Employee perceptions of performance appraisal

0 downloads 0 Views 807KB Size Report
While most research has studied performance appraisal effectiveness in terms ... panies: the appraiser's knowledge of the employee's job and performance; the ...
Employee perceptions of performance appraisal fairness in two organizations ELAINE M. EVANS AND STEVEN L. McSHANE Department of Psychology and Faculty of Business Administration, Simon Fraser University

ABSTRACT While most research has studied performance appraisal effectiveness in terms of the quality of the performance measurement, a growing literature has begun to recognize that appraisal effectiveness can also be influenced by employee attitudes towards the system. The present study extended this research by exploring: 1) which characteristics comprise the appraisal process from the employees' perspective, and 2) how those characteristics arerelatedto employee perceptions of performance appraisal fairness. Based upon factor analysis of survey data from management and professional employees in two large Canadian organizations, six characteristics were found to comprise the appraisal process in both companies: the appraiser's knowledge of the employee's job and performance; the opportunity for employee participation; the establishment of specific andrelevantjob goals; discussion of employee development goals; discussion of company and department goals; and follow-up (feedback) on appraisal results. However, the relative importance of these characteristics to overall appraisal fairness ratings was quite different in the two companies. In addition, the favourableness of the mostrecentappraisal was associated with the employee's belief in the overall fairness of the appraisal process in both companies.

Performance appraisal effectiveness has typically been described and studied from the perspective of how accurately employee performance is measured. Whether the focus of the research is on the properties of the appraisal instrument or the context within which the appraisal is conducted, the objective is usually the same - to identify the conditions which affect the validity, reliability, and other qualities of employee performance measurement (Hgen & Feldman, 1983; Kane & Lawler, 1979; Landy & Farr, 1980). Notwithstanding the relevance of these traditional considerations, performance appraisal effectiveness should also be recognized in terms of employee attitudes toward the system (Bernardin & Beatty, 1984; Carroll & Schneier, 1982; DeMarco & Nigro, 1983). It has been argued, for example, that employee acceptance of appraisal decision outcomes depends at least partly upon their belief that the observed characteristics of the performance appraisal system are consistent with fair process (Folger&Greenberg, 1985; Greenberg, 1986). Such beliefs about the fairness and/or accuracy of the appraisal system are thought to influence employee motivation. Additionally, these attitudes towards the appraisal system might influence other, more subtle, outcomes such as organizational commitment and turnover (DeMarco & Nigro, 1983). The authors would like to extend their thanks to Raymond Koopman for his helpful suggestions in the data analysis and to Gary Latham for his editorial comments on an earlier draft. CANAD. J. BEHAV. SCI./REV. CANAD. SCI. COMP. 20(2), 1988

178

EVANS & McSHANE

Researchers have begun to explore the relationships between various characteristics of performance appraisal systems and employee attitudes towards the appraisal systems. While most of these studies have relied on single-item measures and have been restricted to single samples, they present several characteristics of performance appraisal systems which should be included in future research. The most frequently studied feature of performance appraisals is the extent to which ratees are given the opportunity to participate in the process (Burke, Weitzel, & Weir, 1978; Dipboye & de Pontbriand, 1981; Greenberg, 1986; Greller, 1975; Hillery & Wexley, 1974; Landy, Barnes, & Murphy, 1978; Wexley, Singh, & Yukl, 1973). With few exceptions (e.g., Hillery & Wexley, 1974), employees have more positive attitudes towards appraisals when they are given the opportunity to participate in the process. This includes the opportunity for ratees to voice their opinions on relevant issues, explain their point of view, and request clarification regarding the ratings they receive. Goal setting is another possible predictor of employee reactions to the performance appraisal process (Burke etal., 1978; Dipboye &de Pontbriand, 1981;Greller, 1975; Landy et al., 1978). In most studies, employee reactions to performance appraisals have been more positive when the appraisal outcome is based upon established and agreed-upon goals, and when performance objectives are discussed in the appraisal interview. Apparently, goal setting and the use of standard criteria for measuring goal achievement enhance the employee's belief in the objectivity of the appraisal results. This is in marked contrast to the situation where personality traits of the employee (e.g., honesty) are merely rated with no clear standard for comparison. A third characteristic is the extent to which the appraiser is perceived to be knowledgeable of the employee's job and performance level. Landy et al. (1978) reported that employee attitudes towards the fairness and accuracy of the performance appraisal system increased with the supervisor's perceived knowledge of the ratee's performance (also Falk, Brief, & Barr, 1985; Greenberg, 1986). One possible extension of this characteristic is the degree to which the rater is also familiar with the job duties of the person being appraised. Two other appraisal characteristics which have received some attention are the perceived relevance of the factors on which the job incumbent is being evaluated (Dipboye & de Pontbriand, 1981) and the extent to which the appraisal is conductedfrequently and includes adequatefollow-up (Landy et al., 1978). Numerous researchers have noted that the relevance of the performance dimensions is a major factor in performance appraisal effectiveness (Bernardin & Beatty, 1984; Kane and Lawler, 1979). Dipboye and de Pontbriand (1981) found the relevance of the performance dimensions to be an important predictor of employee attitudes towards the system. With respect to the frequency of appraisal, Landy and his colleagues (1978) found more positive reactions to appraisals which were conducted at least once each year. Bernardin (1979; reported in Bernardin & Beatty, 1984) noted that

PERFORMANCE APPRAISAL FAIRNESS

179

respondents to his survey on performance appraisal were generally dissatisfied with appraisals which were conducted only once or twice a" year. A related issue might be the extent to which the employee has informal performance discussions throughout the year with the person who conducts the formal appraisal. A major objective of the present study was to identify the characteristics that comprise the performance appraisal process from the employee's perspective. Unlike most research in the area, which utilizes single-item statements of each dimension in the analysis, this study began with multiple items written around the five appraisal characteristics or dimensions listed above. Although the items are intended to cover aspects of the dimensions uncovered in the review of previous research, the purpose was very much to explore whether these dimensions or others would emerge in the analysis. Thus, through factor analyses of employee ratings of the items one can begin to explore whether these predicted characteristics truly comprise employee perceptions of the appraisal process. The use of samples from two separate organizations would also increase the reliability of the characteristics uncovered. The second major objective of this study was to examine the association of those characteristics with ratee perceptions of appraisal fairness. While various attitudes towards performance appraisals have been examined in previous research, the perception of fairness was specifically selected for this study based upon the procedural justice literature (Cohen, 1985; Greenberg, 1978,1981,1986; Thibaut & Walker, 1975; Tyler, Rasinski, & Spodick, 1985). Procedural justice refers to the individual's belief that fairness exists when allocative procedures satisfy certain criteria (Leventhal, Karuza, & Fry, 1980). Generally, an individual will prefer procedures meeting criteria that, based on his/her past experience, are likely to generate a fair distribution of resources or outcomes. According to the alternate theory of distributive justice, the individual makes a judgment of fairness solely through the process of comparing his/her outcomes to the outcomes of others (Leventhal et al., 1980). In cases where the outcomes of others are unavailable for comparison, one would anticipate that procedures would become more salient for judging fairness of the outcome. Performance appraisal has generally been recognized as such a situation, since a comparison group (performance ratings of peers) is not usually available. The procedural justice literature also proposes that the more ratees believe that the evaluation process is fair, the more they will tend to endorse the system and accept its outcomes. Thus, it would be of significant value to performance appraisal systems to identify those process characteristics which enhance ratee perceptions of fairness. Regression analysis was used for this purpose by utilizing summary scores (derived from the dimensions uncovered in the factor analyses) to explore the relative importance of each characteristic to fairness perceptions. This study also included the perceived favourableness of the latest performance appraisal as a predictor in the fairness analysis. Dipboye and dePontbriand (1981)

180

EVANS & McSHANE

found that favourable appraisal ratings can induce a halo bias in the ratee's opinions of the appraisal system. This is also consistent with the justice literature which has found a strong positive association between the favourableness of a personal outcome and the perceived fairness of the process from which that outcome was derived (Cohen, 1982; Greenberg, 1978, 1981). However, the procedural justice literature has found that belief in fair procedures can at least partially balance the bias in favour of personal gain (LaTour, 1978; Lind, Kurtz, Musante, Walker, & Thibaut, 1980; Thibaut & Walker, 1975). One would anticipate that the above effect would be particularly relevant to situations such as performance appraisal where the meaning of the outcome is somewhat vague because a comparison group is not generally available. Consequently, the perceived favourableness of the latest appraisal was included in this study in order to test the extent to which favourableness of the latest appraisal outcome would influence overall fairness perceptions. METHOD Research sites Theresearchsites are two large Canadian private sector organizations. Company A is a telecommunications service firm which employs approximately 13,000 people, 2500 of whom are in management and professional positions. Company B is a financial institution with more than 2000 employees, 850 of whom are in management and professional positions. Both organizations use performance appraisal systems which combine Management by Objectives (MBO) with a trait rating scale and global assessment. The two systems had been in place for approximately two years prior to the study. In Company A, the performance review process begins with the supervisor completing most parts of the appraisal form alone. This includes: (1) listing the employee's majorresponsibilitiesand objectives over the previous year and describing therelevantstandards; (2) rating the employee on fourteen traits (e.g., leadership, initiative, decision making, etc.) using a four-point scale and describing pertinent examples of the presence or absence of these traits in the employee; and (3) rating the employee on a global five-point scale. The supervisor then discusses the appraisal with a performance review committee consisting of local managers. The purpose of the committee is to get a uniform understanding of performance and ensure that the supervisor has carefully thought through the performance evaluation. Next, the supervisor and employee meet to discuss the evaluation and develop a new performance plan for the forthcoming year. Prior to the meeting, the employee is given the option of completing a self-appraisal worksheet and bringing it to the meeting for discussion purposes. The completed appraisal form is signed by the supervisor, employee, and the next level of management. Finally, the supervisor is encouraged to communicate the employee's progress throughout the year and, on a specially prepared form, write down critical incidents. In Company B, the appraiser also begins by completing the performance appraisal form alone. First, the previously established objectives for the performance period are listed, including the weighting of each objective. Beside each objective, the employee's actualresultsare described, any discrepancies are explained, and the extent to which the objective has been met or exceeded is rated on a seven-point scale. It should be noted that, unlike Company A, the appraisal form and accompanying manual for Company B repeatedly emphasize the use of measurable ratings (e.g., 15 percent increase in commercial loans over the next twelve months). Next, the employee is rated on twelve traits (e.g., planning, organization, leadership, etc.) using the same seven-point scale. In the third section, the appraiser lists the employee's overall strengths and identifies specific areas for improvement. Finally, a global rating

PERFORMANCE APPRAISAL FAIRNESS

181

of the employee's performance is indicated on a seven-point scale. As in Company A, each level of the scale is anchored on the appraisal form and described more fully in the accompanying manual. When the appraisal form has been completed, it is taken to another manager (usually the appraiser's superior) for review and signoff. Then the performance appraisal is reviewed with the employee. Space is provided on the form for any comments that the employee wishes to document. New objectives are negotiated with the employee and the supervisor is encouraged to provide ongoing feedback with regard to those objectives throughout the year. Indeed, formal quarterly reviews are conducted with the employee to ensure that feedback is given. Subjects Questionnaires were distributed through inter-office mail to 300 randomly selected management and professional employees in each of the two participating organizations. These questionnaires were completed anonymously and were mailed directly to the first author. In Company A, 214 (71%) of the questionnaires were completed; in Company B, 183 (61%) were completed. The average respondent from Company A was in the 35-39 age group, had been with the organization for 11.8 years, and had been in his or her present position for 3.8 years. Women represented 38% of this sample. The average respondent from Company B was in the 30-34 age group, had been with the company for six years, and had been in his or her present position for 2.S years. Women represented 43% of this sample. Measures Consistent with the previous research, the characteristics of the performance appraisal system were identified through employee self-reports. There are two reasons for this. First, in spite of clearly written guidelines to the appraisers, there are differences in the way that the appraisal is conducted from one employee to the next. Indeed, some appraisal characteristics, such as the appraiser's knowledge, cannot be standardized across the organization. The other reason is that the perceived characteristics of the appraisal system, not the actual characteristics, directly influence the perceived fairness of the process. In this respect, the appraisal must give the impression of being fair as well as actually being accurate. Thirty-eight statements were written to tap thefiveaspects of performance appraisals identified from previous research: (1) employee participation in the process (nine items); (2) goal setting (eleven items); (3) appraiser's knowledge (eight items); (4) relevance of appraisal dimensions (seven items); and (S) appraisal frequency and follow-up (three items). These items were derived from previous studies (Burke et al., 1978; Dipboye & de Pontbriand, 1981; Landy et al., 1978) as well as the conceptual literature on performance appraisal. Although these statements were written around the five dimensions, the purpose was very much to explore whether these dimensions or others would emerge in the analysis. All 38 items used a seven-point Likert-type response scale ranging from "strongly agree" to "strongly disagree." Ten statements were negatively worded and reverse-scored to minimize acquiescent response sets. The statements were randomly ordered. In a separate section of the questionnaire, respondents were asked to indicate their overall perception of the fairness of their company's appraisal system. These answers were indicated on a seven-point scale ranging from "very fair" to "very unfair." Respondents were also asked to rate the favourableness of the results of their most recent performance appraisal.

RESULTS

Factor analyses Principal components factor analysis with varimax rotation was employed in order to examine the dimensions of the appraisal process as perceived by the employees

182

EVANS & McSHANE

in this study. The factor analysis was conducted in two stages. First, the 38 items describing the performance appraisal characteristics were factor analyzed using the data from Company A. The results of this initial orthogonal analysis revealed a number of items which highly overlapped amongst the different factors or did not load highly on any of the factors. There was no discernible improvement using oblique rotation. These items were removed and the same data were factor analyzed again with the remaining 21 items. The remaining 21 items were also factor analyzed using the data from Company B. The two factor analytic solutions were then compared and found to be very similar. The 21 item statements, along with means and standard deviations for the two samples, are shown in Table 1. The two factor analyses appear in Table 2. Based upon a scree plot of the eigenvalues, a six-factor solution was determined for both samples. The anticipated dimensions occurred as factors with certain exceptions. Three of the six factors were very similar to the dimensions anticipated on the basis of the literature review. These factors were Employee Participation, Appraiser's Knowledge, and Appraisal Follow-up. Items from the presumed Goal Setting dimension appeared on two factors: Goal Establishment and Employee Development. The Goal Establishment factor appears to be concerned with the process by which performance-related goals for the employee's current job are established, whereas the Employee Development factor focuses on the discussion of more general self-improvement goals for the employee. The presumed Job Relevance of Appraisal items did not form a separate factor. Instead, two of the Job Relevance items (items 15 and 16) loaded on the Goal Establishment factor. In hindsight, this finding is understandable because the performance dimensions in MBO systems are established through the goal-setting process, and both these items were concerned with basing performance appraisal on specified job standards, responsibilities, and goals. Finally, two items expected to form part of the Employee Participation factor formed a separate factor which we have labelled Goal Discussion. Unlike the Employee Participation factor items which suggest that the employee is given the opportunity to state his/her opinions, these two items only indicate that company and department goals are discussed during the appraisal. Based upon the factor analysis results, composite variables were created for each of the six factors by averaging the items representing each performance appraisal dimension. Item analyses of the composite variables indicated that there is quite a high level of homogeneity amongst the items, with perhaps the exception of the Goal Discussion items. The means, standard deviations, internal consistencies (a), and intercorrelations of these composite variables, as well as the measures of performance appraisal favourableness and fairness, were calculated for each sample and are displayed in Table 3.

PERFORMANCE APPRAISAL FAIRNESS

183

TABLE 1 Means and standard deviations of the twenty-one performance appraisal characteristic items Item Statement* 1. My appraiser has a good understanding of the skills required to perform my job. 2. My appraiser is familiar with all phases of my work. 3. My appraiser has excellent personal knowledge of my performance level in my current position. 4. My appraiser does not have a good knowledge of my position's duties. (R) 5. My appraiser has observed my performance under both routine and pressured conditions. 6. I am given the opportunity to state my side of all the issues discussed during my performance appraisal. 7. I am given the opportunity to express my feelings during my performance appraisal. 8. There is ample opportunity to discuss all aspects of my job during my appraisal. 9. I am encouraged to express my opinions on how my duties could be more efficiently performed. 10. I have a clear understanding of the reasons behind the appraisals I receive. 11. Possible means of self-improvement which I could take in my current position are not discussed during my performance appraisal. (R) 12. My personal development needs are discussed during my performance appraisal. 13. Goals for my job are not discussed during my appraisal. (R) 14. Possible actions which I could take to improve performance in my present position are discussed during my performance 15. My performance appraisal is based on specified performance standards for my position. 16. My performance is appraised according to previously established responsibilities, standards, and goals. 17. Any plans or objectives for my job are established and mutually agreed upon by my appraiser and myself. 18. My formal appraisals are connected to informal meetings between my appraiser and I which take place throughout the entire year. 19. Discussion and review of my performance is a continuous process, not one which occurs only during my formal performance appraisal. 20. Overall company goals are not discussed during my performance appraisal. (R) 21. The goals of my department are discussed during my performance appraisal.

Company A Mean SD

Company B Mean SD

5.22

1.90

5.91

1.47

5.00 4.88

1.90 2.02

5.54 5.26

1.79 1.89

5.13

1.93

5.71

1.74

5.44

1.88

5.95

1.48

5.80

1.61

5.80

1.53

5.96

1.55

6.07

1.35

5.48

1.75

5.49

1.75

5.11

1.69

5.26

1.72

5.26

1.87

5.70

1.57

5.05

1.89

4.86

2.03

4.96

1.74

4.81

1.87

5.29 4.84

1.93 1.84

4.98 5.19

2.00 1.84

4.66

1.97

5.62

1.67

5.16

1.73

5.75

1.57

5.54

1.66

5.16

1.91

3.68

2.13

3.78

2.20

3.59

2.15

3.89

2.22

3.68

2.10

4.48

2.15

4.58

2.06

5.56

1.70

Note: Items have been renumbered and grouped according to factors. (R) indicates reverse scored items. Scoring ranged from 1 (strongly disagree) to 7 (strongly agree).

TABLE 2 Factor analyses of performance appraisal characteristics

III.

V.

VI.

iv

v

VI

i iii i

ra

iiii i

ii

.32

.86 .82 .63 .54 .54

.39

.34

.31

.34

.35 .35

iiii l

.35

.82 .84 .80 .70 .64

i iii i

Employee Participation 6. 7. 8. 9. 10.

.83 .80 .71 .79 .53

.77 .81 .64 .63 .48

.38

.44

.48

Employee Development .79 .69 .61 .57

12. 13. 14. IV.

i

Appraiser's Knowledge 2. 3. 4. 5.

II.

Company B Factor Loadings

VI

iiii i

I.

Company A Factor Loadings H ID IV V

i i i i i

Factor Name & Item Numbers*

Goal Establishment 15. 16. 17.

.32

.39

.51 .34

-

.79 .73 .61

Appraisal Follow-up 18. 19.

.82 .84

Goal Discussion 20. 21.

percent of variance accounted for

.34

16.8

14.5

11.9

11.0

-

-

-

_

_

-

.31

.86 .69

9.0

7.4

16.9

14.0





-

-

.55

-

-

_

.83 .80

-

.81 .53 .57 .67

-

-

.84

-

.81)

-

-

-

.84 .75

11.2

10.2

9.7

7.8

Note: *Item numbers correspond to the item statements listed in Table 1. For clarity, only loadings of .30 and higher are presented.

1 9?

i

TABLE 3 Means, standard deviations, internal consistencies (a), and intercorrelations of study variables for both samples Variable 1. 2. 3. 4. 5. 6. 7. 8.

Appraiser's Knowledge Employee Participation Employee Development Goal Establishment Appraisal Follow-up Goal Discussion Favourableness Fairness

1 _ .53 .60 .40 .41 .32 .17 .33

2

.66 .61 .42 .51 .20 .41 .37

3

.58 .57 _ .45 .53 .41 .24 .29

4

.63 .60 .57 _ .33 .26 .08 .34

5

.39 .41 .43 .41 .31 .24 .26

6

7

8

.37 .40 .38 .39 .41 .14 .24

.37 .46 .35 .42 .36 .26 .38

.52 .48 .42 .43 .32 .18 .36 -

Mean 5.14 5.52 5.03 5.12 3.63 4.13 5.38 4.77

ConipanyA SD

.64 .34 1.42 1.52 1.91

.76 .34 .56

Company B

a

Mean

.90 .85 .80 .73 .75 .61 _ -

5.67 5.66 4.96 5.51 3.84 5.02 5.55 5.10

.38 .22 .54 .35 '.03

.66 .28 1.54

a

§

.88 .83

M

|

.69 .80 .82 .64 _ -

;

Note: Correlations above the diagonal are for Company A (n •• 214); correlations below the diagonal are for Company B (n = 183). When r > . 16, p < .05 for Company A. When r > . 18, p < .05 for Company B.

3

186

EVANS &McSHANE

TABLE4 Standardized multiple regression weights of performance appraisal system fairness Company A (n = 214) Without Full favourableness equation

Predictors Appraiser's Knowledge Employee Participation Appraisal Follow-Up Goal Establishment Employee Development Goal Discussion Favourableness R

Ri, */>