Leniency Effect Research Articles

An information-processing approach to leniency in performance judgments.

The traditional rendering-bias explanation of leniency in performance judgments was postulated to be incomplete. To demonstrate this insufficiency, an encoding-bias model of leniency was proposed, and hypotheses generated from both models were tested in a laboratory experiment. Sixty students viewed a videotape of a lecture, and their expectations of providing face-to-face feedback to the instructor were manipulated preobservation, postobservation, or not at all. Dependent variables included measures of encoding, private appraisals, retrieval, and public appraisals. As expected, results supported both explanations of leniency. However, the pattern of results suggests that a retrieval bias may also contribute to leniency in performance jugdments. A more complex view of leniency effects is proposed on the basis of principles of motivated cognitive processing

Leniency effect as a function of rating format, purpose for appraisal, and rater individual differences

The effect of rating format and non-performance variables on rating leniency were studied in two law enforcement organizations. One of these variables, trust in the appraisal process, was defined as the extent to which a rater believes that fair and accurate appraisal will be made in the organization. A measure of trust in appraisal accounted for a significant proportion of variance in performance ratings. The purpose of appraisal (i.e., feedback or promotion) also accounted for rating variance. A mixed-standard rating format showed less susceptibility to the non-performance variables on the extent of leniency. Discussion centers on the usefulness of rater and organizational variables in performance appraisal research.

Confounding among Measures of Leniency and Halo

Because the variability of rating distributions decreases with leniency, common indices of the psychometric performance of rating data may be confounded. Lenient rating distributions, for example, should show lower average scale intercorrelations (a measure of halo) than do less lenient rating distributions. Although psychometric theory would predict such confounding, it does not permit the estimation of the magnitude of this phenomenon. Consequently, the interrelationships among halo and leniency measures were examined using simulated rating data. Specifically, effects of leniency on 3 levels of halo were examined. As expected, results showed that as leniency increased, halo decreased when measured by dimension intercorrelations but increased when measured by standard deviations across dimensions. The magnitude of the convergent validities calculated for halo and leniency measures also decreased as leniency increased. In addition, although halo and leniency were orthogonal when no leniency was present, substantial confounding of these constructs occurred as leniency increased. Implications of these results for the use and integration of the various measures are discussed.

Self versus superior assessment of social science research employees

Self-assessment of performance was compared to superior's assessment on a sample of 60 personnel research workers. It was hypothesized that norms of self-acceptance and relatively low defensiveness will characterize members of this sample as a result of their educational background in Psychology and related areas. Therefore the previously noted leniency effect in self-assessment will not be expressed. The subjects and their superiors (N=10) rated the subjects performance on a thirty-item behaviorally based scale as well as on a global performance assessment item. No significant differences between superior's and subordinates' ratings were found for any of the scale items. The results also suggest that subordinates' ratings were significantly more highly correlated with the importance they attributed to the performance items than superiors' ratings. There were also noted differences between superiors and subordinates in the relationships between the assessment of various performance areas and the global performance assessment. The results are discussed in terms of the need for exploring the characteristics of self-assessment in unique samples and the possible advantages of incorporating self-assessment into personnel evaluation systems.

THE EFFECTS OF DEFENDANT AND JUROR ATTRACTIVENESS ON SIMULATED COURTROOM TRIAL DECISIONS

An experiment was conducted to investigate the interactive effects of the physical attractiveness of hypothetical defendants and mock jurors on judicial decisions. Seventy-eight college students rated their own physical attractiveness and then evaluated attractive, moderately attractive, and unattractive defendants as to the defendants' guilt or innocence, responsibility for the charges being brought, trustworthiness, happiness, honesty, intelligence, and likeability as well as recommended punishment for those convicted. As expected, more as opposed to less attractive defendants were convicted less, punished less severely, rated as less responsible for the charges being brought, happy, likeable, and trustworthy. Attractive subjects were more likely to convict than acquit unattractive defendants, while less attractive subjects did not differentially convict or acquit defendants across all levels of defendant physical attractiveness. Both attractive and less attractive subjects recommended the least severe punishment for attractive defendants; however, attractive subjects were harshest on unattractive defendants, while unattractive subjects were harshest on moderately attractive defendants. The results are discussed in terms of leniency effects when judging others with similar attributes.

Effect of Reciprocal Leniency on the Relation between Consideration Scores from the Leader Behavior Description Questionnaire and Performance Ratings

This study tested the hypothesis that “reciprocal leniency” moderated the relationship between Consideration scores on the Leader Behavior Description Questionnaire—Form XII and performance ratings. Reciprocal leniency was defined as a response style in which scores on the questionnaire are affected by harsh, lenient, or fair ratings made by the supervisor. Results partially supported the hypothesis.

PSYCHOMETRIC PROPERTIES OF SUBORDINATE RATINGS OF MANAGERIAL PERFORMANCE

The purpose of the present study was to compare the psychometric properties of subordinate ratings of managerial performance to those of supervisor and self ratings. Subordinate ratings (n = 365), self‐ratings (n = 80) and supervisor ratings (n= 80) were obtained on eight performance dimensions for middle level managers in a multinational corporation. Multitrait‐multimethod analysis of variance indicated significant effects for convergent validity, halo error, and discriminant validity. Comparisons between rating sources indicated that subordinate ratings are more similar to supervisor ratings than to self ratings in terms of convergent validity and leniency effect. A comparison of the present results to other studies indicated less convergent validity, more discriminant validity, and less halo error than found by others.

A fairer examination of rating scales when used for performance appraisal in a real organizational setting

AbstractFifty‐seven managers in an aircraft plant who had received rater training, appraised their subordinates (n = 376) using both a BARS format and a graphic scale. Less leniency and halo effect occurred with the BARS format. In addition, the use of BARS resulted in higher interrater reliability than ratings obtained with the graphic scale. Finally, the overall (global) evaluation derived with the BARS form was significantly less susceptible to leniency effect. A tentative conceptual framework to explain this finding was suggested.

Should Henrietta be Punished or Rewarded? The Effects of Name Desirability on Responsibility Attribution and Sanction Assignment

AbstractThis study investigated effects of first name desirability on both attribution of responsibility (AR) and sanction assignment (AS) to assess the generality of name bias effects on the attribution process. Subjects (n = 112) were equally divided by sex, and randomly assigned to four groups which received stimulus stories with all male or all female actors. The stories represented three of Heider's levels with high magnitude positive or negative outcomes produced by male or female actors with desirable or undesirable first names.Analysis of AR ratings revealed significant main effects for both Level (L) and Outcome (0) as well as a significant L x 0 interaction; no significant main effects or interactions involving subjects’ sex or actor's sex (A) or name (Na) were revealed, although there was an Lx Na trend (p = .07). For AS, significant main effects were found for L, 0, and name desirability (Na). Significant two-(S × A, Lx 0), and (L × Na) and three-way (L × 0 × Sand Lx 0 × Na) interactions were found. Evidence of the effect of name desirability on AS is provided by the significant main effects for name desirability and by the L × Na interaction and by the L × 0 × Na interaction. Punishment ratings exceeded reward ratings, and significantly less reward was assigned to male actors with undesirable names and significantly less punishment was assigned to female actors with undesirable names, implying that perhaps a cognitive factor is involved, generating a type of gender appropriate leniency effect. These data also indicate that the impact of name desirability is limited primarily to subjective judgmental areas and can be weakened by focusing upon objective features of causal evidence and target's behavior.

EFFECTS OF ITEM SEQUENCING AND GROUPING IN LOW‐INFERENCE TYPE QUESTIONNAIRES

In the literature on questionnaire construction, conflicting conclusions have been reached concerning the possible effects of presenting items within labeled groupings and/or in a different order on the conclusions drawn from a study. For example, Sudman and Bradburn (1974) suggest that, without additional research, recommendations cannot be made regarding the possible influence of item order and grouping, but Kidder (1981) suggests that different forms should be administered and Schriesheim (1981b) suggests that items should be ordered randomly. One apparent reason for the conflicting judgments is that relatively few studies have been conducted which could serve as bases for recommendations. A second reason is that available findings generally do not support one another. For example, Krant, Wolfson and Rothenberg (1975) concluded that item location on a questionnaire could affect the average response, but Clancy and Wachsler (1971) concluded that item location was of only minor importance. Also, Metzner and Mann (1953) concluded that item grouping did not affect the item correlations, but Schriesheim (198 la, 1981b) and Schriesheim and DeNisi (1980) concluded that grouping items into logical categories instead of randomly ordering the items impaired the magnitude of the correlations indicative of the convergent and divergent validities and resulted in a leniency effect on the item means. Another example is Schriesheim and DeNisi (1980) who observed that item order (and grouping) affected factor analytic results, but Kane (1971) concluded that different orders for semantic differential items did not affect the factor compositions. Because only a limited number of studies have been conducted, one might view the mixed findings as resulting from some investigators having observed either a unique situation or a chance result. However, the methodologies of the reported studies differed enough that it might not even be meaningful to compare the results from the studies, that is, the differences in the methodologies might be an alternative explanation for the mixed findings. One particular feature about the methodologies used to date is that no study employed a sufficient number of questionnaire forms to examine independently the effects of grouping and order on the item means and correlations. For example, in the Schriesheim studies, only two forms were used, one on which the items were ungrouped and randomly ordered and a second on which items were grouped according to logical classifications and appeared in a different order. Thus, order and grouping effects were confounded and it probably is not meaningful to compare the Schriesheim findings to those of studies that investigated only either a grouping or an order effect. By using three different forms of a questionnaire in this study, it was possible to examine the effects of both order and grouping in separate analyses. The criteria for assessing the effects were item means, item correlations, and factors underlying the

Prediction of performance in initial Officer training using reference reports*

The usefulness of reference reports has tended to be neglected in personnel selection research, and available evidence is not encouraging. In the study reported, ratings on the headteacher's report form used in Royal Navy Officer selection were examined and compared with training performance. Although the ratings of candidates (n = 263) were subject to restriction of range and possible leniency effects, amongst entrants (n = 103) a total score derived from the seven individual scales had a correlation (corrected for restriction) of ***0–36 with overall training performance. It is concluded that under some circumstances head‐teacher's reports can make a useful contribution to the selection process. It is suggested that such predictive power as the ratings have derives from their providing, within carefully constructed format, samples of characteristic behaviour similar to the criterion.

The Effect of Grouping or Randomizing Items on Leniency Response Bias

This study was concerned with the effects of item presentation mode on the degree of leniency bias inherent in responses to standard field research questionnaires. Two types of modes were examined : the first with items measuring the same dimensions grouped together and the second with items measuring the same dimensions distributed randomly throughout the questionnaire. Sixty respondents completed questionnaires containing items from the Leader Behavior Description Questionnaire (Stogdill, 1963) and the Michigan Four Factor Leadership Questionnaire (Taylor and Bowers, 1972); there were thirty respondents in each of the modes (grouped and random). The random relative to the grouped mode showed substantially less leniency response bias, as assessed by both correlational and factor-analytic procedures. However, the magnitude of these effects was still considerable for the random mode. There were no notable differences in leniency effects between the two questionnaires, but some differences were obtained for specific leadership dimensions. Based upon these and the total set of findings, implications for questionnaire validity and for future research are discussed.

The Effect of Reciprocal Leniency on the Relationship Between Consideration Scores from the LBDQ and Performance Ratings.

This study tested the hypothesis that reciprocal moderated the relationship between Consideration scores on the LBDQ-Form XII and performance ratings. Reciprocal leniency was defined as ...

The effect of leniency on leader behavior descriptions

This article reports the results from five studies designed to estimate the magnitude and effects of leniency on subordinate-provided descriptions of leader Consideration and Initiating Structure. Using a newly developed scale to measure leniency response bias (the tendency to describe others in favorable but probably untrue terms), the results of the studies indicate that subordinate reports of leader Initiating Structure are not particularly susceptible to the effects of leniency. The results for Consideration, however, showed that: (1) consideration items were not socially neutral and were susceptible to leniency; (2) consideration reflected an underlying leniency factor when applied in a field setting; (3) leniency explained a substantial proportion of the variation involved in Consideration-dependent variable correlations; and (4) this did not appear to result from conceptual overlap between Consideration and leniency, but rather from spurious correlation through the susceptibility of the measures to the effects of leniency. The importance of these findings for leadership research are discussed, and several alternatives for the control of leniency in leader behavior description are discussed.

Theoretical and Applied Issues in Self-Control

Recent developments in self-control research with children are critically reviewed, and a distinction is drawn between studies investigating self-determined performance standards and those investigating self-reinforcement. The reinforcing function of self-control in humans and animals is discussed, as is the potential functional value inherent in labeling certain strategies “self-determined.” The maintaining capacities of self-determined systems, as” well as their effects on initial performance, are compared to those of externally controlled systems. At present, self-control strategies are at least as effective, and in some cases more effective, than externally controlled strategies in promoting desired behavior change. The effects of stringency and leniency of performance standards and reinforcement levels are evaluated, and methods are advanced to elicit the self-selection of stringent standards. Future directions for research on self-control are proposed.