The spillover effect at the group level

(1)

1

Thesis

The spillover effect at the group level

Name: Suzan van Hesteren Student number: 10642846 Thesis supervisor: dr R.S. Ghita Date: 26 June 2017

Word count: 9,606

MSc Accountancy & Control, specialization Control

(2)

2 Statement of Originality

This document is written by student Suzan van Hesteren who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

The aim of this paper is to investigate whether the spillover effect is different in a situation in which only one employee is evaluated than in a situation in which multiple employees are evaluated.

The spillover effect exists when an objective performance measure of an activity influences the subjective evaluation of another activity, even though both activities are completely independent of each other.

Literature states that the spillover effect exists. However, a limitation of this study is that it only examines the situation in which only one employee is evaluated. It may be that the spillover effect is different when multiple employees are evaluated, because evaluators make more reasoned decisions when they can compare employees. The research question in this paper is therefore: ‘How does the amount of evaluated employees affect the spillover effect? This question will be answered by using a scenario-based experiment.

This paper adds to the existing literature, because it investigates the spillover effect at the group level. This study tests whether the results that are knownabout the spillover effect can be generalized to a situation in which multiple employees are evaluated.

Found is that the spillover effect indeed exists when only one employee is evaluated. However, it has not been found that the spillover effect exists when multiple employees are evaluated. This might indicate that the results which are found in existing literature cannot be generalized. There is also no significant difference between the spillover effect when only one employee is evaluated and the spillover effect when multiple employees are evaluated.

The results of this paper are relevant, because many supervisors evaluate multiple employees. Knowing that the spillover effect does not exist when multiple employees are evaluated can save firm owners time and money because then they do not have to take actions to minimize the spillover effect. When multiple employees are evaluated firm owners do not have to spend time and money on organizing trainings to make employees more aware of the spillover effect. They also do not have to spend time and money on finding solutions on how they can achieve that a supervisor evaluates an employee without knowledge about previous performance of that employee.

(4)

4

1 Introduction

Most incentive systems include objective and subjective measures. They are used both,

because this creates a more complete view about employee performance (Bol & Smith, 2011). However, there is a complex trade-off with subjective evaluations. They can give a more complete view, but they cannot be verified by a third party. Therefore supervisors can be influenced by their own preferences when subjectively evaluating performance (Bol & Smith, 2011).

Even in the absence of preferences a supervisor’s subjective evaluation can be biased. The availability of objective measures on previous independent activities can unintentionally influence a supervisor’s judgment. So the different measurement types can influence each other, even though the objective and subjective evaluation might measure separate dimensions of employee performance. The knowledge of unrelated information to the activity that is evaluated can influence subjective judgements about the activity that is evaluated (Bol & Smith, 2011). Bol and Smith (2011) found that supervisors’ subjective evaluations were significantly higher when the level of the independent objective measure was high. They also found that the supervisors’ subjective evaluations were significantly lower when the level of the independent objective measure was relatively low. This is called the spillover effect. Bol and Smith (2011) explained that the spillover effect arises because supervisors unintentionally bias their subjective evaluations to be consistent with the known level of performance. This is called cognitive distortion. A supervisor who knows that the employee performed well on an activity, is likely to interpret information about the employee’s performance on another activity more favourably than when the employee would have performed badly. Even if this activity is totally independent from the one measured before. Bol and Smith (2011) found this spillover effect by conducting an experiment. In this experiment participants had to evaluate only one employee. However, many supervisors evaluate more than one employee. When multiple employees are evaluated, employees will compare themselves to other employees (Bol & Smith, 2011). Employees get angry or get distressed when they notice that a peer gets a higher reward for a similar amount of work or when a peer gets the same reward for a lower amount of work. This may lead to reduced effort and confrontations (Bol, Kramer & Maas, 2016). Supervisors might feel more need to justify their decisions to reduce confrontations when they evaluate multiple employees. This leads to more reasoned thinking (Bazerman, Tenbrunsel & Wade-Benzoni, 1998). Bohnet, van Geen and Bazerman (2015) state that more reasoned decision making takes place when

(6)

6

supervisors can compare the ones they have to evaluate and that more intuitive decision making takes place when they cannot compare the one they have to evaluate. This

comparative element might influence the spillover effect. The research question of this paper is therefore: How does the amount of evaluated employees affect the spillover effect? As mentioned before, Bol and Smith (2011) state that the spillover effect exists. However, a limitation of this literature is that it only investigates the spillover effect at the individual level. It might be that the result found by Bol and Smith (2011) cannot be applied to a situation in which multiple employees are evaluated. This research contributes to the existing literature, because it also investigates the spillover effects at firms where an employer evaluates multiple employees.

It might be helpful to answer the research question, because it might be that the result found by Bol and Smith (2011) is not generalizable to a situation in which multiple employees are evaluated. The spillover effect is a problem, because the influence of the objective

measure on the subjective evaluation might limit the role of the subjective evaluation in creating a more complete view about employee performance. It might be helpful to know if a difference exists between the situation in which one employee is evaluated and the situation in which multiple employees are evaluated, because if it turns out that the spillover effect is not that strong when evaluating multiple employees the result found by Bol and Smith (2011) might be less of a problem than expected.

The results of this paper are relevant for the firm owner, because it is in their best interests that the performance reviews are right. They are the ones who are responsible for making sure that managers give accurate evaluations. If the spillover effect does not exist the firm owners do not have to take actions to minimize the spillover effect and can trust that in this regard the evaluator evaluates his employees right.

A scenario-based experiment is used to find an answer on the research question. Participants are asked to subjectively evaluate one or two managers. The objective score of another activity is given to them and information is provided on which the participants have to base their subjective evaluation.

The results indicate that the spillover indeed exists when one employee is evaluated. However, the results do not show that the spillover exists when two employees are evaluated. This might indicate that the results which are found in earlier literature cannot be generalized. The results also do not show that a difference exists between the situation where only one employee is evaluated and the situation were multiple employees are evaluated. This is

(7)

7

inconsistent because a spillover effect does exist when one employee is evaluated, but not when multiple employees are evaluated.

This thesis is structured as follows. In the first part of the paper the theoretical constructs are explained. There is described what objective and subjective performance

evaluations are, what the spillover effect is and what the difference is between evaluating only one employee and multiple employees. After that the hypothesis development is discussed. Subsequently the methodology is described and after that the results are given.

(8)

8

2 Objective and subjective performance evaluation

This paragraph explains what objective and subjective performance evaluations are. With subjective performance evaluations the supervisor makes a judgment about whether the results reflect a good or bad performance. This is instead of using formal, numerical calculations to determine the performance, which is used with objective performance evaluations (Merchant & van der Stede, 2012).

Subjective performance evaluations can evaluate actions and efforts that objective measures cannot evaluate. Subjective evaluations can therefore be used to reduce risk to employees and improve the incentive alignment of the firm’s performance system (Baker et al. 1993). When objective measures are used, it is impossible to correct for bad luck or good luck. Subjective performance evaluations give the opportunity to the employer to use their knowledge of the situation faced by the employee in evaluation how good the employee performed in a given period (Merchant & van der Stede, 2012).

However, subjectivity in evaluations can also increase the risks for employees. The risk increases when the performance evaluations are unfair, inconsistent, or biased (Merchant & van der Stede, 2012). Subjective performance evaluations cannot be verified by a third party, whereas objective measures can be verified. This causes that supervisors can be

influenced by their own preferences when subjectively evaluating performance (Bol & Smith, 2011). The spillover effect which will be explained in the next paragraph and which is the main focus in this research shows that subjective evaluations are indeed not perfect and can be biased.

(9)

9

3 Spillover effect

In this paragraph is explained what the spillover effect is.

Employees judge if their pay is fair by looking at two different things. First, they determine whether their pay is determined in a fair way. So, is the procedure used to

determine which pay they get fair? Second, they judge on whether their pay is fair compared to others. So, does someone who exerts the same amount of effort get the same pay?

(Hartmann & Slapničar, 2012). The spillover effect, which will be explained in this paragraph is an example of a situation where the procedure used to determine the pay is not right. So if the spillover effect exists employees will probably not think that their pay is fair. The other reason for perceived unfairness, which is the perceived unfairness after comparison, will be discussed in the next paragraph.

The spillover effect means that an existing objective measure of an independent activity influences the subjective evaluation of another activity. So even though the objective and subjective evaluation measure separate dimensions of employee performance, they can influence each other (Bol & Smith, 2011). The procedure followed to determine the reward is not fair in this case, because the performance that has to be measured is adjusted with

information that should not be included in the evaluation. Many situations exist where an objective performance measure is known before the subjective evaluation takes place. For example, it is likely that a supervisor observes an employee’s sales performance in a period before subjectively determining their leadership or teamwork skills. If the objective measure of the employees sales performance affects the subjective evaluation of for example the employees’ leadership or teamwork skills, the spillover effect exists (Bol & Smith, 2011). Bol and Smith (2011) found that supervisors’ subjective evaluations were indeed significantly higher when the level of the manager’s objective measure on previous activities was relatively high. They also found that the supervisors’ subjective evaluations were

significantly lower when the level of the manager’s objective measure was relatively low. It is possible that supervisors are influenced by their own preferences when evaluating performance, as subjective performance evaluations cannot be verified by third parties. However, even in the absence of preferences the evaluation of supervisors can be biased. When the spillover effect exists supervisors unintentionally bias their subjective evaluations to be consistent with the known level of performance on the objective measure. This is called cognitive information distortion. A supervisor who knows that the employee performed well on an activity, is likely to interpret information about the employee’s performance on another

(10)

10

activity more favourably than when the employee would have performed badly. Even if this activity is totally independent from the one measured before (Bol and Smith, 2011).

Subjective evaluations are used because they can evaluate actions that objective measures cannot evaluate. The disadvantage is that these evaluations can be influenced by cognitive distortion and that the procedures used to determine which pay someone gets are not fair. This is harmful for an organization, because a fair procedure has several benefits for an organization. First of all, a fairly determined pay leads to increased organizational citizen behaviour (Moorman, Neihoff & Organ, 1993; Naumann & Bennet, 2000). Lind (1995) and Tyler and Dawes (1993) found that it also causes people to become more willing to accept the decisions of authorities. Other studies found that employees are more obedient to the rules (Kim & Mauborgne, 1993) and that they invest more of their personal resources in the organization (Sapeinze & Korsgaard, 1996). It is also found that a fairly determined pay decreases employee theft (Greenberg, 1990; Greenberg & Scott, 1996), workplace assaults (Folger, Robinson, Dietz, McLean-Parks & Baron, 1998), suing of co-workers (Sitkin & Bies, 1993) and employee withdrawal (Colquitt, Conlon, Wesson, Porter & NG, 2001).

There are some possible solutions for the spillover effect. In some settings it may be possible to subjectively evaluate someone before observing the objective measures of other tasks. This adjustment in timing will eliminate the risks of the cognitive distortion. Another possibility is to train evaluators to avoid the influences of the objective measure (Bol & Smith, 2011). So making the evaluators aware of their cognitive distortion. People can only correct their bias if they are aware of their bias (McPherson Frantz, 2006).

In summary, supervisors unintentionally bias their subjective evaluations to be consistent with the known level of performance on the objective measure. A supervisor who knows that the employee performed well (badly) on an activity, is likely to interpret

information about the employee’s performance on another independent activity more (less) favourably than when the employee would have performed badly (Bol & Smith, 2011). So the expectation is that the subjective evaluation is higher (lower) when the objective performance measure about a previous activity is high (low). Therefore the first hypothesis is:

H1: When evaluating one employee the subjective evaluation will be higher (lower) when the objective performance measure of an unrelated previous activity is higher (lower).

(11)

11

4 Spillover effect when evaluating multiple employees

In this paragraph the comparative element is explained, which exists when multiple employees are evaluated. There is also explained what influence this comparative element might have on the spillover effect.

As mentioned before, employees determine if their reward is fair by evaluating whether the procedure used to determine the pay is fair and by comparing their pay to others. The benefits of a fair procedure which are mentioned in the previous paragraph also exist when multiple employees work together. If employees work in teams additional benefits of a fair procedure are that employees will be more cooperative (Sinclair, 2003) and helpful towards others (Naumann and Bennet, 2000). However, when multiple employees are evaluated they will be able to compare themselves to others. Therefore it will also become important that pay is fairly distributed among employees when multiple employees are evaluated (Hartmann & Slapničar, 2012).

The reason for why pay is not always divided in a fair way between employees is because the ones who evaluate often consider their own personal costs and benefits when making subjective evaluation decisions instead of considering firm value. This can happen in organizations because most supervisors who evaluate employees are middle managers whose interests are not perfectly aligned with those of the firm’s owners (Bol et al., 2016). Bol et al., (2016) state that it is in the supervisor’s best interest to not set the rewards lower than what the employees think they deserve. This is because these rewards will probably result in confrontations. Most of the employees do not see themselves as below average (Alicke & Goverun, 2005). Managers therefore give higher rewards than the ones that are below average deserve, which leads to performance evaluation compression (Bol et al., 2016).

The extent of outcome transparency affects the personal costs and benefits for the supervisor (Bol et al., 2016). Outcome transparency is the extent to which employees can observe the rewards given to their peers. Bol et al., (2016) state that performance evaluation compression will be more likely with low outcome transparency. When employees do not know what the rewards of their peers are, supervisors can prevent themselves from confrontations by giving everybody the rewards that they expect.

However, when outcome transparency is high, so when employees can observe each other’s rewards, satisfaction of their own rewards will also rely on how much their peers get (Bol et al., 2016). The judgment of this comparison affects their perception of fairness. Employees will react negatively towards the compensation system and will not get motivated

(12)

12

by it when they notice that a peer gets the same reward for a lower amount of work or when a peer gets a higher reward for the same amount of work (Bol, 2011). They will get angry and distressed due to the perceived unfairness, which will lead to confrontations and reduced effort (Bol et al., 2016). When working in teams, this perceived unfairness is also a contributor to conflict among team members (DeMatteo et al., 1998). The confrontations increase the personal costs for the supervisors. So when outcome transparency is high, performance evaluation compression will lead to more personal costs for the supervisor (Bol et al., 2016).

In table 1 an overview is given about the evaluations of satisfaction and fairness when an employee’s reward is less than their peer’s reward, equal to their peer’s reward or higher than their peer’s reward. This data is from the research of Austin, McGinn and Susmilch (1979).

Employees are more tended to accept the distributions in their rewards if the procedures to determine their rewards are fair (Cropanzano & Greenberg, 1997). This acceptance might lead to less confrontations, which might indicate that it is even more

important that the procedure used to determine the reward is fair when multiple employees are evaluated. So the existence of the spillover effect might have a larger negative effect when evaluating multiple employees.

However, expected is that there will be an effect of the objective measure on the subjective evaluation if supervisors evaluate multiple employees. McPherson Frantz (2006) states that people are blind to their biases and convinced of their fairness. So expected is that even though someone wants to be fair, the unintentional spillover effect cannot be fully eliminated, because supervisors are not aware of their bias. People can only correct their bias

(13)

13

if they are aware of their bias and motivated to do so. If people are unaware of their bias, they cannot correct for it (McPherson Frantz, 2006). This leads to the second hypothesis:

H2: When evaluating multiple employees the subjective evaluation will be higher (lower) when the objective performance measure of an unrelated previous activity is higher (lower).

(14)

14

5 Separate evaluation versus joint evaluation

As shown in the previous paragraphs, the expectation is that the spillover exists in both the situation where one employee is evaluated and the situation in which multiple employees are evaluated. However, there might be a difference between both situations. This paragraph elaborates on this.

There is a difference between situations in which multiple employees are evaluated simultaneously and can be easily compared and situations in which employees are evaluated one at a time and in isolation. The former is called a joint evaluation and the latter is called a separate evaluation (Hsee, Loewenstein, Blount & Bazerman, 1999).

People are assumed to have two distinct modes of thinking, namely intuitive and automatic or reflective and reasoned. It has been suggested that a lack of comparison information, which is the case at a separate evaluation, leads to more intuitive decision making. In contrast, more reasoned choices are made in a joint evaluation. In a joint evaluation more information about the employee’s relative performance is known, which updates prior biased beliefs (Bohnet et al., 2015).

In a joint evaluation supervisors might feel a higher need to justify their decisions. For example because they feel the need to justify why they gave someone a lower evaluation score than the other. This increased need to justify decisions increases analytical thinking, which leads to more rational thinking rather than intuitive thinking (Bazerman et al., 1998). Bazerman et al., (1998) state that they believe that a separate evaluation will lead to a reduced need to justify choices, which will lead to more intuitive thinking.

The more reasoned thinking in a joint evaluation leads to the expectation that the spillover effect will be smaller when supervisors evaluate multiple employees than when supervisors evaluate only one employee. Supervisors are able to compare the ones they evaluate and they will probably feel the need to justify their decisions if they have to evaluate multiple employees. They might therefore make more reasoned decisions. In the situation where only one employee is evaluated supervisors make more intuitive decisions and might therefore be more influenced by the independent objective score.

This paper will investigate whether the spillover effect is smaller when multiple employees are evaluated than when only one employee is evaluated. In other words, there will be investigated if an independent objective score influences the subjective evaluation less if multiple employees are evaluated than when only one employee is evaluated.

(15)

15

If the expectation is true the subjective evaluation of an employee who works together with others and performed well on a previous activity will be lower than the subjective evaluation of someone who does not work with others and also performed well on the previous activity. Also, the subjective evaluation of an employee who works together with others and scored badly on a previous activity should be higher than the subjective evaluation of someone who does not work with others and also performed bad on a previous activity. The subjective evaluation in the situation where multiple employees are evaluated might be more in line with the actual performance when there are objective performance measures, because it might be that this evaluation is less influenced by the objective performance measure.

The preceding will be illustrated with an example: Employee A and B are evaluated by the supervisor. They both scored good on the activity that will be subjectively measured. However, employee A has no one to compare his reward with, while employee B can compare his reward with other colleagues. Both performed equally bad on a previous independent activity that is measured objectively. When the supervisor gives employee B a bad subjective evaluation, this might lead to more costly confrontations than when he gives employee A a bad subjective evaluation, because employee B can compare his reward to others, and he will probably not understand it. Therefore expected is that employee B will get a higher subjective evaluation than employee A. So the reward of employee B is less

influenced by the bad score on the previous activity.

To determine whether the amount of employees moderates the spillover effect the following hypotheses will be investigated:

H3a: The spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is high.

In other words, H3a indicates that the score of the employee that is evaluated in a joint evaluation is less adjusted upwards due to the high objective score than the score of the employee that is evaluated in a separate evaluation. This means that if H3a is supported the subjective evaluation of the employee in the situation where multiple employees are evaluated is significantly lower than the subjective evaluation of the employee in the situation where only one employee is evaluated.

(16)

16

H3b: The spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is low

In other words, H3b indicates that the score of the employee that is evaluated in a joint evaluation is less adjusted downwards due to the low objective score than the score of the employee that is evaluated in a separate evaluation. This means that if H3b is supported the subjective evaluation of the employee in the situation where multiple employees are evaluated is significantly higher than the subjective evaluation of the employee in the situation where only one employee is evaluated.

(17)

17

6 Hypotheses

This paragraph presents the hypotheses that are based on the previous literature.

Previous literature has found that supervisors unintentionally bias their subjective evaluations to be consistent with the known level of performance on the objective measure (Bol & Smith, 2011). This leads to the first hypothesis:

H1: When evaluating one employee the subjective evaluation will be higher (lower) when the objective performance measure of an unrelated previous activity is higher (lower).

McPherson Frantz (2006) states that people are blind to their biases and convinced of their fairness. So expected is that even though supervisors want to be fair due to the comparative element, the unintentional spillover effect cannot be fully eliminated because supervisors are not aware of their bias. This leads to the second hypothesis:

H2: When evaluating multiple employees the subjective evaluation will be higher (lower) when the objective performance measure of an unrelated previous activity is higher (lower)

In the previous paragraph is suggested that more reasoned decision making exists in joint evaluations and more intuitive decision making exists in separate evaluations. In a joint evaluation more information about the employee’s relative performance is known, which updates prior biased beliefs (Bohnet, van Geen & Bazerman, 2015). Supervisors might also feel a higher need to justify their decisions to reduce confrontations (Bazerman, Tenbrunsel & Wade-Benzoni, 1998). In the situation where only one employee is evaluated supervisors make more intuitive decisions and might therefore be more influenced by the independent objective score. This theory leads to the following hypotheses:

H3a: The spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is high.

H3b: The spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is low.

(18)

18

7 Methodology

In order to test the hypotheses an experiment is conducted. The participants are placed in a hypothetical case setting. In this hypothetical case setting participants must assume that they are the regional director for an industrial pipe and fitting company. The role of the regional director is to oversee district managers who are responsible for two separate activities, namely individual sales and district office administration. These activities are completely independent of each other. The participants will be asked to subjectively evaluate the performance of one district manager or of two district managers, while the manager’s score on an individual sales measure is known.

7.1 Participants

The link with the experiment was sent to as many people as possible. The investigation is a psychological issue, so everyone could participate in the experiment. The control questions exist to check if the participants understood the case. The average age of the participants was 28, ranging from 17 till 65. The educational level of the participants ranged from high school to master. In table 2 till 4 the age distribution, educational level distribution and gender distribution shown.

7.2 Experimental design:

The conducted experiment was a scenario experiment. There were fivecases. Three cases for the situation where only one employee was evaluated, one case for the situation were two employees are evaluated and onecontrol condition were two employees are evaluated. The district managers their scores on the sales measure were manipulated to be one of the two different levels (high and low). In the control condition was no information provided about the objective sales measure. The experimental design of this thesis was based on the research design of Bol and Smith (2011). This research design was provided by the authors.

(19)

19

7.3 Independent variables:

There are three independent variables in the case. The first independent variable is the

district’s manager’s individual sales performance. The second independent variable is the amount of employees that are evaluated in the case. The last independent variable is which employee has to be evaluated.

7.3.1 Individual sales performance

The independent variable in the experiment was the district manager’s individual sales performance. The level of this performance was manipulated to be either high or low. More specifically, the score was manipulated to be 2 (low) or 9 (high) on a 10-point scale.

7.3.2 Amount of evaluated employees

The second independent variable in the case is the amount of employees that are evaluated in the case. Participants are assigned to a case in which they have to evaluate either one or two employees.

7.3.3 Which employee

The participants are assigned to a case in which information is provided about the office administration performance of David or Lou (or both).

7.4 Dependent variables:

The dependent variable is the subjective evaluation of the district manager’s office

administration performance. In the case participants had to read information which included personal notes about the district managers’ administration performance and interviews with manager’s office staff who had observed the district manager’s administration performance. Information was available about the administrative performance of district manager David and about the administrative performance of district manager Lou.

In the two cases were only David was evaluated the provided subjective performance information was the same. In the case were only Lou was evaluated the provided subjective performance information was different from the information that was provided about David. In the case were two employees were evaluated the participants read about the performance of both David and Lou. The information provided about David was here the same as in the case were David was the only one who was evaluated and the provided information about Lou was the same as in the case were only Lou was evaluated.

After reading the information the participants were asked to give a score for the district manager’s administration performance on a scale from 0 to 10. .

(20)

20

7.5 Procedures:

Participants conducted an online experiment between May 6 2017 and June 4 2017 through the distribution of an anonymous URL. The participants were randomly assigned to one of the five cases. The experiment took on average 5 minutes.

In the online experiment participants assumed that they were the regional director of an industrial pipe and fitting company. The case stated that the role of the regional director was to oversee ten district managers who are responsible for two separate activities, namely individual sales and district office administration.

In the cases was described that the individual sales activity and the district office administration were completely independent of each other. So performance on one of the job functions was unrelated to performance on the other job function. To determine the overall performance the individual sales activity and the district office administration were equally weighted. On both activities was scored on a 0-10 scale. The average of these score was taken to determine the overall evaluation score.

To all participants, except for the participants in the control condition, the objective sales score of the managers was shown in a table. In this table the sales performance scores of 10 managers was shown. The objective score(s) of the manager(s) which the participants had to evaluate was in bold.

In table 5 is described what the differences between the five cases are. In the first case participants were asked to give the performance evaluation of David Sutton. David’s objective score was manipulated to be 9 in this case. In the second case participants were also asked to provide the performance evaluation score of David, only now his objective score was 2. In the third case the same was asked for Lou Archibald, who had an objective score of 2. In the case where multiple employees were evaluated the participants were asked to give the performance evaluation score of two of the district managers, namely David and Lou. For the objective evaluation score again a table was given of all ten district managers. The objective evaluation score of David was manipulated to be 9 and the objective score of Lou was manipulated to be

(21)

21

2. In the control condition in which also both David and Lou had to be evaluated no information was given about the objective evaluation score.

For the subjective evaluation the participants received information about the district manager’s office performance. Described was that the performance could not be directly observed so they had to base their evaluation on personal notes made when overseeing the manager and on conducted interviews with members of the staff of the district manager. The personal notes and the conducted interviews were provided in the case. The information about David was the same in all cases and the information about Lou was also the same in all cases. After reading the information the participants were asked to give a score of the district manager’s office administration on a 0-10 scale. In both cases where one district manager was evaluated and the case where two district managers were evaluated, the participants were asked to average the office administration performance score with the individual sales performance score to come up with the overall evaluation. In the control condition

participants were only asked to give the score of the administration performance because no objective score was given.

In the cases where multiple employees were evaluated it was stated that there is high outcome transparency. This because employees can compare themselves to others if

employees know each other’s rewards. So if high outcome transparency exists the cases in which multiple employees are evaluated are most different from the cases where only one employee is rewarded by the supervisor. Also, if low outcome transparency exists, the scores of the employees may be closer to each other, because of performance evaluation

compression. This means that scores are closer to each other because evaluators are afraid of confrontations.

At the end of the experiment some demographic questions were asked and some manipulation checks were done. At this point participants could not change their subjective scores anymore. In all cases, except for the control condition, participants had to recall if the

(22)

22

objective sales score was high or low and whether this sales score was indicative for the office administrative performance. In the cases were multiple employees had to be evaluated an additional control question was whether managers knew the evaluation score that was given to the other. In all cases was also asked whether the provided information about the managers’ office administration performance allowed the participants to make a reasonable judgment about the district manager’s administration performance.

The employees who failed the manipulation checks were excluded from all further analyses. 56 of the participants answered one question wrong. 27 of these participants answered the question wrong in which was asked whether the two employees know each other’s evaluation score. The question in which was asked whether the sales performance is indicative for the office administration performance was answered wrong 28 times. Three of the participants answered the question wrong in which was asked whether the employee that they had to evaluate had a low or high sales performance. Four of the participants answered two questions wrong. Two of these answered both the question wrong in which was asked whether the two employees know each other’s evaluation score and the question in which was asked whether sales performance is indicative for office administration performance. The other two answered the questions wrong in which was asked whether the employees had a high or low sales performance and the question in which was asked whether sales

performance was indicative for office administration performance.

In total there where 177 participants. 60 of them answered a manipulation question wrong. These 60 are excluded from all analyses, so the analyses is done with 117 participants.

(23)

23

8 Results

8.1 Information sufficiency and descriptive statistics:

The participants were asked to indicate on a 5 point scale whether they thought that they were

able to make a reasonable judgement regarding the manager’s administration performance. Participants gave on average a score of 3.7. With a one-sample T-test is found that this score is significantly higher than 3, which is the scale midpoint (p<0.01). This means that

participants felt like they had sufficient information about the manager’s office administration performance to make a reasonable judgment.

To test whether there are no significant differences between the different groups regarding the information sufficiency measure a one-way Anova test is done. This test showed that there are no significant differences between groups regarding this measure (p=0.238).

In table 6 the descriptive statistics are shown. The mean scores of the subjective performance evaluation, standard errors and the number of observations are shown for every case.

8.2 Hypotheses tests:

8.2.1 Hypothesis 1: Spillover effect 1 employee

Hypothesis 1 predicts that an objective score of an activity influences the subjective

(24)

24

activities are independent of each other. To test this hypothesis the mean scores of the condition in which one employee scored high on his sales performance and the condition in which the same employee scored low on his sales performance is analysed. The same

employee is analysed so the provided subjective information is identical in the two compared cases.

To test the first hypothesis, the independent samples test is used. Before using the T-test it is checked whether equal variances could be assumed. Levene’s T-test for equality of variances showed that this could not be assumed (F=22.25; p<0.001). To correct for the violation of this assumption the Welch-Satterthwaite T-test is used, which adjusts the degrees of freedom. The result of this test is shown in table 7. The test showed a significant difference between the score for the employee in the high sales condition (M=6.65; SD=1.05) and the score for the employee in the low sales condition (M=4.67; SD =2.11); (T=3.91; p<0.001). This result supports the first hypothesis, suggesting that the information of an independent objective performance measure influences the subjective evaluation. So even though the performance on the subjective evaluation was exactly the same in both conditions, the score of the employee in the high sales condition was significantly higher than the score of the employee in the low sales condition.

8.2.2 Hypothesis 2: Spillover effect multiple employees

The second hypothesis also predicts that an independent objective score influences the subjective evaluation, only now in the case when multiple employees are evaluated. The two cases in which two employees are evaluated are analysed. So the case in which the objective scores are known is compared with the case in which no information about the sales

performance is given. It is tested whether the average difference between the employee that scored high and the employee who scored low is significantly higher than the average difference between the two employees in the case where no objective score is given. The compared information about the subjective performance information is the same in the two cases.

(25)

25

This hypothesis is also tested with an independent samples T-test. Again, first is tested whether equal variances could be assumed with Levene’s test for equality of variances. This test showed no significant result (F=2.44; p=0.125), indicating that equal variances could be assumed and the independent samples T-test could be used without making adjustments. The result of the independent samples T-test is shown in table 8. The test showed no significant difference between the average difference in the case with the given objective scores

(M=0.28; SD=1.45) and the average difference in the case where no objective score is given (M=0.06; SD=0.84); (T=0.640; p=0.526). The second hypothesis is therefore not supported, suggesting that the information of an unrelated objective performance score does not

influence the subjective evaluation when multiple employees are evaluated.

8.3.3 Hypothesis 3: Amount of evaluated employees as moderator

The third hypothesis predicts that an independent objective score has more influence on the subjective evaluation when one employee is evaluated than when multiple employees are evaluated. This hypothesis is subdivided in two hypotheses.

Hypothesis 3a predicts that the spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is high. Predicted is that for the situation in which a high objective score is given, the subjective evaluation score in the case where only one employee is evaluated is higher than the score of the employee who is evaluated in a joint evaluation. The cases in which the employees had a high objective score are compared. The two cases which are compared are the cases in which the provided subjective information is identical. So the score of the employee in the case in which one employee is evaluated is compared with the score of that same employee only than in the case in which he is evaluated together with another employee. To test the hypothesis the mean scores of the two cases are analyzed to check whether a significant difference exists. Hypothesis 3b predicts that the spillover effect is stronger when one employee is evaluated than when multiple employees are evaluated when the objective measure is low. Predicted is that for the situation in which a low objective score is given, the subjective evaluation in the case where only one employee is evaluated is lower than the score of the

(26)

26

employee who is evaluated in a joint evaluation. The cases in which the employees had a low objective score are compared. Again, the two cases that are compared are the cases in which the provided subjective information is identical. So the score of the employee in the case in which one employee is evaluated is compared with the score of that same employee only than in the case in which he is evaluated together with another employee. To test the hypothesis the mean score of these two cases are analyzed to check whether a significant difference exists. Both hypotheses are tested by using an independent samples T-test. First is tested whether equality of variances could be assumed. Levene’s test of equality showed that equality of variances could be assumed for both the cases tested in hypothesis 3a (F=0.13; p=0.717) and the cases tested in hypothesis 3b (F=0.01; p=0.25). Therefore the independent samples T-test could be used without making adjustments.

The independent samples T-test showed no significant differences for hypothesis 3a. There is no difference found between the employee’s score in the case where only one employee is evaluated (M=6.65; SD=1.05) and the employee’s score in the case where both employees are evaluated (M=6.63; SD=1.09); (T=0.049; p=0.961). Hypothesis 3a is therefore not supported, indicating that it does not matter whether the supervisor evaluates one or multiple employees when the independent objective score is high.

The test also showed no significant differences for hypothesis 3b. No significant difference is found between the employee’s score in the case where only he is evaluated (M=5.77; SD=1.41) and the employee’s score in the case where both managers are evaluated (M=6.35; SD=1.20); (T=-1.51; p=0.139). Hypothesis 3b is therefore also not supported, which indicates that it does not matter whether the supervisor evaluates one or multiple employees when the independent objective score is low. The results of both tests are in table 9.

8.4 Analysis of results

In line with Bol and Smith (2011) the results show that an independent objective measure influences the subjective evaluation when one employee is evaluated. A limitation of the

(27)

27

study of Bol and Smith (2011) is that they do not consider the spillover effect at the group level. This paper shows that the spillover effect does not exist when multiple employees are evaluated. This indicates that the spillover effect found by Bol and Smith (2011) cannot be generalized to a setting in which multiple employees are evaluated.

However, in this paper it was expected that the spillover effect does exist when multiple employees are evaluated because people are blind to their biases and convinced of their fairness (McPherson Frantz, 2006). A reason why this expectation does not hold may be that the increased rationalized decision making in a joint evaluation is able to compensate for the bias. More information about the employee’s relative performance might update prior biased beliefs (Bohnet et al., 2015).

This paper also found no significant difference in the spillover effect between the situation in which one employee is evaluated and the situation in which multiple employees are evaluated. This result is against expectation and inconsistent, since there does exist a spillover effect in the case of one evaluated employee and not in the case of two evaluated employees.

Additional analyses are required to test how this discrepancy arises. In table 10 is shown which mean scores were used for testing the hypotheses. The table points out that 1Dl and 1Ll are only used once. 1Dl is only used for the first hypothesis and 1Ll is only used for the third hypothesis. These are the hypotheses which show a discrepancy. The fact that these scores are only used once explains why the results of hypothesis 1 and 3 are inconsistent, because an (untabulated) independent samples T-test shows that a significant difference exists between 1Dl and 1Ll (T=2.31; p=0.026). This is not as expected since in both cases the employees had exactly the same objective score and it can also be stated that the provided subjective information contained the same message. Namely, an (untabulated) independent samples T-test does not find a significant difference between 2DC and 2LC (T=0.296;

(28)

28

p=0.768). In these cases the employees were solely evaluated based on the provided

subjective information, no objective score was shown. Therefore it can be assumed that the provided subjective information contained the same message. The only possible explanation for why a significant difference exists between 1Dl and 1Ll and no significant difference exists between 2DC and 2LC is that the latter is a joint evaluation instead of a separate evaluation. So it may be that the scores of 2DC and 2LC are closer to each other due to performance evaluation compression, which means that managers give higher scores than the ones that below average deserve to avoid confrontations. But for now it cannot be fully explained why a significant difference exists between 1Dl and 1Ll.

For the investigation in this paper it would have been better if more different cases were available. For hypothesis 1 is now only the score of 1Dl compared with the score of 1Dh. It would have been useful if a case of 1Lh was available so it could also be investigated whether the spillover effect exists if 1Ll and 1Lh are compared. Also, for clarification of hypothesis 3 cases of 2Lh and 2Dl would have been useful. For hypothesis 3a is in this paper 2Dh is compared with 1Dh. For this hypothesis it would have been useful to know the score of 2Lh to be able to compare 2Lh with 1Lh, which should as explained above also be added to clarify hypothesis 1. If 2Lh and 1Lh are compared there could also be investigated whether a significant difference exists between the situation in which one employee is evaluated and the situation in which multiple employees are evaluated when the objective measure is high. For hypothesis 3b this paper compared 2Ll with 1Ll. For this hypothesis it would have been useful to know the score of 2Dl to be able to compare 2Dl with 1Dl. With this test there could be also be investigated whether a significant difference exists between the situation in which one employee is evaluated and the situation in which multiple employees are evaluated when the objective score is low.

The two new cases (2Lh and 2Dl) can also be used to strengthen hypothesis 2. For hypothesis 2 the difference between the cases of 2Dh and 2L was compared with the

difference between 2Dc and 2Lc. With the cases 2Lh and 2Dl there can also be tested whether the spillover effect exists when multiple employees are evaluated by comparing the difference between 2Lh and 2Dl with the difference between 2Dc and 2Lc. At last, control condition cases for the situation in which one employee is evaluated would have been useful (1Dc and 1Lc). In these cases participants evaluate only one employee based solely on subjective information. With these cases it can be checked whether it can be assumed that the provided subjective information is equal. With more available tests conclusions can be drawn with more certainty. In this paper the current available data and tests are used to draw conclusions.

(29)

29

9 Conclusion

The aim of this study was to investigate how the amount of evaluated employees affects the spillover effect. It is investigated whether an independent objective performance score has a different effect on the subjective evaluation when multiple employees are evaluated than when only one employee is evaluated.

Literature has shown that companies use subjective evaluations because they give a more complete view about the performance of an employee. However, subjective evaluations cannot be verified by a third party. This is a problem, because supervisors can be biased and give unfair subjective evaluations. An example of a situation where the supervisor is biased and gives the wrong evaluation is a situation in which the spillover effect exists. If the spillover effect exists the subjective evaluation is not determined in a fair way. The spillover effect means that an already existing objective measure of an

independent activity influences the subjective evaluation of another activity. This is caused by a cognitive distortion, which means that supervisors unintentionally bias their subjective evaluations to be consistent with the known level of performance on the activity that is objectively measured.

This paper builds upon the research of Bol and Smith (2011). However, a limitation of their study is that they only investigated the spillover effect at the individual level. It may be that this effect is different at the group level. In reality most supervisors evaluate multiple employees, so it is useful to know if the spillover effect found by Bol and Smith (2011) is generalizable to a situation in which multiple employees are evaluated. It is expected that the spillover effect is different in a situation in which multiple employees are evaluated, because supervisors think more reasoned when they evaluate multiple employees instead of only one employee.

This paper will again investigate whether the spillover effect exists when one employee is evaluated. There will also be investigated whether the spillover effects exists when multiple employees are evaluated and there will be investigated whether a difference in the spillover effect exists between a situation in which one employee is evaluated and a situation in which multiple employees are evaluated

This paper found that supervisor’s subjective evaluations were influenced by the independent objective measure when one employee is evaluated. This is in line with the research done by Bol and Smith (2011). However, it is not found that the supervisor’s subjective evaluation is influenced by the objective measure when multiple employees are

(30)

30

evaluated. This might indicate that the results found by Bol and Smith (2011) are not

generalizable to a situation in which multiple employees are evaluated. This paper did expect that the spillover also exists when multiple employees are evaluated, because people are blind to their biases. That the results show that the spillover does not exist when evaluating multiple employees may be due to increased rationalized thinking when employees can be compared. More information about the employee’s relative performance might update prior biased beliefs.

This paper also investigated whether a difference exists in the spillover effect between a situation in which one employee is evaluated and a situation in which multiple employees are evaluated. Expected is that the spillover effect is smaller when multiple employees are evaluated because managers think more reasoned when they compare multiple employees. Supervisors might feel a higher need to justify their decisions in order to avoid confrontations when multiple employees are evaluated, which leads to more reasoned thinking. If supervisors evaluate only one employee they will make more intuitive decisions. In a joint evaluation employee’s prior biased beliefs can be updated due to more available information about the employee’s relative performance which might reduce the spillover effect.

This paper however did not find a significant difference between the situation in which only one employee is evaluated and the situation in which multiple employees are evaluated. This is inconsistent because it is shown that the spillover effect does exist when one employee is evaluated and not when multiple employees are evaluated. So there should be a difference between those two situations. That no significant difference is found is because the scores of different employees are used to test hypothesis 1 and 3. It was expected that the scores of these two employees would not differ significantly since their scores in the control condition did not differ significantly and the given objective scores were the same. However, a test pointed out that the scores did differ significantly.

The results of this paper are relevant, because many supervisors evaluate multiple employees. This study found that the spillover effect does not exist when multiple employees are evaluated. This means that there may be no spillover effect in practice. So the spillover effect might be less of a problem than that is expected by Bol and Smith (2011). This result is relevant for the firm’s owners, because it is in their best interestthat employees are rewarded in a way that enhances firm value. So it is in the firm owner’s best interestthat the subjective evaluation is accurate. The firm owners are responsible for making sure that managers give accurate evaluations. They should take actions if the threat of the spillover effect exists. If the spillover effect does not exist the firm owners do not have to take actions to minimize this

(31)

31

effect and can trust that in this regard the evaluator evaluates his employees accurately. This study found that the spillover effect does not exist when multiple employees are evaluated. Since most managers evaluate multiple employees, most managers will save time and money because they do not have to take actions to minimize the spillover effect. Actions to minimize the spillover effect are to subjectively evaluate someone before observing the objective measure or to train evaluators to avoid the influences of the objective measure. So if multiple employees are evaluated firm owners do not have to spent time and money on organizing trainings to make employees more aware of the spillover effect. They also do not have to spent time and money on finding solutions on how they can achieve that a supervisor evaluates an employee without knowledge about previous performance of that employee. However, this paper has some limitations. First of all, the group of participants was small. There were 20-25 participants per case. This number might be too small to make the right conclusions because the standard errors are large. Second, it is not tested whether people who have work experience in giving subjective evaluations give different scores than people that never evaluated employees. It might be that they are less susceptible to the spillover effect due to their experience. At last, the experiment in this paper would have been better if more cases were available. With more available tests conclusions can be drawn with more certainty. The unexpected difference between the two different employees in the low control condition could have been analyzed in a better way if more information was available. These limitations lead to some suggestions for further research. First of all, this paper should be repeated with more participants to see whether there are really no significant results for the third hypothesis when the sample size is bigger. Second, further research should test whether work experience as evaluator moderates the spillover effect. It should be tested if people with more work experience as evaluator are less influenced by the objective measure. At last, the tests in this paper should be done with more different cases. With more cases the unexpected difference between the two different employees in the low control condition could be better analyzed and conclusions could be drawn with more certainty.

(32)

32

References

Alicke, M. D., & Govorun, O. (2005). The better-than-average effect. The self in social

judgment, 1, 85-106.

Austin, W., McGinn, N. C., & Susmilch, C. (1980). Internal standards revisited: Effects of

social comparisons and expectancies on judgments of fairness and

satisfaction. Journal of Experimental Social Psychology, 16(5), 426-441.

Baker, G., Gibbons, R., & Murphy, K. J. (1993). Subjective performance measures in optimal

incentive contracts (No. w4480). National Bureau of Economic Research.

Bazerman, M. H., Tenbrunsel, A. E., & Wade-Benzoni, K. (1998). Negotiating with yourself

and losing: Making decisions with competing internal preferences. Academy of

Management Review, 23(2), 225-241.

Bohnet, I., Van Geen, A., & Bazerman, M. (2015). When Performance Trumps Gender Bias:

Joint vs. Separate Evaluation. Management Science, 62(5), 1225-1234.

Bol, J. C., & Smith, S. D. (2011). Spillover effects in subjective performance evaluation: Bias

and the asymmetric influence of controllability. The Accounting Review, 86(4), 1213-

1230.

Bol, J. C., Kramer, S., & Maas, V. S. (2016). How control system design affects performance

evaluation compression: The role of information accuracy and outcome

transparency. Accounting, Organizations and Society, 51, 64-73.

Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O., & Ng, K. Y. (2001). Justice at the

millennium: a meta-analytic review of 25 years of organizational justice research.

Cropanzano, R., & Greenberg, J. (1997). Progress in organizational justice: Tunneling

through the maze. International review of industrial and organizational

psychology, 12, 317-372.

(33)

33

evidence. Research in organizational behavior, 20, 141-183.

Folger, R., Robinson, S. L., Dietz, J., Parks, J.M., & Baron, R.A. (1998, August). When

Colleagues Become Violent: Employee Threats And Assaults As A Function Of

Societal Violence And Organizational Injustice. In Academy of Management

Proceedings (Vol. 1998, No. 1, pp. A1-A7). Academy of Management.

Greenberg, J. (1990). Organizational justice: Yesterday, today, and tomorrow. Journal of

management, 16(2), 399-432.

Greenberg, J., & Scott, K. S. (1996). Why do workers bite the hands that feed them?

Employee theft as a social exchange process. Research in organizational behavior, 18,

111-156

Hartmann, F., & Slapničar, S. (2012). Pay fairness and intrinsic motivation: the role of pay

transparency. The International journal of human resource management, 23(20),

4283-4300.

Hsee, C. K., Loewenstein, G. F., Blount, S., & Bazerman, M. H. (1999). Preference reversals

between joint and separate evaluations of options: A review and theoretical

analysis. Psychological bulletin, 125(5), 576.

Kim, W. C., & Mauborgne, R. A. (1993). Procedural justice, attitudes, and subsidiary top

management compliance with multinationals' corporate strategic decisions. Academy

of management journal, 36(3), 502-526.

Lind, E. A. (1995). Justice and authority relations in organizations. Organizational politics,

justice, and support: Managing the social climate of the workplace, 83-96.

Merchant, A. & Van der Stede, A. (2012). Management Control Systems. Pearson Education.

McPherson Frantz, C. (2006). I AM being fair: The bias blind spot as a stumbling block to

seeing both sides. Basic and Applied Social Psychology, 28(2), 157-167.

(34)

34

organizational citizenship behavior: Sorting the effects of job satisfaction,

organizational commitment, and procedural justice. Employee responsibilities and

rights journal, 6(3), 209-225.

Naumann, S. E., & Bennett, N. (2000). A case for procedural justice climate: Development

and test of a multilevel model. Academy of Management journal, 43(5), 881-889.

Sinclair, A. L. (2003). The effects of justice and cooperation on team

effectiveness. Small Group Research, 34(1), 74-100.

Sapienza, H. J., & Korsgaard, M. A. (1996). Procedural justice in entrepreneur-investor

relations. Academy of management Journal, 39(3), 544-574.

Sitkin, S. B., & Bies, R. J. (1993). Social accounts in conflict situations: Using explanations

to manage conflict. Human Relations, 46(3), 349-370.

Tyler, T. R., & Dawes, R. M. (1993). Fairness in groups: Comparing the self-interest and

social identity perspectives. Psychological perspectives on justice: Theory and

(35)

35

Appendix

[This shows the experimental materials for the condition in which both David and Lou are

evaluated with a known objective score. However, all manipulations are reflected in this version of the materials. Dash-lined boxes indicate what participants in the control condition did not see. In the cases where only David is evaluated, all provided information below about Lou is not shown. In the cases where only Lou is evaluated, all provided information below about David is not shown. In addition, some additional information about what was different in the cases where only one employee is evaluated is presented between brackets]

Welcome to this study!

In the first part of this study you will read about a hypothetical case. After that you will be asked to answer some questions based on this case. Please read it carefully, because in the end there will be some questions to check whether you understood the case and once you go to the next page, you can't go back.

Your answers will be anonymous and will only be used for this study.

Thank you for participating in this study!

SCREEN BREAK Assume the following hypothetical case:

You are a regional director at Dexter Pipe Inc., a company that manufactures and sells

industrial pipes and fittings. There are ten district managers in your region, and they all report to you.

A compressed organizational chart for your region is shown below.

SCREEN BREAK

Regional Director

Sales Managers

Sales Associates Administrative Staff

(36)

36

District managers have separate responsibilities in two main areas: (1) sales and (2) office administration.

1. Sales. District managers are expected to devote about 50% of their time to generating individual sales.

2. Office administration. District managers are expected to devote the other 50% of their time to managing the district office, which includes processing sales-related

paperwork, record-keeping and administration. Duties such as monitoring budgets, filing reports and other documents, and managing the administrative staff fall under this responsibility.

Each district manager’s overall performance evaluation is made up of two separate measures: an objective measure based on sales and a subjective measure based on office administration performance (which you evaluate).

The company has decided that two measures are necessary because district managers’ two main job functions (sales and office administration) are completely independent of each other. That is, performance on one dimension is not indicative of performance on the other

dimension.

The two measures are weighted equally in the evaluation system. Each measure is a score on a scale from 0 to 10, and the measures are averaged together to compute the overall score. Thus, the overall evaluation is a score that will be in the range of 0 to 10.

Your task in this case is to conduct the performance evaluation for David Sutton and Lou

Archibald, two of the sales managers in your region. David and Lou can observe the reward

that is given to the other based on the performance evaluation and they also know how the other performed.

[Conditions where only one employee is evaluated: It is not told that David and Lou can observe each other’s reward]

SCREEN BREAK

The first measure of the overall evaluation is a score based on the manager’s individual sales during the six-month period. A manager receives one point for each $50,000 of sales he or she generates. The table below shows the sales score for the ten district managers in your region over the most recent six-month period. Each of the ten districts contains about the same number of businesses and potential customers

(37)

37

You can see that David Sutton’s individual sales score for the most recent six-month period is 9 and that Lou Archibald’s individual sales score for the most recent six-month period is 2. [In the condition in which David had an objective score of 2: In the table the name of Lou is replaced by the name of David. For all cases in which one employee is evaluated only the name of the one that is evaluated is in bold]

Market demand was stable during the period. There were no unforeseen or uncontrollable events that impacted David’s or Lou’s sales.

SCREEN BREAK

The second measure of the overall evaluations of David and Lou is a score based on your personal evaluations of the district managers’ office administration performances.

Based on the interactions you have had with David and Lou, you have compiled some notes over the past six months. Those notes are provided below. However, as regional director you are not able to directly observe your district managers on a day-to-day basis. Because of that, you have also conducted interviews with some members of David and Lou’s administrative staff. Their answers to your questions regarding David’s and Lou’s office administration performance are also provided below.

David’s office administration performance:

Personal Notes regarding David Sutton, Division C District Manager

Wasn’t on time to the office the first morning I stopped by…. Always has a big smile, staff seem to really like him. Budget estimates are pretty accurate. Some minor issues with the completeness of other reports, but generally OK.

Interview Responses, Division C Sales and Administrative Staff

Question: Can you please comment on David’s office administration performance over the

The spillover effect at the group level

Thesis