FIRST IMPRESSIONS VERSUS GOOD IMPRESSIONS: THE EFFECT OF SELF-REGULATION ON INTERVIEW EVALUATIONS

By CYNTHIA R. NORDSTROM, ROSALIE J. HALL, and LYNN K. BARTELS Journal of Psychology Interdisciplinary & Applied, September 1998, Vol. 132 Issue 5, pp. 477-491.

ABSTRACT. Self-regulation may interfere with the ability to fully use situational information to form impressions ofothers. The demands posed by self-regulation were studied in a job interview situation. In the context of a simulated job interview, participants in the high cognitive load condition (interviewers) were less able to correct their initial characterizations of a job applicant with situational information than participants in the low cognitive load condition (observers). Results are discussed in terms of the person-perception literature and how cognitive load effects could be minimized to increase the comprehensiveness of interviewer assessments.

A PERSON'S ABILITY TO PERFORM one cognitive task may be impaired when he or she is simultaneously
attempting to perform other cognitive tasks (Gilbert, Pelham, & Krull, 1988; Logan, 1979; Navon & Gopher,
1980). A typical job interview provides an excellent example of a situation in which both interviewer and interviewee are trying to accomplish multiple competing tasks, including impression management, ingratiation and influence attempts, self-presentation, and evaluation of each others' traits and motives. Thus, in such a situation, one attempts to form accurate impressions of others while concurrently engaged in a host of self-regulatory behaviors (Gilbert & Osborne, 1989; Swann, 1984).

When a person monitors, visualizes, or mentally reviews his or her own actions (e.g., planning what to say before actually saying it), he or she is performing self-regulatory activities (Wegner, Schneider, Carter, & White, 1987). If cognitive resources are indeed limited, as some models suggest (Nisbett & Ross, 1980; Simon, 1955), self-regulatory behavior may diminish cognitive "supplies," hindering additional cognitive processing. Thus, in a social situation, a person may become so immersed in his or her own self-regulatory activities that the ability to draw accurate inferences about social partners is limited because of the lack of necessary cognitive resources. In the
current study, we examined how increased attentional demands stemming from self-regulation affect an interviewer's
ability to integrate relevant situational information about a job candidate in a simulated job interview.

Person Perception

Understanding how self-regulatory behavior may impair the ability to form comprehensive social perceptions requires consideration of the process of impression formation. Gilbert, Pelham, and Krull (1988) suggested that, to arrive at their perceptions of others, people typically use three sequential processes that differ in the amount of cognitive resources required: (a) categorization--perceiving the actions of others, (b) characterization--forming dispositional inferences about an actor, and (c) correction--modifying those inferences with information about situational constraints.

Categorization is viewed as a relatively automatic process that occurs immediately and does not require conscious attention (Lackoff, 1987; Rosch, 1978). Characterization and correction, on the other hand, are seen as more effortful processes, requiring conscious attention (Higgins & Bargh, 1987). Thus, one may characterize people fairly effortlessly but may neglect to correct those impressions to take into account situational information that conflicts with the initial characterizations (Gilbert, Krull, & Pelham, 1988; Gilbert & Osborne, 1989; Reeder, Fletcher, & Furman, 1989; Smith & Miller, 1983). Perceivers do not appear to lack the situational constraint information needed for correction; they are simply unable to use the information to modify their original characterizations.

Person Perception and the Selection Interview

Gilbert and his associates' (Gilbert et al., 1988; Gilbert & Osborne, 1989) model seems particularly applicable to the typical selection interview situation. During a typical job interview, both interviewers and interviewees perform substantial self-regulation (e.g., formulating questions and answers; revising answers while trying to appear knowledgeable, competent, and poised), placing both individuals under considerable cognitive load. According to the model, both interviewers' and interviewees' mutual impressions may be incomplete because of self-regulatory demands.

For instance, interviewers engaged in self-regulation activities may have the cognitive resources available to categorize and characterize the behavior of interviewees, but they may not have enough resources available tocorrect their characterizations to account for situational constraint information. Unfortunately, this situation implies that under cognitive load conditions, interviewers may not be accurately assessing a prospective job candidate's true strengths and weaknesses and consequently may make a poor hiring decision.

To address this issue, we used Gilbert and his associates' model of person perception to examine the selection interview process. We asked participants in the study to evaluate a job candidate twice-once after reading through
the job candidate's written credentials (Time 1 ratings) and again after meeting the candidate in a simulated job
interview situation (Time 2 ratings). We manipulated the participants' self-regulation demands by asking one
participant to actively conduct the interview with the job candidate ("interviewer" role) while another participant passively observed the interview taking place ("observer" role). Thus, interviewers had a higher cognitive load, imposed by the need to self-regulate. Passive perceivers, on the other hand, had the luxury of being able to form impressions of a job candidate without concerning themselves with the mechanics of social interaction. We expected that interviewers and observers would assess the candidate similarly on the Time 1 ratings but that their Time 2 ratings would differ substantially because of the varying cognitive load demands.

The methods used in this research were modeled on those used by Gilbert, Krull, and Pelham (1988). However, we improved on their methodology by collecting both pre- and post-cognitive load manipulation ratings. Additionally, in this study we asked participants to make judgments about a wide range of the target person's characteristics (e.g., introversion, friendliness, intelligence, truthfulness, effectiveness) consistent with real-life impression formation
situations (Ambady & Rosenthal, 1992, 1993; Bassili & Smith, 1986; Fiske & Cox, 1979). The following
hypotheses were tested:

Hypothesis 1: Low self-regulating participants (observers) and high self-regulating participants (interviewers) will not
differ in their recall of the interview situation information.

Hypothesis 2: Low self-regulating participants who receive disconfirming situational information will show a greater change in their ratings (trait, hiring, and salary recommendations) of the job applicant than will high self-regulating participants who receive the same information.

                                         Method

Overview

We asked participants to evaluate a job candidate for a teaching position in order to give her feedback about how she might be perceived by prospective interviewers. The participants based their initial hiring decisions on written information from an application packet. The decisions were indicated by participants' responses to three measures we created: a trait rating, a hiring recommendation, and a salary recommendation measure. Participants were then assigned to pairs in which one interviewed the job candidate and the other observed the interview as it took place.
The role of the job candidate was played by a research associate.

Half of the participants observed an interview that confirmed the information given earlier in the application materials. The other half observed a disconfirmatory interview. After the interview, the participants again completed the trait rating, hiring recommendation, and salary recommendation measures. They also were asked to recall information about the job candidate and to complete a manipulation check measure, the Self-Consciousness Scale (Fenigstein, Scheier, & Buss, 1975), and an interview experience measure.

Participants

We recruited a sample of 92 undergraduate students from a midwestern university's psychology subject pool. Participants were given extra credit for participation. Sixty-one of the participants were female, 17 were male, and 14 did not indicate their sex. Sixty-three of the participants were Caucasian, 13 were Black, 7 were Asian, 3 were Hispanic, and 2 were Native American. Three participants classified themselves as "other," and one participant did not designate ethnicity. The average age of the participants was 24 years.

Most of the participants had at some time been interviewed for a job (97%). However, only 12 participants (13%) considered interviewing job candidates as one of their major job duties.

Interviewers' and observers' responses on the Self-Consciousness Scale (Fenigstein et al., 1975) and the interview
experience measure were compared. Interviewers and observers did not significantly differ in their responses to the
Self-Consciousness Scale, t(90) = -.37, and the interview experience measure, t(90) = 1.57, p > .05.

Procedure

Pre-interview procedure. The study was depicted as a joint venture between the psychology and education departments. The job candidate (the associate of the researchers) was described to the participants as an education major who was ready to enter the job market and who could benefit not only from role playing the interview, but from any information the participant could provide about the impression she made. Participants were led to believe that one of the purposes of the study was to provide the job candidate with feedback.

Information about the job candidate came from two sources: an application packet and a face-to-face job interview. First, we asked the participants to review an application packet containing the following items about the job candidate: job description, resume, transcript, job application, and three letters of recommendation. The packet was designed to lead participants to form negative trait attributions (e.g., low grades). We asked the participants (a) to
indicate their impressions of the job candidate by rating her on a list of job-relevant traits, (b) to indicate whether they would recommend hiring the candidate at this point, and (c) to recommend a starting salary for the job candidate if hired.

Self-regulation manipulation. To introduce the self-regulatory manipulation, participants were randomly assigned to play the role of either observer (low self-regulation condition) or interviewer (high self-regulation condition). Participants who passively observed the interview taking place were considered to have lower self-regulatory demands and were thus more likely to perceive and integrate the situational information presented in the interview. Interviewers were considered higher in self-regulation demands (or "cognitively busy") as they were actively conducting the interview.

To further increase the interviewers' cognitive loads, we gave them a list of 10 desirable behaviors (e.g., smile, maintain good eye contact) to study in preparation for their role play with the job candidate. The job candidate and the interviewer were seated at a table directly across from one another; the observer was seated at the end of the same table so that both interviewer and observer were equally distant from the job candidate. The close physical proximity of the observer was also expected to enhance the interviewer's cognitive load.

Situational constraint information. The interviewers were provided with a list of 16 questions to ask the job candidate (structured job interview format). The face-to-face interview took approximately 10 min. During each interview, one of the two associates who played the role of the job candidate responded with answers from either a
confirmatory or a disconfirmatory script (both associates role-played both scripts). The disconfirmatory script contained plausible situational explanations for some of the negative information that was contained in the application packet. For example, the candidate explained that she had poor grades (on the transcript) because she had been injured in a car accident. These mitigating responses were intended to produce situational attributions for the less than impressive information in the application packet.

In the confirmatory script condition, the information in the script was consistent with the application packet information. That is, both sets of materials provided negative applicant information. For instance, the candidate rationalized that she received poor grades because teachers took points off for "really minor, stupid reasons." The interview responses in the confirmatory script were intended to confirm the negative dispositional attributions created by the application packet. Thus, half of the pairs heard answers from the disconfirmatory script, which contained situational information to be integrated with the negative information from the application packet, and the other half heard answers from the confirmatory script, which corroborated the negative information from the application packet.

Post-interview procedure. After the interview, the participants were again asked to complete the trait-rating scale,
the hiring recommendation measure, and the salary recommendation item. In addition, participants completed a manipulation check measure, an incidental recall measure, the Self-Consciousness Scale, and an interview experience measure that included demographic questions. Upon completing these measures, the participants were
thanked and debriefed.

Measures

Manipulation check. We created an 18-item manipulation check measure to compare the self-regulatory behaviors of participants assigned to the interviewer and observer role conditions. That measure contained questions such as "I thought about how well I was doing." Items were modeled on those used by Kanfer and Ackerman (1989). Responses were made on a Likert-type scale ranging from 1 to 7. The internal consistency reliability estimate for this measure (Cronbach's alpha) was .76.

Incidental recall measure. The recall measure consisted of a set of items designed to measure how much information the participants could remember about the job candidate. It contained questions such as "What were the
interviewee's two college majors?" To answer these questions, participants had to recall information presented in the
application packet (4 items) and in the face-to-face interview with the job candidate (6 items). Responses to the
items were scored as correct (1) or incorrect (0). The internal consistency reliability estimates (KR-20) were .38 for the resulting packet recall scale and .47 for the interview recall scale. These low reliabilities are not surprising, given the non-unidimensionality of the recall items that assessed memory for various aspects of the job candidate.

Interview experience measure. On the basis of previous research (Osborne & Gilbert, 1992), we believed that interviewing experience might affect the participants' responses by decreasing the attentional demands of self-regulation. Therefore, we created an interview experience measure based on the combined information from the participants' responses to two interview experience questions: "Have you ever been interviewed as part of applying for a job? How many times have you had such an interview?" and "Have you ever had a job where one of your major duties was conducting interviews for the purpose of hiring new employees?" Demographic information (e.g., age, sex) also was collected.

Self-Consciousness Scale. Self-consciousness is the tendency to direct attention inward toward the self. We thought that individual differences in self-consciousness might affect individuals' tendency to self-regulate. We used the Fenigstein, Scheier, and Buss (1975) measure, which consists of 23 items rated on a scale ranging from extremely uncharacteristic (0) to extremely characteristic (4). The scale is composed of three subscales: private self-consciousness, public self-consciousness, and social anxiety. The internal consistency reliability estimate (Cronbach's alpha) for the entire Self-Consciousness Scale was .86. Because all three components of self-consciousness might reasonably be expected to impair a participant's ability to process and integrate social information, we used the total score on the Self-Consciousness Scale in all analyses.

Ratings of job candidate. The focus of the study was on the participants' ability to correct or adjust their impressions based on new situational constraint information. Therefore, we used shift scores for the trait, hiring, and salary dependent variables. The trait rating shift score consisted of the sum of the absolute values of all shifts in trait ratings (post-interview/pre-interview) across all 19 traits measured. The interview provided participants with a richassortment of new information that could potentially lead to a change in trait ratings. Furthermore, although we constructed the interview script to lead to predominantly positive changes in ratings, other sources of information (such as job candidate appearance and nonverbal behavior) had the potential to lead to either positive or negative adjustments on specific trait items.

The 19-item trait rating measure was designed to solicit participants' impressions of the job candidate on a variety of
trait descriptors. Items were chosen to represent those traits that would be important in assessing a job applicant
(e.g., capable/incapable). The measure was administered after exposure to the packet information and again after the job interview. The resulting responses were used to form a composite shift score, as described in the preceding paragraph. The internal consistency reliability estimate (Cronbach's alpha) for the trait shift score was .78, indicating that those participants who made large adjustments on one trait tended to do so on others.

Similarly, the recommendation shift score was the sum of the absolute value post-interview/pre-interview differences in responses to the recommendation items. The hiring recommendation measure consisted of nine hiring recommendation items, such as "Would you recommend hiring this person for the teaching job?" The scale was designed to measure the likelihood that the participants would hire the candidate. Responses were made on a 5-point Likert-type scale. Once again, this measure was administered to the participants twice. The internal consistency reliability (Cronbach's alpha) for the hiring recommendation shift score was .77.

On the salary recommendation measure, the participants were asked to respond to the following question: "What
salary would you recommend offering this candidate?" The salary shift score was the post-interview/pre-interview
difference between responses to the salary item. To provide the participants with a common frame of reference, we
gave them a range of starting teaching salaries extending from $18,000 to $28,000, with an average annual salary of $25,000.

                                         Results

We used analyses of variance (ANOVAs) to examine the data. Although 92 individuals participated in the study, 2 participants failed to fill out measures completely and thus were eliminated from some of the analyses. The degrees of freedom used in the F tests varied slightly as a function of these missing data.

Manipulation Check

To ensure that the self-regulatory manipulation was effective, we constructed a composite manipulation check score of the mean of the 18 manipulation check items. The results suggested that there were differences in the participants'cognitive loads, depending on self-regulation condition, t(90) = -5.06, p < .001. Interviewers (high self-regulation) had significantly higher scores on the manipulation check measure (M = 3.14, SD = .40) than did observers (low self-regulation; M = 2.65, SD = .51).

We anticipated that participants who habitually focused attention on themselves (as indicated by higher scores on the Self-Consciousness Scale; M = 2.26, SD = .64) would also report higher scores on the manipulation check measure if it was indeed detecting self-regulated attention. As expected, the two measures were significantly related to each other (r = .26, p < .05). The positive sign of this correlation suggests that the more self-conscious participants were, the more they reported regulating their behavior during the simulated interview. The manipulation check also was correlated significantly with interview experience, r = -.30, p < .01, suggesting that the attentional demands were lower for experienced interviewers than for inexperienced interviewers.

Recall

Gilbert, Pelham, and Krull's (1988) model suggests that the degree of self-regulation does not affect recall, but it does affect the ability of the individual to adjust or correct for the effects of the fundamental attribution error when new situational information is presented. Thus, in our study, we did not expect differences across conditions in memory for the situational information presented in the interview scenario. Accordingly, the recall measure was broken into "packet information" and "interview information," based on how the information was presented. An ANOVA was used to analyze participants' recall for packet information and interview information, in order to determine if there were, in fact, recall differences across conditions. Interviewers (M = 2.72, SD = 1.11) and observers (M = 2.67, SD = 1.12) had very similar recall for candidate information contained in the packet materials. This difference clearly was not significant, F(1, 88) = .03, ns. (Statistical power to detect a medium effect size in this analysis was approximately .65.)

In addition, consistent with Hypothesis 1, the ANOVA results were not significant for recall of information presented during the interview, F(3, 88) = .01, ns. Again, interviewers (M = 4.69, SD = 1.35) and observers (M = 4.67, SD = 1.23) demonstrated similar recall for situational information presented during the interview, F(1, 88) = .01, ns. Participants assigned to the disconfirmatory script conditions (M = 5.00, SD = 1.13) demonstrated better recall for interview information than those assigned to the confirmatory script conditions (M = 4.34, SD = 1.36), F(1, 88) = 6.30, p < .05. The interaction between role and script did not reveal a significant effect on the interview recall measure, F(1, 88) = .33, ns.

Dependent Variables

The second hypothesis suggested that interviewers would be less likely to modify their initial impressions of the job candidate because of their increased self-regulation. The three dependent variables were significantly correlated with each other: trait ratings with recommendation shift (r = .50, p < .001); trait ratings with salary shift (r = .22, p < .05); recommendation shift with salary shift (r = .39, p < .001). Because the self-regulation manipulation affected participants experienced in interviewing less strongly than it affected those with less experience, in all hypothesis tests we controlled for the effect of interview experience. We also controlled for which associate played the role of interviewee.

For the ANOVA performed on the trait rating shift variable, a significant regulation main effect was found, F(1, 86) = 7.76, p < .01 (see Table 1 for the cell means). As expected, the low-regulating observers made greater adjustments in their initial trait ratings (M = 28.09) than did the high-regulating interviewers (M = 22.57). The effect size was relatively small (R2 = .08). In contrast, there were no significant differences attributable to the script factor, F(1, 86) = .11, ns, nor was the regulation by script interaction significant, F(1, 86) = .40, ns. Thus, a main effect for regulation was found on the trait rating shift variable.

Table 1 contains the cell means and standard deviations for the trait, hiring-recommendation, and salary-shift variables. The test of the recommendation-shift variable showed a significant script effect, F(1, 86) = 9.97, p < .01. The means for the script effect on the recommendation-shift score are as follows: confirmatory script (M = 7.74), disconfirmatory script (M = 10.42). As one might expect, participants who were exposed to two sources of conflicting information (negative application packet/disconfirming interview script) shifted their hiring recommendations more than the participants who were presented with consistent information (negative application packet/confirmatory interview script). That is, those participants who received the situational constraint information incorporated that information into their hiring recommendations for the job candidate.

The analysis also indicated a marginally significant regulation by shift interaction, F(1, 86) = 3.05, p < .10. The recommendation shift means (Table 1) show the pattern that would be expected. Participants who were observers and received disconfirming information (M = 11.70) shifted their recommendations significantly more than participants who were interviewers and received disconfirming information (M = 9.13), t(44) = 2.20, p < .05. High and low self-regulating participants in the confirmatory script condition did not differ significantly in their hiring recommendation ratings (see Figure 1).

Contrary to our hypothesis, there were no significant differences attributable to self-regulation for the salary shift variable, F(1, 85) = 65, p = .42. Neither the script effect, F(1, 85) = 2.06, p = .15, nor the script by regulation interaction was significant, F(1, 85) = .01, ns.

                                        Discussion

In general, our results are consistent with Gilbert's and others' research (Nordstrom, Williams, & LeBreton, 1996). People under cognitive load tended to fail to correct their personality trait attributions with relevant situational information. The results suggest that simply conducting a job interview can generate enough cognitive load to obstruct corrective thinking.

Interestingly, we obtained a somewhat different pattern of results for the trait ratings than for the hiring recommendations. Trait ratings showed a simple main effect for self-regulation. Compared with the observers, the interviewers under high levels of self-regulation were less likely to correct their impressions of the job candidate and incorporate relevant situational constraint information. This finding is particularly interesting because situational information appeared to be equally available in memory for participants in both the high-and low-regulation conditions, as revealed by the interview recall measure.

In contrast to the strong main effect for self-regulation on the trait-rating variable, the hiring-recommendation variable showed a weak interaction effect. As predicted by theory, the difference in the hiring-recommendation change for low versus high self-regulating participants was present only when the interview script disconfirmed packet information. Low self-regulating participants engaged in more correction than did high self-regulating participants.

What might account for the different pattern of results in the two types of variables? One possibility is that the trait- and hiring-recommendation ratings initiated somewhat different depths of cognitive processing from the participants. We suggest that the formation of trait impressions was largely effortless and unconscious. In contrast, making specific hiring recommendations may have triggered more thorough, less automatic processing from the participants. Moreover, the hiring-related questions call for global evaluations of the job candidate, requiring the integration of multiple sources of information that would likely prompt more controlled processing. The trait items, on the other hand, call for a simple targeted matching process (does the job candidate possess this trait?), thereby limiting the depth of cognitive processing required.

Generalizability of Results

Although the results generally supported the hypotheses, we suggest that our findings likely underestimate the impact that cognitive load may have on the processing of job candidate information in selection interviews. The simulated selection interview was highly scripted. In real-life situations, the cognitive load on the interviewer will be increased if he or she also generates questions during the course of the interview (e.g., both initial as well as follow-up questions). Moreover, by using a scripted list of interview questions, we limited the opportunities for different interviewers to pursue different lines of questioning and thereby elicit different information from the job candidate Dipboye & Gaugler, 1993). Thus, we prevented high self-regulating interviewers from using the interview to gather
additional characterization-confirming information (i.e., no interviewer-generated questions were allowed).

Finally, interviewers in the present study had the luxury of making their decisions free from other managerial duties. In actuality, of course, managers seldom operate in single-task environments (Mintzberg, 1989). Nordstrom, Williams, and LeBreton (1996) suggested that when managers operate in busy work environments, their ability to correct impressions is hampered.

To date, the external validity of the effect of cognitive load on attributional correction has not been fully demonstrated--all research has been conducted within the lab. However, we tested Gilbert's model in a lab situation that simulated a real-world event. Much of the interview research has limited generalizability because paper and pencil or recorded stimulus materials were used (Arvey & Campion, 1982). Noting this trend, Ferguson and Fletcher (1989) stated, "Such methods ignore or avoid many of the factors inherent in social context and render results ungeneralizable." Thus, it is likely that much interview research has consisted of efforts to equate the evaluations made by passive observers in the lab to the assessments made by active, self-regulating interviewers in the field. Results from the current study suggest that this may not be a valid comparison.

Practical Implications

The interview is currently the most widely used selection technique (e.g., Harris, 1989), despite ongoing concerns about its validity as a predictor of future job performance. To improve the validity of the interview, researchers have advocated using interview panels and increasing the expertise of individual interviewers (e.g., through interviewer training; Campion, Palmer, & Campion, 1996).

In a panel interview, the same questions may be posed by a single interviewer while the other panel members observe the interview process (Campion, Pursell, & Brown, 1988). Afterward, all panel members evaluate the job candidate. Our research suggests that this procedure may place greater cognitive demands on the panel member asking the questions than on those members observing the job candidate. As an active perceiver, the panel member conducting the interview is less likely to correct his or her initial assessment of the job candidate than are the panel members observing the interview. An advantage of this interview procedure is that the nonparticipating panel members are not hindered by excessive self-regulation demands while they form impressions of the job candidate. Thus, the panel members, as passive perceivers, may be more likely to correct their initial assessment of the job candidate and take into account additional situational information.

Our finding of a significant correlation (r = -.30) between the manipulation check and the interview experience measures suggests that a second means of reducing the cognitive demands imposed by the interview process is to use practiced interviewers. Experienced interviewers may be less vulnerable to self-regulation effects as they already have interview scripts (cognitive structures held in memory, which describe the appropriate sequencing of events in a typical interview) at their disposal. Because scripts are believed to save cognitive resources (Abelson & Black, 1986; Gioia & Poole, 1984; Lord & Kernan, 1987), experts' performances would not be expected to suffer as a result of self-regulation (i.e., they would have cognitive resources to allocate to accurate person perception).

Our results provide evidence that self-regulatory demands affect interviewers' impressions of job candidates. Thosedemands also are expected to exert an influence on the types of impressions interviewees come away with from an interview. Just as interviewers may not be drawing accurate inferences because of self-regulation effects, so too might interviewees' views of their potential employer/prospective organization be distorted because of cognitive load effects. Further studies, similarly modeled, could be directed toward uncovering how self-regulatory processes influence the perceptions of interviewees.

                                        TABLE 1

Mean Trait, Hiring, and Salary Recommendation
Shifts, by Regulation and Script Condition

                                     Regulation condition

                                     Low              High
Script condition                  (observer)      (interviewer)

Trait-rating shift

Confirmatory script

M                                   27.13             22.85
SD                                  10.97              8.67

Disconfirmatory script

M                                   29.04             22.28
SD                                   8.34             10.13

Hiring-recommendation shift

Confirmatory script

M                                    7.58              7.90
SD                                   3.04              3.21

Disconfirmatory script

M                                   11.70              9.13
SD                                   4.95              4.06

Salary shift

Confirmatory script

M                                  503               784
SD                               1,312             1,162

Disconfirmatory script

M                                1,003             1,276
SD                               2,002             1,679

Note. Means are adjusted for confederate
and interviewer experience effects.
 
 

  FIGURE 1. Mean shift in hiring recommendation ratings as a function of level of self-regulation and type of
                                      interview script.

                                      REFERENCES

Abelson, J. A., & Black, J. B. (1986). Introduction. In J. Galambos, R. Abelson, & J. Black (Eds.), Knowledge
structure (pp. 1-18). Hillsdale, NJ: Erlbaum.

Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequence: A meta-analysis. Psychological Bulletin, 111, 256-274.

Ambady, N., & Rosenthal, R. (1993). Half a minute: Predicting teacher evaluations from thin slices of nonverbal
behavior and physical attractiveness. Journal of Personality and Social Psychology, 64, 431-441.

Arvey, R. D., & Campion, J. E. (1982). The employment interview: A summary and review of recent research.
Personnel Psychology, 35, 281-322.

Bassili, J. N., & Smith, M. C. (1986). On the spontaneity of trait attribution: Converging evidence for the role of
cognitive strategy. Journal of Personality and Social Psychology, 50, 239-245.

Campion, M. A., Palmer, D. K., & Campion, J. E. (1996). A review of structure in the selection interview.
Unpublished manuscript.

Campion, M. A., Pursell, E. D., & Brown, B. K. (1988). Structured interviewing: Raising the psychometric
properties of the employment interview. Personnel Psychology, 41, 25-42.

Dipboye, R. L., & Gaugler, B. B. (1993). Cognitive and behavioral processes in the selection interview. In N.
Schmitt, W. C. Borman, & Associates (Eds.), Personnel selection in organizations. San Francisco: Jossey Bass.

Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private self-consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43, 522-527.

Ferguson, J., & Fletcher, C. (1989). An investigation of some cognitive factors involved in person-perception during selection interviews. Psychological Reports, 64, 735-745.

Fiske, S. T., & Cox, G. (1979). Person concepts: The effects of target familiarity and descriptive purpose on the
process of describing others. Journal of Personality, 47, 136-161.

Gilbert, D. L., Krull, D. S., & Pelham, B. W. (1988). Of thoughts unspoken: Social inference and the self-regulation
of behavior. Journal of Personality and Social Psychology, 55, 685-694.

Gilbert, D. L., & Osborne, R. E. (1989). Thinking backward: Some curable and incurable consequences of cognitive busyness. Journal of Personality and Social Psychology, 57, 940-949.

Gilbert, D. L., Pelham, B. W., & Krull, D. S. (1988). On cognitive busyness: When person perceivers meet persons perceived. Journal of Personality and Social Psychology, 54, 733-740.

Gioia, D. A., & Poole, P. P. (1984). Scripts in organizational behavior. Academy of Management Review, 9,
449-459.

Harris, M. H. (1989). Reconsidering the employment interview: A review of recent literature and suggestions for
future research. Personnel Psychology, 42, 691-726.

Higgins, E. T., & Bargh, J. A. (1987). Social cognition and social perception. Annual Review of Psychology, 38,
369-425.

Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative aptitude-treatment
interaction approach to skill acquisition. Journal of Applied Psychology, 74, 657-690.

Lackoff, G. (1987). Women, fire and dangerous things. Chicago: University of Chicago Press.

Logan, G. D. (1979). On the use of concurrent memory load to measure attention and automaticity. Journal of
Experimental Psychology: Human Perception and Performance, 5, 189-207.

Lord, R. G., & Kernan, M. C. (1987). Scripts as determinants of purposeful behavior in organizations. Academy of
Management Review, 12, 265-277.

Mintzberg, H. (1989). Mintzberg on management: Inside our strange world of organizations. New York: The Free
Press.

Navon, D., & Gopher, D. (1980). Task difficulty, resources, and dual-task performance. In R. S. Nickerson (Ed.),
Attention and performance, VIII. Hillsdale, NJ: Erlbaum.

Nisbett, R., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice-Hall.

Nordstrom, C. R., Williams, K. B., & LeBreton, J. M. (1996). The effect of cognitive load on the processing of
employment selection information. Basic and Applied Social Psychology, 18, 305-318.

Osborne, R. E., & Gilbert, D. T. (1992). The preoccupational hazards of social life. Journal of Personality and
Social Psychology, 62, 219-228.

Reeder, G., Fletcher, G., & Furman, K. (1989). The role of observers' expectations in attitude attribution. Journal of Experimental Social Psychology, 25, 168-188.

Rosch, E. (1978). Principles of categorization. In E. Rosch & B. B. Lloyd (Eds.), Cognition and categorization (pp.
28-48). Hillsdale, NJ: Erlbaum.

Simon, H. A. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 99-118.

Smith, E. R., & Miller, F. D. (1983). Mediation among attributional inferences and comprehension processes: Initial
findings and a general model. Journal of Personality and Social Psychology, 44, 492-505.

Swann, W. B. (1984). Quest for accuracy in person perception: A matter of pragmatics. Psychological Review, 91,
457-477.

Wegner, D. M., Schneider, D. J., Carter, S. R., & White, T. L. (1987). Paradoxical effects of thought suppression.
Journal of Personality and Social Psychology, 53, 5-13.