. On the one hand, negative feedback is more psychologically threatening than positive feedback. Thus, individuals may avoid negative feedback in order to maintain their self-esteem (Janis & Mann, 1977; Miller, 1976). On the other hand, because it points out where their behavior is “off track,” negative feedback is helpful to goal attainment (cf., Ashford & Cummings, 1985; Festinger, 1954). Anything that heightens ego-protection concerns should deter inquiry. Performance expectations may be just such a factor. Individuals expecting to perform poorly at a task should be more likely than those with high performance expectations to expect that feedback existing in the information environment also will be negative. These individuals should restrict their feedback search “so as to avoid exposure to disquieting messages” (Janis & Mann, 1977, p. 218).

Several studies have examined the hypothesis that low expectations lead to feedback avoidance. These studies have examined the effects of low expectations (induced by providing failure feedback) on the choice of subsequent tasks-those for which norms exist and those for which norms were unavailable (e.g., Meyer & Starke, 1982; Sachs, 1982; Trope, 1982; Zuckerman, Brown, Fox, Lathin, & Minasian, 1979)-or on the choice of items to include in a test that subjects would subsequently take (norm-available vs norm-unavailable items) (e.g., Carver, Antonio, & Scheier, 1985; Trope & Ben-Yair, 1982). In these task-choice and item- creation studies, feedback is discovered privately by working on the cho- sen task or by reviewing responses to the items generated. While in these studies low expectations have increased feedback seeking, the results may be of limited utility for scholars interested in feedback processes in organizations. Inquiry in organizational contexts is a social and therefore public process. It is doubtful that task-choice (private discovery) para- digm results will generalize to feedback inquiry in a social context where the act of seeking or receiving feedback may be quite public.

Brickman and Bulman (1977) captured a more social process in their study of the effects of low expectations on the desire for social comparison information. They found that individuals who received feedback indicating that they were “below average” on a symbol discrimination test expressed preferences not to exchange performance information with others. While this task was more social in nature, it suffered from a different weakness (true also of all task-choice research): it tapped subjects’ preferences to receive feedback rather than their behavioral attempts to seek it.

One prior study has examined reported feedback-seeking behavior in


the field. Ashford (1986) hypothesized that individuals would defend their egos by avoiding feedback and that people with negative beliefs about their performance would be particularly likely to engage in this defensive strategy. Surprisingly, Ashford (1986) found that individuals reporting low performance expectations also reported greater feedback-seeking through both inquiry and monitoring strategies. While Ashford (1986) interpreted this result as supportive of the instrumental value of feedback, the cross-sectional, single-source nature of her data allows alternative explanations. For instance, frequent seeking may give individuals information that leads them to be more pessimistic about their subsequent performance. Given that Ashford’s (1986) finding is inconsistent with a number of past studies (cited above), an examination of the impact of low expectations on inquiry for feedback in a controlled setting seems warranted. Thus, the first hypothesis to be tested in this research is

HI: Individuals with low performance expectations will engage in less feedback inquiry than those with high expectations.

This hypothesis and its test will add to past findings by focusing on feedback seeking from a social source rather than task choice, on actual seeking behaviors rather than preferences to receive feedback, and will examine these issues in a controlled setting.

Feedback Context and Inquiry

Beyond the desire to defend the ego by avoiding disquieting messages, there is a second deterrent to feedback inquiry-self-presentation. People typically alter their self-presentations based on what they believe others expect of them (Riordan, James, & Dunaway, 1985), what will maximize performance (Leary, Robertson, Barnes, & Miller, 1986), or what will procure them benefits (Baron, 1986). An important motivation in self- presentation is to avoid ridicule or loss of esteem in the eyes of various important audiences (Baumeister, 1982). The primary cost of seeking help or feedback is that inquiry exposes the seeker’s uncertainty and need for help or information (Gergen, 1974). Such exposure may be at odds with a need to maintain a self-assured, confident self-presentation (Goffman, 1956). Thus, asking for feedback, like asking for help of any kind, may be socially costly (cf., Broll, Gross, & Piliavin, 1974; Gergen, 1974; Nadler, 1986).

Self-presentation ought to be particularly important to feedback inquiry because anonymity is impossible in inquiry. At least one person-the feedback provider-becomes aware of the seeker’s “need to know” when feedback is requested. The costs of public inquiry should be even higher when, in addition to the feedback provider, others observe the seeking act. Public inquiry easily could be interpreted by observers as a


sign of insecurity (Schoeneman, 1981). To the extent that individuals recognize this interpretation, they should avoid public inquiry. Though Ashford (1986) found no direct support for this contention in her field research, she did find that longer tenured employees sought feedback less, even though they considered feedback as valuable as did their less tenured colleagues. Ashford (1986) interpreted this result as reflecting the inhibitory self-presentational pressures faced by more senior employees. These employees probably believed that seeking feedback would under- mine their standings as confident and self-assured veterans, and accord- ingly sought less feedback. No previous studies have examined the effects of a public seeking context on feedback inquiry in a controlled setting.

This study explored two aspects of inquiry publicness: (a) the publicness of the seeker’s request, and (b) the publicness of feedback delivery. Hypothesis 2 follows the argument stated above:

H2: Individuals will reduce feedback inquiry when others can observe their seeking requests.

Thus, the most inquiry is expected when feedback requests can be made privately. The predictions regarding the effects of public receipt of feedback are more complicated. An interaction between performance expectations and public feedback receipt is anticipated. Specifically,

H3: Individuals with low performance expectations will reduce inquiry when feedback delivery is public. Individuals with high performance expectations will not decrease seeking behavior when feedback delivery is public.

The logic of Hypothesis 3 is similar to that underlying Hypothesis 1. However, instead of curtailing seeking to protect their egos from direct harm, Hypothesis 3 suggests that individuals reduce inquiry to protect their egos from indirect harm-the public embarrassment caused by having others hear of their poor performance. When performance expectations are low, this concern will inhibit feedback inquiry.

Individual Differences and Inquiry

The underlying theoretical justification for this research is that feedback and feedback inquiry may be threatening to an individual’s self- concept and social persona. Potentially negative feedback, however, may not be equally threatening to all people. Individual characteristics or personality factors that reduce or enhance the threat of negative feedback also should influence feedback seeking. Self-esteem may be one such factor. Individuals with high self-esteem may not see negative feedback as a threat to their egos (e.g., McFarlin & Blascovitch, 1981). Rather, they may feel that this information is instrumentally valuable and that they can


use it to alter their behavior. In contrast, low self-esteem individuals may feel particularly threatened by potentially negative feedback (Bergles & Jones, 1978; Kolditz & Arkin, 1981). While much research has focused on how high and low self-esteem individuals react to feedback (McFarlin & Blascovitch, 1981; Shrauger & Lund, 1975; Swann, Griffin, Predmore, & Gaines, 1987), no empirical work has addressed how these two groups might differ in their willingness to seek out feedback. Brickman and Bul- man (1977, p. 160) do note that

. . . if people are for any reason self-conscious, anxious, or just sensitive about their own position in a group, we might expect them to be especially interested in avoiding potentially unfavorable comparisons.

Potentially negative feedback may make low self-esteem individuals anxious for three reasons. First, as argued above, low self-esteem individuals have weaker egos to begin with and, therefore, potentially negative feedback may seem more threatening than it might be to a high self-esteem individual. Second, low self-esteem individuals have fewer alternative defense mechanisms available to them for coping with the negative message; therefore, avoidance is an attractive strategy. High self-esteem individuals are much more likely to cope with negative messages by labeling the rater as incompetent or disparaging the accuracy of the feedback than are their low self-esteem counterparts (Swann et al., 1987). Finally, negative feedback may have a more severe effect on subsequent efficacy expectations for low rather than high self-esteem individuals. Shrauger & Sorman (1977), for example, found that the persistence and performance of low self-esteem individuals were severely affected by the receipt of negative feedback, while that of high self-esteem individuals were not. Low esteem individuals may feel that they cannot improve their behavior on the basis of negative feedback. Given its likely costs in terms of reduced efficacy expectations, these individuals should view feedback as more potentially threatening. Thus,

H4: When performance expectations are low, low self-esteem individuals will reduce inquiry.

A final hypothesis was tested in this study, primarily to rule out an alternative explanation of feedback inquiry behaviors. It could be argued that the valence of obtained feedback (negative or positive) will influence an individual’s subsequent seeking behavior, irrespective of ego defense or self-presentation concerns. If feedback is positive, inquiry might increase simply because behavior that is reinforced tends to be repeated (Skinner, 1953). By the same logic, if the feedback obtained is negative, subsequent seeking should decrease. Thus, Hypothesis 5 states that

H5: The valence of received feedback should influence subsequent


feedback seeking; individuals receiving positive feedback early should be more likely to seek later, and individuals receiving negative feedback early should be less likely to seek later.

These five hypotheses were tested in the context of a controlled laboratory experiment in which subjects could request and receive performance feedback. An additional, exploratory focus of this study was the dynamics of feedback seeking across different types of feedback. Theo- rists have identified at least two types of feedback: direct information about individuals and their performance, and information that allows individuals to compare their performance with the performance of others (Jones & Gerard, 1967). Individual outcome feedback is not always useful for task learning (cf., Hoffman, Earle, & Slavic, 1981; Schmitt, Coyle, & Saari, 1977). Particularly in competitive situations, individuals may need comparative feedback in order to calibrate the value of their own accom- plishments-to understand where their performance falls relative to the performance of others. Given the value of these two very different types of feedback, it seems important to examine whether the dynamics of feedback seeking are the same for both. No a priori hypotheses regarding feedback type were proposed.

METHOD

Subjects

Seventy-eight undergraduates in the College of Business at the Univer- sity of Arizona participated as subjects. Each subject received credit for participating in partial fulfillment of a course requirement. Subjects were run in small groups (four to seven subjects per group). Groups were randomly assigned to experimental conditions.

Design Overview

Each subject assumed the role of investment portfolio manager and managed a stock portfolio (starting with $100,000 in cash) through a four- period stock market simulation. There were five stocks in the market and each stock began the exercise valued at $100 per share. During each period of the exercise, the price of each stock either fell or rose $10 per share. At the beginning of each period of the exercise, subjects received a computer printout summarizing stock recommendations from five different brokerage houses to buy or sell each of the five stocks. Subjects completed “transaction request forms” during each period of the exercise indicating how many shares of each stock the subject wished to buy or sell. The movement of stock prices was a random sequence determined in advance by the computer and was the same for all subjects. Each bro-


kerage house perfectly predicted the movement of two of the five stocks and was randomly correct for the other three stocks.

To assist subjects’ deliberations, feedback was available by request. Subjects could request either portfolio feedback, social comparison feedback, or both. Portfolio feedback summarized the current contents of an individual’s personal portfolio including the number of shares held of each stock, value of shares held, available cash reserves, and total cash- equivalent portfolio value. Social comparison feedback did not provide direct outcome information. Instead, social comparison feedback provided a comparison of an individual’s portfolio with other participants in the exercise, including rank of the subject’s portfolio in cash-equivalent value (e.g., “3rd”), and size and most recent change in value of the subject’s portfolio, relative to those of other participants in the exercise (e.g., “Your portfolio has $1200 less than the largest portfolio; last period your portfolio increased by $200 less than the largest increase.“)

There were two independent variables. The first independent variable manipulated the performance expectations of subjects. Prior to beginning the exercise, each subject completed one page of “Brokerage Aptitude Exam sample questions.” The page of questions appeared to be an ex- cerpt from a longer test. There were two versions of the page of questions: Low Expectations and High Expectations. In the Low Expecta- tions condition, the questions were statistically sophisticated (e.g., joint probability calculations, confidence intervals) and had been pilot tested to appear familiar but be virtually impossible for this subject population. In the High Expectations condition, the questions were straightforward (e.g., simple arithmetic) and had been pilot tested to be familiar and easy for this subject population. All subjects also received answer keys to the questions against which to check their own attempts.

A second independent variable manipulated Feedback Context. Feed- back context refers to the publicness of subjects’ (1) requests for and (2) delivery of feedback. Requests for both types of feedback were made either publicly or privately. In the Private conditions, there were two blanks corresponding to portfolio and social comparison feedback on each subject’s “transaction request form.” To request feedback, the subject needed only to check one or both of the blanks corresponding to the type of feedback desired. In the Public conditions, request slips were made available to subjects that were 3 by 4 in. and blue (social comparison feedback) or pink (portfolio feedback) with v2 in. letters that said “Social Feedback” or “Portfolio Feedback.” These request slips had to be paper-clipped to the “transaction request form” if feedback was desired and were clearly visible to the other participants.

Delivery of feedback also was either Private or Public. In the Private conditions, requested feedback was included with the computer printout


of brokerage house recommendations received by every subject at the beginning of every period of the exercise. In the Public conditions, requested feedback was written into a matrix on a large blackboard clearly visible to all subjects. If portfolio feedback was requested, the portfolio total cash-equivalent value was written into the matrix; if social comparison feedback was requested, the current rank of the subject’s portfolio and its total value relative to other portfolios both were written into the matrix. In addition, subjects in the Public delivery condition also received the complete versions of portfolio and social comparison feedback on their computer printouts.

Thus, the experimental design was a 2 x 3 factorial, with two levels of Performance Expectations (Low, High) and three levels of Feedback Context (Private Request/Private Delivery; Public Request/Private Deliv- ery; Public Request/Public Delivery). A fourth possible Feedback Con- text condition-private request/public delivery-was omitted from the design; because public delivery de facto made requests public, there was no private request/public delivery condition.

Procedures

Subjects signed up to participate in groups of six. Upon arrival, subjects were ushered into a small conference room which contained a conference table, ten chairs, and a large blackboard. Through a window subjects could see a computer on a desk in an outer office. Subjects seated themselves around the table and the experimenter made some brief remarks explaining the general nature of the experiment. In particular, subjects were told that this was an experiment about decision-making and that they would be making some decisions in the context of a stock market simulation. To stimulate interest, subjects were informed that two gift certificates for dinners at a local restaurant would be awarded for the best overall portfolio performance and for the best single-period portfolio performance in their group during the simulation. Subjects then were each given an instruction booklet to read through. A small placard placed in front of each subject publicly identified the subject’s account number for the remainder of the exercise.

The instruction booklet contained a general introduction to the process of buying and selling stocks, types of feedback available and how to request them, and two short questionnaires. The first questionnaire as- sessed the self-esteem of each subject. The second questionnaire was the Performance Expectation manipulation described above. In the opening remarks, subjects had been told that their responses to these questionnaires would help the experimenter understand what factors predicted performance on the experimental task.

The first questionnaire was described to subjects as testing some per-


sonal characteristics that might influence performance. This first questionnaire contained the revised Jan&Field Feelings-of-Inadequacy (self- esteem) scale, which was developed by Eagley (1%7) and has been used extensively in research (e.g., Brockner & Swap, 1983; Brockner, Davy, & Carter, 1985). The obtained sample mean in this study (69.15) is consistent with previous research using business students (Brockner, personal communication). For analysis purposes, subjects’ self-esteem scores were dichotomized via a median split. The second questionnaire was the “Brokerage Aptitude Exam sample questions,” which the experimenter noted might predict performance on the experimental task. A mock graph had been drawn on the blackboard suggesting a strong positive relationship between number of questions correctly answered on the second questionnaire and performance on the experimental task. The experimenter gestured to this graph when describing the second questionnaire. Subjects also were informed that the answer key for the second questionnaire had been attached because previous subjects had found it interesting to check their own answers against the key.

When subjects had read through the instructions and completed both questionnaires, the experimenter quickly summarized the instructions and fielded questions concerning the logistics of the experimental task. In order to prevent “end-play” effects, subjects were not told how many periods the exercise would last. The experimenter then distributed “transaction request forms” and first-period brokerage house stock recommendations. For the Public Request conditions, piles of feedback request slips and paper clips were put in the center of the table. For the Public Delivery conditions, a large matrix was drawn on the blackboard in which to record publicly requested feedback. Subjects were identified in the matrix by account number. Prior to starting the exercise, each subject was asked to turn over the second questionnaire and record an estimate of expected performance using a l-to-7 scale (where 7 = “I will do very well”). Subjects then were told to fill out their first period “transaction request form.”

When subjects completed their “transaction request forms,” (including attachment of feedback request slips in the Public Request conditions), the experimenter collected the forms, exited the conference room, and entered the subjects’ transaction requests on the computer within sight of the subjects. The computer produced a print-out for each subject which presented the next period’s brokerage house recommendations and any feedback requested by the subject. The experimenter then returned to the conference room, recorded updated stock prices on the blackboard, and delivered the printouts to begin the next period of the exercise. If the subjects were in a Public Delivery condition, the experimenter also wrote any requested feedback into the matrix on the blackboard.


At the end of the fourth period, the simulation was halted. Subjects were given a follow-up questionnaire to complete that contained several manipulation checks. Final portfolio performance figures were provided for subjects, and the two gift certificates were awarded to participants in each group. Finally, subjects were debriefed, thanked for their participation, and dismissed.

RESULTS

Manipulation Checks

The manipulation checks confirmed the effectiveness of the experimental manipulations. First, subjects in the High Expectation conditions reported doing significantly better on the “Brokerage Aptitude Exam sample questions” than subjects in the Low Expectations conditions (5.22 vs 1.86, F(1,77) = 79.6, p < .OOl). Analysis of variance also revealed that subjects’ reported expectations were significantly higher in the High Ex- pectations than in the Low Expectations conditions (5.00 vs 3.83, F( 1,74) = 15.56, p < .OOl), but were unrelated to subjects’ self-esteem scores (p > .lO). Subjects in the Private Delivery conditions reported their feedback to be significantly more private than subjects reported in the Public Delivery conditions (4.76 vs 3.15 and 3.31, F(2,76) = 5.83, p < .Ol). Subjects in the Private Request conditions reported their feedback requests to be significantly more private than subjects reported in the Public Request conditions (3.41 vs 4.88 and 4.69, F(2,76) = 4.09, p < .05).

The experimental groups did not differ significantly in their ratings of the usefulness of feedback received or the importance of feedback about performance on the experimental task. It is worth noting that the overall mean “importance of feedback about performance” rating was 5.87 on a 7-point scale. Subjects in the High Expectations conditions reported their participation in the experiment to be significantly more enjoyable than subjects in the Low Expectations conditions (2.16 vs 2.85, F(1,77) = 4.57, p < .05).

Experimental Hypotheses

The mean number of Portfolio and Social Comparison feedback requests by subject (up to four requests possible) are displayed by condition and period in Tables 1 and 2. A preliminary analysis of variance with Feedback Type (two levels, Portfolio and Social Comparison) as a repeated measure was run to see if there were significant differences in the amounts of Portfolio and Social Comparison feedback requested, and to see if the pattern of effects across the two types of feedback warranted separate analyses. This preliminary analysis revealed significant differences in the amounts of Portfolio and Social Comparison feedback re-


TABLE 1 MEAN NUMBER OF PORTFOLIO FEEDBACK REQUESTS BY CONDITION

Feedback condition

Hiih expectations High Self-esteem

Period 1 2 3 4

Low Self-esteem Period 1

2 3 4

Low expectations High Self-esteem

Period 1 2 3 4


2 3 4

Private/private Public/private Public/private

4.00 3.33 4.00 1.00 X.7 1.00 1.00 .83 1.00 1.00 1.00 1.00 1.00 .83 1.00

3.50 3.00 3.80 SO .33 1.00

1.00 .89 1.00 1.00 .89 .80 1.00 .89 1.00

3.67 3.40 2.40 .78 .80 .oo .89 .80 ho

1.00 .80 1.00 1.00 1.00 .80

3.83 3.60 1.20 .83 .80 .oo

1.00 .80 .20 1.00 1.00 A0 1.00 1.00 ho

quested (subjects requested significantly more Portfolio than Social Com- parison feedback, F(1,66) = 47.41, p < .OOl), and significant effects across both Feedback Types for Feedback Context (F(2,66) = 25.81, p < .OOl), Performance Expectations (F(1,66) = 5.24, p = .025), Feedback Context by Self Esteem (F(2,66) = 3.49, p = .036), and Feedback Con- text by Performance Expectations (F(2,66) = 6.10, p = .004). This preliminary analysis also revealed a significant Feedback Context by Feed- back Type interaction (F(2,66) = 4.27, p = .018), a significant Feedback Context by Performance Expectations by Feedback Type interaction (F(2,66) = 3.63, p = .032), and a significant Feedback Context by Per- formance Expectations by Self Esteem by Feedback Type interaction (F(2,66) = 5.61, p = .006). Due to the presence of these significant Feedback Type interactions, subsequent analyses examined Portfolio and Social Comparison feedback requests separately. Specific tests of the experimental hypotheses follow below.

Hypothesis 1 proposed that low performance expectations would inhibit inquiry. As shown in Table 3, Hypothesis 1 was supported for Port- folio feedback. Low Expectations subjects requested significantly less


TABLE 2 MEAN NUMBER OF SOCIAL COMPARISON FEEDBACK REQUESTS BY CONDITION

Feedback condition

Private/private Public/private Public/private

High Expectations High Self-esteem

Period 1 2 3 4


2 3 4

Low Expectations High Self-esteem

Period 1 2 3 4


2 3 4

2.63 2.83 1.83 SO .33 .50 .63 .83 .50 .63 .67 .33 .88 1.00 SO

3.25 2.67 .80 .50 .22 .20 .75 .78 .20

1.00 .67 .20 1.00 1.00 .20

3.67 1.40 1.00 .67 .20 .20

1.00 .40 .oo 1.00 .40 .40 1.00 A0 .40

2.17 3.60 .40 .50 .60 .oo .50 1.00 .20 .50 1.00 .I0 .67 1.00 .lO

Portfolio feedback than did High Expectations subjects (F( 1,66) = 12.67, p < .OOl). Any interpretation of this effect should be tempered by the significant interaction found between Performance Expectations and Feedback Context. (As described below, feedback seeking was extremely low when both performance expectations were low and feedback delivery was public.) As shown in Table 4, Hypothesis 1 was not supported for Social Comparison feedback.

Hypothesis 2 argued that making feedback requests public would inhibit inquiry. Hypothesis 2 was tested by contrasting feedback requests in the Private-Request/Private-Delivery and Public-Request/Private-De- livery conditions. Hypothesis 2 was only marginally supported for Port- folio feedback. For all four mean comparisons (High and Low Expecta- tions by High and Low Self-esteem) subjects requested less feedback when requests were Public than when requests were Private. Thus, the pattern of means for all cells was consistent with the hypothesis. Analysis of variance revealed this effect to be only marginally significant, however (F(1,66) = 4.30, p < .lO). Hypothesis 2 was not supported for Social Comparison feedback.


TABLE 3 ANALYSIS OF VARIANCE: PORTFOLIO FEEDBACK

Between subjects effects Feedback context Performance expectations Self-esteem Feedback context by performance

expectations Other effects Error

Within subjects effects Trials Feedback context by performance

expectations by trials Other effects Error

SS df MS

4.81 2 2.41 2.07 1 2.07 1.16 1 1.16

5.42 2 2.71 1.57 5 .31

10.79 66 .16

4.16 3 1.39

2.27 6 .38 2.47 27 39

14.35 198 .07

F P

14.71 <.ool 12.67 <.OOl 7.11 .Ol

16.58 C.001 l.% ns

19.15 <.OOl

5.23 C.001 1.30 ns

Hypothesis 3 proposed that public feedback delivery would inhibit inquiry only if performance expectations were low. Hypothesis 3 was tested by contrasting feedback requests in the Public-Request/Private-Delivery and Public-Request/Public-Delivery conditions. The predicted interaction between Performance Expectations and Public Delivery was significant for Portfolio feedback (F( 1,66) = 27.45, p < .OOl). For Social Comparison feedback, however, the predicted interaction was not significant. Instead Public Delivery (when requests were public) significantly decreased in-

TABLE 4 ANALYSIS OF VARIANCE: SOCIAL COMPARISON FEEDBACK

Between subjects effects Feedback context Performance expectations Self-esteem Feedback context by self-esteem

by performance expectations Other effects Error

Within subjects effects Trials Feedback context by trials Other effects Error

ss df MS

15.73 2 7.86 .26 1 .26 .26 1 .26

3.82 2 1.91 .64 3 .21

27.68 66 .42

4.22 3 1.41 1.77 6 .30 2.14 27 .08

18.37 198 49

F P

18.75 1.001 .62 ns .62 ns

4.56 .014 .50 ns

15.15 <.OOl 3.19 .005

.88 ns


quit-y for both High and Low Expectations subjects (F( 1,66) = 15.93, p < .Ol).

Hypothesis 4 proposed that Low Self-esteem subjects would be particularly likely to reduce inquiry when performance expectations were low. Hypothesis 4 was not supported. Rather than the expected interaction, self-esteem demonstrated a significant main effect on Portfolio feedback inquiry. As shown in Fig. 1, High Self-esteem subjects requested significantly more Portfolio feedback than did Low Self-esteem subjects (3.54 vs 2.90 requests, respectively; F(1,66) = 7.11, p < .Ol). In contrast, as shown in Fig. 2, for Social Comparison feedback there was a significant 3-way interaction among Feedback Context, Performance Expectations, and Self-esteem (F(2,66) = 3.82, p = .014). Self-esteem influenced the inquiry of Low Expectations subjects when requests were Public and

Requests

/

Reqwis Private

Delivery Private

Public

Private

Public

Public

Feedback Context

(b)

Number of Feedback Requests

4.0

1

em

LOW Self-Esteem

t , , , l&q”&3 Private Publlc Public

Dell”erlJ Private Prl”*te Public

Feedback Context

FIG. 1. Mean number of portfolio feedback requests by period for (a) high and (b) low self-esteem subjects.


(a)


High Self-Esteem

Requests Delivery

Private Private

Public Private

Public Public

Feedback Context

(b)


4.0 -

Low Self-Esteem

I I I I

Requests Private Public Public Delivery Private Private Public

Feedback Context

FIG. 2. Mean number of social comparison feedback requests by period for (a) high and (b) low self-esteem subjects.

delivery was Private; Low Self-esteem subjects requested significantly more feedback then their High Self-esteem counterparts in this condition (t = 2.68,~ < .OS).

Two repeated-measures analyses explored possible changes in the effects of the independent variables on feedback inquiry across the four periods of simulation. For Portfolio feedback, there was a significant increase in inquiry across trials (61.5% of subjects requesting feedback in Period 1 vs 91 .O% of subjects requesting feedback in Period 4, F(3,198) = 19.15, p < .OOl). A significant Feedback Context by Performance Expec- tations by Trials interaction also was found for Portfolio feedback (F(6,198) = 5.23, p < .OOl); inspection of the means suggests that inquiry in the Low-Expectations/Public-Request/Public-Delivery condition in-


creased more over trials than in any other condition. For Social Compar- ison feedback, inquiry also increased significantly over trials (from 35.9 to 67.9% of subjects, F(3,198) = 15.15, p < .OOl). A significant Feedback Context by Trials interaction also was found for Social Comparison feedback (F(6,198) = 3.19, p = .005); inquiry increased less over trials in the Public-Request/Private-Delivery condition than in other conditions.

Hypothesis 5 offered an alternative explanation for any effects found in testing the first four hypotheses. Consistent with the theoretical position of this paper, Hypothesis 5 was not supported. To test Hypothesis 5, each feedback message received by each subject was coded as either punish- ment (if portfolio value had decreased) or reinforcement (if portfolio value had increased or remain unchanged). Probabilities were calculated to contrast inquiry likelihoods for subjects who had just been punished and reinforced. No significant differences in subsequent inquiry likelihood were found between reinforced and punished subjects for either Portfolio or Social Comparison feedback.

DISCUSSION

This study examined the influence of several factors-performance expectations, feedback context, and self-esteem--on an individual’s willingness to request personal performance and social comparison feedback. Each of these factors affected feedback inquiry, even though subjects across experimental conditions reported no differences in perceived usefulness of or desire to acquire performance feedback. The picture that has emerged from this study is a complex one, however. Personal performance and social comparison feedback inquiry apparently obey quite different dynamics. Further, while self-esteem significantly influences inquiry, the precise nature of its effect is not clear.

For personal performance feedback (portfolio feedback in this study), the effects of performance expectations and feedback context are straightforward. In support of the ego defense argument, individuals engage in less inquiry when performance expectations are low, or when self-esteem is low, even when feedback is delivered privately. In line with the impression management argument, public delivery of feedback inhib- its feedback inquiry, particularly when performance expectations are low. These results are consistent with the contention that feedback inquiry is affected by both self-presentation and ego-defense concerns. Individuals forego inquiry if they believe the feedback will be negative, and especially if others will hear it.

The pattern of seeking behavior appears quite different for social comparison feedback-feedback that compares an individual’s performance with that of others. The most striking characteristic of social comparison


feedback inquiry in this study was the absence of an interaction between performance expectations and feedback seeking context. Performance expectations had no measurable effect on the amount of social comparison feedback inquiry. Public feedback delivery altered (specifically, decreased) social comparison feedback inquiry. The effects of self-esteem on social comparison inquiry found in this study are perplexing. In particular, public delivery of social comparison feedback appears to have strikingly different effects on inquiry depending on the self-esteem and performance expectations of the feedback seeker.

One possible explanation for the differences found between personal performance and social comparison feedback is that personal performance feedback requests are seen as more legitimate than social comparison feedback requests. After all, the desire for personal performance feedback is likely to be seen as an appropriate attempt at self- improvement. Social comparison feedback requests, on the other hand, may be perceived as crass, competitive attempts at one-upsmanship. If this view of social comparison feedback is commonly held, then the self- presentational pressures are clear: One should request less social comparison feedback to avoid negative reactions from observers. However, this study found that public delivery of feedback reduced social comparison feedback seeking rather than public requests. Perhaps the simple act of turning in a blue-colored card to request social comparison feedback was not sufficiently and publicly vivid to deter inquiry. However, when the social comparison feedback was posted, the competitive motivation of the requesting subject would have been clear. Recognizing this may have led subjects to reduce their social comparison feedback-seeking in the public delivery condition. Future research will need to explore further the differences in motives for personal and social comparison feedback inquiry .

These results have several practical implications for the design of feedback delivery systems in organizations. First, as shown in previous research, performance expectations play a role in the feedback process. In this study, low expectations inhibited inquiry for personal performance feedback. This finding contradicts Ashford’s (1986) results from a corre- lational field study. Unfortunately, low performance expectations-the very circumstances in which feedback probably would be most valuable-seem likely to decrease workers’ willingness to ask for information that could improve performance.

Given the reluctance of supervisors to give negative feedback (Fisher, 1979; Ilgen & Knowlton, 1980), a picture develops of those subordinates fearing poor performance not requesting the information they need to improve. Since feedback is important to performance, the absence of


feedback for these individuals just might guarantee that they remain in the very condition they fear-that of continued poor performance. This study also has shown feedback context (specifically, the publicness of feedback requests and delivery) to be an important factor affecting feedback inquiry. If managers wish to encourage subordinates to solicit feedback in order to improve their performance, the influence of contextual factors needs to be considered. In the tield study mentioned previously in which low performers sought more feedback than high performers (Ashford, 1986), the culture of the organization may have supported feedback inquiry, regardless of performance expectations. The design of feedback systems may have to include norms and incentives to promote poor performers’ feedback seeking. At the very least, providing some means by which workers can acquire and digest feedback in private would seem to be an effective way to increase the probability that performance feedback will be sought and utilized.

While this study suggests that individuals are sensitive to the ego and self-presentational costs of feedback inquiry, more research is needed to document the dynamics influencing inquiry behavior. This study, for example, may have been a relatively conservative test of the hypotheses due to the nature of the task employed. In this stock market simulation, there is a very clear, objectively determined bottom-line indicator of performance. In this situation observers’ impressions of an individual’s self- confidence, strength, and so forth may not matter as much as they would in a task where peers had an input into performance ratings, or simply where the definition of good performance was more subjective and eval- uators might be influenced by such factors as how an individual’s peers seemed to respond to him or her. The fact that this study found any effects for the public request and delivery of feedback suggests that these effects are robust ones.

Two issues remain unaddressed in this study. First is the possible com- plementary nature of inquiry and monitoring strategies. It may be that those factors (such as feedback delivery publicness) which are likely to impede inquiry will lead to increased monitoring. After all, monitoring is not subject to impression-management concerns and, therefore, makes a suitable (if less trustworthy) substitute when inquiry becomes problem- atic.

A second unresolved issue concerns the dual content (task and performance information) of most feedback messages. In this task, the personal portfolio feedback requested by subjects included both information about the task (i.e., how many shares of each stock in the portfolio) and about the subject’s performance (i.e., did the portfolio’s value go up or down?). This suggests that multiple motives could be driving interests in available


“feedback.” Future research in this arena should explore whether pure information seeking-which does not reference the self-is influenced by the same ego-defense and impression-management concerns as feedback inquiry and help seeking.

Future research also should further explore feedback seeking contexts. In the current study, even “private” inquiry was public to at least one person-the feedback source or “target.” Technological advances are making it possible for workers to get feedback in the absence of even a target, for instance from a machine. One study (Northcraft & Earley, 1989) has demonstrated the importance of such “absolutely private” sources in promoting feedback use; to date no studies have examined what factors determine whether targets are treated more like audiences and when targets are treated more like “absolutely private” sources of feedback. For instance, if the target also is the seeker’s performance evaluator, are heightened impression-management concerns likely to di- minish inquiry in favor of increased monitoring? Finally, future research should attempt to directly measure subjects’ perceptions of the riskiness of requesting feedback.

In conclusion, this study does suggest that if researchers are to better understand self-generated feedback processes as Herold and Parsons (1985) suggest, then they must understand the world as individual performers see it. That world appears to be one of competing benefits and risks in which openly seeking feedback has costs in terms of the image performers may want to convey and the beliefs they hope to maintain about themselves. Steps that organizations can take to minimize these perceived costs should produce more direct inquiries for feedback. In this regard, research efforts are needed to document the effects of group or organizational norms regarding feedback seeking. Norms validating inquiry for feedback should substantially reduce the perceived cost of inquiry. Further, inquiry is an inherently public strategy in that at least one other person (the target) becomes aware of the seeker’s desire for feedback. This suggests that perceived characteristics of the target such as the target’s evaluativeness and trustworthiness will influence the use of inquiry to obtain feedback irrespective of the presence or absence of observers. Target influences also are worthy of further study. An alternative conclusion that might be drawn from this research is that because of social and ego costs of seeking by inquiry, researchers should focus their efforts on better understanding how individuals obtain feedback using their own observations of the world around them (i.e., monitoring). This less public and less costly strategy may be a primary means by which self-generated feedback is obtained. This conclusion would lead researchers to consider questions such as: What cues are particularly salient to


individuals? How do they resolve discrepancies between cues? And what are the typical biases that monitors fall prey to in interpreting the cues around them?

REFERENCES

Ashford, S. J. (1989). Self-assessments in organizations: A literature review and integrative model. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior, Vol. 11, 133-174.

Ashford, S. J. (1986). Feedback seeking in individual adaptation: A resource perspective. Academy of Management Journal, 29, 465-487.

Ashford, S. J., & Cummings, L. L. (1983). Feedback as an individual resource: Personal strategies of creating information. Organizational Behavior and Human Performance, 32, 370-398.

Ashford, S. J., & Cummings, L. L. (1985). Proactive feedback seeking: The instrumental use of the information environment. Journal of Occupational Psychology, 58, 67-79.

Baron, R. A. (1986). Self-presentation in job interviews: When there can be “too much of a good thing.” Journal of Applied Social Psychology, 16, 16-28.

Baumeister, R. (1982). A self-presentational view of social phenomena. Psychological Bul-

letin, 91, 3-26. Bergles, S., & Jones, E. F. (1978). Drug choice as a self-handicapping strategy in response

to non-contingent success. Journal of Personality and Social Psychology, 36,405-417.

Blumberg, H. H. (1972). Communication of interpersonal evaluations. Journal of Person- ality and Social Psychology, 23, 157-162.

Brickman, P., & Bulman, R. J. (1977). Pleasure and pain in social comparison. In J. Suls & R. L. Miller (Eds.), Social comparison processes: Theoretical and empirical perspec- tives. Washington, D.C.: Halsted-Wiley.

Brockner, J., Davy, J., & Carter, C. (1985). Layoffs, self-esteem, and survivor guilt: Mo- tivational, affective, and attitudinal consequences. Organizational Behavior and Hu-

man Decision Processes, 36, 229-244. Brockner, J., & Swap, W. C. (1983). Resolving the relationships between placebos, misat-

tribution, and insomnia: An individual-differences perspective. Journal of Personality

and Social Psychology, 45, 32-42.

Broil, L., Gross, A., & Piliavin, I. (1974). Effects of offered and requested help on help seeking and reactions to being helped. Journal of Applied Social Psychology, 4, 255-

258. Carver, C. S., Antonio, M., & Scheier, M. F. (1985). Self-consciousness and self-

assessment. Journal of Personality and Social Psychology, 48, 117-124. Eagley, A. H. (1967). Involvement as a determinant of response to favorable and unfavor-

able information. Journal of Personality and Social Psychology, Monograph 7. Fedor, D. B., Eder, R. W., &Buckley, M. R. (1989). The contributory effects of supervisor

intentions on subordinate feedback responses. Organizational Behavior and Human Decision Processes, 44, 396-414.

Felson, R. B. (1980). The effect of self-appraisals of ability on academic performance. Journal of Personality and Social Psychology, 47, 944-952.

Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117- 140.

Fisher, C. D. (1979). Transmission of positive and negative feedback to subordinates: A laboratory investigation. Journal of Applied Psychology, 64, 533-540.


Gergen, K. (1974). Toward a psychology of receiving help. Journal of Applied Social Psy- chology, 4, 184193.

Goffman, E. (1956). Embarrassment and social organization. American Journal of Sociol-

ogy, 62,264-271. Herold, D. M., & Parsons, C. K. (1985). Assessing the feedback environment in work

organizations: Development of the job feedback survey. Journal of Applied Psychol-

ogy, 70, 290-306. Hoffman, P. J., Earle, T. C., & Slavic, P. (1981). Multidimensional functional learning

(MFL) and some new conceptions of F. B. Organizational Behavior and Human De- cision Processes, 27, 290-306.

Ilgen, D. R., Fisher, C. D., &Taylor, S. M. (1979). Consequences of individual feedback on behavior in organizations. Journal of Applied Psychology, 64, 359-371.

Ilgen, D. R., & Moore, C. F. (1987). Types and choices of performance feedback. Journal of Applied Psychology, 12, l-6.

Ilgen, D. R., & Knowlton, W. A. (1980). Performance attributional effects on feedback from subordinates. Organizational Behavior and Human Performance, 25, 441-456.

Janis, I., & Mann, L. (1977). Decision making. New York: Free Press. Jones, E. E., &Gerard, H. B. (1967). Foundations of Social Psychology. New York: Wiley. Kolditz, T., & Arkin, R. (1981). An impression management interpretation of the self-

handicapping study. Journal of Personality and Social Psychology, 90, 492-502. Leary, M., Robertson, R. B., Barnes, B. D., & Miller, R. S. (1986). Self-presentation of

small group leaders: Effects of role requirements and leadership orientation. Journal of

Personality and Social Psychology, 51, 742-748. McFarlin, D. B., & Blascovitch, J. (1981). Effects of self-esteem and performance feedback

on future affective preferences and cognitive expectations. Journal of Personality and Social Psychology, 40, 52 l-53 1.

Meyer, W. U., & Starke, E. (1982). Own ability in relation to self-concept of ability: A field study of information seeking. Journal of Personality and Social Psychology, 34, 901- 906.

Miller, D. T. (1976). Ego involvement and attributions for success and failure. Journal of Personality and Social Psychology, 34, 901-906.

Nadler, A. (1986). Help-seeking as a cultural phenomenon: Differences between city and kibbutz dwellers. Journal of Personality and Social Psychology, 51, 976-982.

Northcraft, G. B., & Earley, P. C. (1989). Technology, credibility, and feedback use. Or- ganizational Behavior and Human Decision Processes, 44, 83-%.

Riordan, C. A., James, M. K., & Dunaway, F. A. (1985). Interpersonal determinants of helping and the transgression-compliance relationship. Journal of Personality and So- cial Psychology, 125, 365-372.

Sachs, P. R. (1982). Avoidance of diagnostic information in self-evaluation of ability. Per-

sonality and Social Psychology Bulletin, 8, 242-246.

Schoeneman, T. J. (1981). Reports of sources of self-knowledge. Journal of Personality, 49, 289-293.

Schmitt, N., Coyle, B. W., & Saari, B. B. (1977). Types of task information feedback in multiple-cue probability learning. Organizational Behavior and Human Performance, 18, 316-328.

Shrauger, J. S., & Lund, A. K. (1975). Self-evaluation and reactions to evaluations from others. Journal of Personality, 43, 94-108.

Shrauger, J. S., & Sorman, P. (1977). Self-evaluations, initial success and failure, and improvement as determinants of persistence. Journal of Consulting and Clinical Psychol- ogy, 45, 784-795.

Skinner, B. F. (1953). Science and Human Behavior. New York: Macmillan.


Swarm, Jr., W. B., Griflin, Jr., J. J., Predmore, S. C., & Gaines, B. (1987). Cognitive- affective cross-fire: When self-consistency meets self-enhancement. Journal ofExper- imental Social Psychology, 18, 201-215.

Trope, Y. (1982). Self-assessment and task performance. Journal of Experimental Social

Psychology, 18, 201-215. Trope, Y., & Ben-Yair, E. (1982). Task construction and persistence as means for self

assessment of abilities. Journal of Personality and Social Psychology, 42, 637-645. Vroom, V. H. (1964). Work and Motivation. New York: Wiley. Walsh, J., Ashford, S., & Hill, T. (1985). Feedback obstruction: The influence of informa-

tion environment on employee turnover intentions. Human Relations, 38, 23-46.

Zuckerman, M., Brown, R. H., Fox, G. A., Lathin, D. R., & Minasian, A. J. (1979). De- terminants of information seeking behavior. Journal of Research in Personality, 13,

161-174.

RECEIVED: June 20, 1988

GREGORY NORTHCRAFT SUSAN ASHFORD - …webuser.bus.umich.edu/sja/pdf/PreservationEverydayLife.pdftors-performance expectations, self-esteem, ... confident self-presentation (Goffman,

Documents