hr chapter 8 – Flashcards

The Process of Performance Management Although many employees have come to dread the annual "performance appraisal" meeting, at which a boss picks apart the employee's behaviors and apparent attitudes from the past year, performance management can potentially deliver many benefits. Effective performance management can tell top performers that they are valued, encourage communication between managers and their employees, establish uniform standards for evaluating employees, and help the organization identify its strongest and weakest performers. ex Consultant Dick Grote asserts that performance appraisals, properly done, meet an "ethical obligation of leadership" by providing information that all members of an organization want to know so they can succeed: "What is it you expect of me? How am I doing at meeting your expectations?" 2 To meet these objectives, performance management includes several activities. As shown in Figure 8.1 , • these are defining performance, • measuring performance, and feeding back performance information. First, the organization specifies which aspects of performance are relevant to the organization. These decisions are based on the job analysis, described in Chapter 4. Next, the organization measures the relevant aspects of performance by conducting performance appraisals. Finally, through performance feedback sessions, managers give employees information about their performance so they can adjust their behavior to meet the organization's goals. When there are performance problems, the feedback session should include efforts to identify and resolve the underlying problems. In addition, performance feedback can come through the organization's rewards, as described in Chapter 12. Using this performance management process helps managers and employees focus on the organization's goals. Computer software and Internet-based performance management systems are available to help managers at various stages of the performance management process. Software can help managers customize performance measurement forms. The manager uses the software to establish a set of performance standards for each job. The manager rates each employee according to the predetermined standards, and the software provides a report that compares the employee's performance to the standards and identifies the employee's strengths and weaknesses. Other software offers help with diagnosing performance problems. This type of software asks questions—for example, Does the employee work under time pressure? The answers suggest reasons for performance problems and ways the manager can help the employee improve

Purposes of Performance Management Organizations establish performance management systems to meet three broad purposes: strategic, administrative, and developmental. Strategic purpose means effective performance management helps the organization achieve its business objectives. It does this by helping to link employees' behavior with the organization's goals. Performance management starts with defining what the organization expects from each employee. It measures each employee's performance to identify where those expectations are and are not being met. This enables the organization to take corrective action, such as training, incentives, or discipline. Performance management can achieve its strategic purpose only when measurements are truly linked to the organization's goals and when the goals and feedback about performance are communicated to employees ex . Just Born, the company that makes Peeps and Mike and Ike candy, meets the strategic purpose of performance management. Its system has employees and managers meet to agree on several personal objectives through which each employee will help meet the objectives of his or her department. Together, they identify whatever training the employee needs and meet regularly to discuss the employee's progress in meeting the objectives. 3 The administrative purpose of a performance management system refers to the ways in which organizations use the system to provide information for day-to-day decisions about salary, benefits, and recognition programs. Performance management can also support decision making related to employee retention, termination for poor behavior, and hiring or layoffs. Because performance management supports these administrative decisions, the information in a performance appraisal can have a great impact on the future of individual employees. Managers recognize this, which is the reason they may feel uncomfortable conducting performance appraisals when the appraisal information is negative and, therefore, likely to lead to a layoff, disappointing pay increase, or other negative outcome. Finally, performance management has a developmental purpose, meaning that it serves as a basis for developing employees' knowledge and skills. Even employees who are meeting expectations can become more valuable when they hear and discuss performance feedback. Effective performance feedback makes employees aware of their strengths and of the areas in which they can improve. Discussing areas in which employees fall short can help the employees and their manager uncover the source of problems and identify steps for improvement. Although discussing weaknesses may feel uncomfortable, it is necessary when performance management has a developmental purpose.

Criteria for Effective Performance Management In Chapter 6, we saw that there are many ways to predict performance of a job candidate. Similarly, there are many ways to measure the performance of an employee. For performance management to achieve its goals, its methods for measuring performance must be good. Selecting these measures is a critical part of planning a performance management system. Several criteria determine the effectiveness of performance measures: • Fit with strategy —A performance management system should aim at achieving employee behavior and attitudes that support the organization's strategy, goals, and culture. If a company emphasizes customer service, then its performance management system should define the kinds of behavior that contribute to good customer service. Performance appraisals should measure whether employees are engaging in those behaviors. Feedback should help employees improve in those areas. When an organization's strategy changes, human resource personnel should help managers assess how the performance management system should change to serve the new strategy. • Validity —As we discussed in Chapter 6, validity is the extent to which a measurement tool actually measures what it is intended to measure. In the case of performance appraisal, validity refers to whether the appraisal measures all the relevant aspects of performance and omits irrelevant aspects of performance. Figure 8.2 shows two sets of information. The circle on the left represents all the information in a performance appraisal; the circle on the right represents all relevant measures of job performance. The overlap of the circles contains the valid information. Information that is gathered but irrelevant is "contamination." Comparing salespeople based on how many calls they make to customers could be a contaminated measure. Making a lot of calls does not necessarily improve sales or customer satisfaction, unless every salesperson makes only well-planned calls. Information that is not gathered but is relevant represents a deficiency of the performance measure. For example, suppose a company measures whether employees have good attendance records but not whether they work efficiently. This limited performance appraisal is unlikely to provide a full picture of employees' contribution to the company. Performance measures should minimize both contamination and deficiency. • Reliability —With regard to a performance measure, reliability describes the consistency of the results that the performance measure will deliver. Interrater reliability is consistency of results when more than one person measures performance. Simply asking a supervisor to rate an employee's performance on a scale of 1 to 5 would likely have low interrater reliability; the rating will differ depending on who is scoring the employees. Test-retest reliability refers to consistency of results over time. If a performance measure lacks test-retest reliability, determining whether an employee's performance has truly changed over time will be impossible. • Acceptability —Whether or not a measure is valid and reliable, it must meet the practical standard of being acceptable to the people who use it. For example, the people who use a performance measure must believe that it is not too time consuming. Likewise, if employees believe the measure is unfair, they will not use the feedback as a basis for improving their performance. • Specific feedback —A performance measure should specifically tell employees what is expected of them and how they can meet those expectations. Being specific helps performance management meet the goals of supporting strategy and developing employees. If a measure does not specify what an employee must do to help the organization achieve its goals, it does not support the strategy. If the measure fails to point out employees' performance problems, they will not know how to improve.

Organizations have developed a wide variety of methods for measuring performance. Some methods rank each employee to compare employees' performance. Other methods break down the evaluation into ratings of individual attributes, behaviors, or results. Many organizations use a measurement system that includes a variety of the preceding measures, as in the case of applying total quality management to performance management. Table 8.1 compares these methods in terms of our criteria for effective performance management

Making Comparisons The performance appraisal method may require the rater to compare one individual's performance with that of others. This method involves some form of ranking, in which some employees are best, some are average, and others are worst. The usual techniques for making comparisons are simple ranking, forced distribution, and paired comparison. Simple ranking Simple ranking requires managers to rank employees in their group from the highest performer to the poorest performer. In a variation of this approach, alternation ranking, the manager works from a list of employees. First, the manager decides which employee is best and crosses that person's name off the list. From the remaining names, the manager selects the worst employee and crosses off that name. The process continues with the manager selecting the second best, second worst, third best, and so on, until all the employees have been ranked. The major downside of ranking involves validity. To state a performance measure as broadly as "best" or "worst" doesn't define what exactly is good or bad about the person's contribution to the organization. Ranking therefore raises questions about fairness. forced-distribution Another way to compare employees' performance is with the forced-distribution method. This type of performance measurement assigns a certain percentage of employees to each category in a set of categories. For example, the organization might establish the following percentages and categories: • Exceptional—5 percent • Exceeds standards—25 percent • Meets standards—55 percent • Room for improvement—10 percent • Not acceptable—5 percent The manager completing the performance appraisal would rate 5 percent of his or her employees as exceptional, 25 percent as exceeding standards, and so on. A forceddistribution approach works best if the members of a group really do vary this much in terms of their performance. It overcomes the temptation to rate everyone high in order to avoid conflict. Research simulating some features of forced rankings found that they improved performance when combined with goals and rewards, especially in the first few years, when the system eliminated the poorest performers. 4 However, a manager who does very well at selecting, motivating, and training employees will have a group of high performers. This manager would have difficulty assigning employees to the bottom categories. In that situation, saying that some employees require improvement or are "not acceptable" not only will be inaccurate, but will hurt morale. paired-comparison method. Another variation on rankings is the paired-comparison method. This approach involves comparing each employee with each other employee to establish rankings. ex Suppose a manager has five employees, Allen, Barbara, Caitlin, David, and Edgar. The manager compares Allen's performance to Barbara's and assigns one point to whichever employee is the higher performer. Then the manager compares Allen's performance to Caitlin's, then to David's, and finally to Edgar's. The manager repeats this process with Barbara, comparing her performance to Caitlin's, David's, and Edgar's. When the manager has compared every pair of employees, the manager counts the number of points for each employee. The employee with the most points is considered the top-ranked employee. Clearly, this method is time consuming if a group has more than a handful of employees. For a group of 15, the manager must make 105 comparisons. In spite of the drawbacks, ranking employees offers some benefits. It counteracts the tendency to avoid controversy by rating everyone favorably or near the center of the scale. Also, if some managers tend to evaluate behavior more strictly (or more leniently) than others, a ranking system can erase that tendency from performance scores. Therefore, ranking systems can be useful for supporting decisions about how to distribute pay raises or layoffs. Some ranking systems are easy to use, which makes them acceptable to the managers who use them. A major drawback of rankings is that they often are not linked to the organization's goals. Also, a simple ranking system leaves the basis for the ranking open to interpretation. In that case, the rankings are not helpful for employee development and may hurt morale or result in legal challenges.

Rating Individuals Instead of focusing on arranging a group of employees from best to worst, performance measurement can look at each employee's performance relative to a uniform set of standards. The measurement may evaluate employees in terms of attributes (characteristics or traits) believed desirable. Or the measurements may identify whether employees have behaved in desirable ways, such as closing sales or completing assignments. For both approaches, the performance management system must identify the desired attributes or behaviors, then provide a form on which the manager can rate the employee in terms of those attributes or behaviors. Typically, the form includes a rating scale, such as a scale from 1 to 5, where 1 is the worst performance and 5 is the best. Rating Attributes The most widely used method for rating attributes is the graphic rating scale. This method lists traits and provides a rating scale for each trait. The employer uses the scale to indicate the extent to which the employee being rated displays the traits. The rating scale may provide points to circle (as on a scale going from 1 for poor to 5 for excellent), or it may provide a line representing a range of scores, with the manager marking a place along the line. Figure 8.3 shows an example of a graphic rating scale that uses a set of ratings from 1 to 5. A drawback of this approach is that it leaves to the particular manager the decisions about what is "excellent knowledge" or "commendable judgment" or "poor interpersonal skills." The result is low reliability, because managers are likely to arrive at different judgments. mix standard scale To get around this problem, some organizations use mixed-standard scales, which use several statements describing each trait to produce a final score for that trait. The manager scores the employee in terms of how the employee compares to each statement. ex Consider the sample mixed-standard scale in Figure 8.4. To create this scale, the organization determined that the relevant traits are initiative, intelligence, and relations with others. For each trait, sentences were written to describe a person having a high level of that trait, a medium level, and a low level. The sentences for the traits were rearranged so that the nine statements about the three traits are mixed together. The manager who uses this scale reads each sentence, then indicates whether the employee performs above ( + ), at (0), or below ( − ) the level described. The key in the middle section of Figure 8.4 tells how to use the pluses, zeros, and minuses to score performance. Someone who excels at every level of performance (pluses for high, medium, and low performance) receives a score of 7 for that trait. Someone who fails to live up to every description of performance (minuses for high, medium, and low) receives a score of 1 for that trait. The bottom of Figure 8.4 calculates the scores for the ratings used in this example. Rating attributes is the most popular way to measure performance in organizations. In general, attribute-based performance methods are easy to develop and can be applied to a wide variety of jobs and organizations. If the organization is careful to identify which attributes are associated with high performance, and to define them carefully on the appraisal form, these methods can be reliable and valid. However, appraisal forms often fail to meet this standard. In addition, measurement of attributes is rarely linked to the organization's strategy. Furthermore, employees tend perhaps rightly to be defensive about receiving a mere numerical rating on some attribute. How would you feel if you were told you scored 2 on a 5-point scale of initiative or communication skill? The number might seem arbitrary, and it doesn't tell you how to improve.

Rating Behaviors One way to overcome the drawbacks of rating attributes is to measure employees' behavior. To rate behaviors, the organization begins by defining which behaviors are associated with success on the job. Which kinds of employee behavior help the organization achieve its goals? The appraisal form asks the manager to rate an employee in terms of each of the identified behaviors. critical-incident method One way to rate behaviors is with the critical-incident method. This approach requires managers to keep a record of specific examples of the employee acting in ways that are either effective or ineffective. Here's an example of a critical incident in the performance evaluation of an appliance repairperson: example A customer called in about a refrigerator that was not cooling and was making a clicking noise every few minutes. The technician prediagnosed the cause of the problem and checked his truck for the necessary parts. When he found he did not have them, he checked the parts out from inventory so that the customer's refrigerator would be repaired on his first visit and the customer would be satisfied promptly. This incident provides evidence of the employee's knowledge of refrigerator repair and concern for efficiency and customer satisfaction. Evaluating performance in this specific way gives employees feedback about what they do well and what they do poorly. The manager can also relate the incidents to how the employee is helping the company achieve its goals. Keeping a daily or weekly log of critical incidents requires significant effort, however, and managers may resist this requirement. Also, critical incidents may be unique, so they may not support comparisons among employees. A behaviorally anchored rating scale A behaviorally anchored rating scale (BARS) builds on the critical-incidents approach. The BARS method is intended to define performance dimensions specifically, using statements of behavior that describe different levels of performance. (The statements are "anchors" of the performance levels.) The scale in Figure 8.5 shows various performance levels for the behavior of "preparing for duty." The statement at the top (rating 7) describes the highest level of preparing for duty. The statement at the bottom describes behavior associated with poor performance. These statements are based on data about past performance. The organization gathers many critical incidents representing effective and ineffective performance, then classifies them from most to least effective. When experts about the job agree the statements clearly represent levels of performance, they are used as anchors to guide the rater. Although BARS can improve interrater reliability, this method can bias the manager's memory. The statements used as anchors can help managers remember similar behaviors, at the expense of other critical incidents. 6 A behavioral observation scale A behavioral observation scale (BOS) is a variation of a BARS. Like a BARS, a BOS is developed from critical incidents. However, while a BARS discards many examples in creating the rating scale, a BOS uses many of them to define all behaviors necessary for effective performance (or behaviors that signal ineffective performance). As a result, a BOS may use 15 behaviors to define levels of performance. Also, a BOS asks the manager to rate the frequency with which the employee has exhibited the behavior during the rating period. These ratings are averaged to compute an overall performance rating. Figure 8.6 provides a simplified example of a BOS for measuring the behavior "overcoming resistance to change." A major drawback of this method is the amount of information required. A BOS can have 80 or more behaviors, and the manager must remember how often the employee exhibited each behavior in a 6- to 12-month rating period. This is taxing enough for one employee, but managers often must rate 10 or more employees. Even so, compared to BARS and graphic rating scales, managers and employees have said they prefer BOS for ease of use, providing feedback, maintaining objectivity, and suggesting training needs. organizational behavior modification Another approach to assessment builds directly on a branch of psychology called behaviorism, which holds that individuals' future behavior is determined by their past experiences—specifically, the ways in which past behaviors have been reinforced. People tend to repeat behaviors that have been rewarded in the past. Providing feedback and reinforcement can therefore modify individuals' future behavior. Applied to behavior in organizations, organizational behavior modification (OBM) is a plan for managing the behavior of employees through a formal system of feedback and reinforcement. Specific OBM techniques vary, but most have four components: 9 1. Define a set of key behaviors necessary for job performance. 2. Use a measurement system to assess whether the employee exhibits the key behaviors. 3. Inform employees of the key behaviors, perhaps in terms of goals for how often to exhibit the behaviors. 4. Provide feedback and reinforcement based on employees' behavior. OBM techniques have been used in a variety of settings. For example, a community mental health agency used OBM to increase the rates and timeliness of critical job behaviors by showing employees the connection between job behaviors and the agency's accomplishments. 10 This process identified job behaviors related to administration, record keeping, and service provided to clients. Feedback and reinforcement improved staff performance. OBM also increased the frequency of safety behaviors in a processing plant. 11 Behavioral approaches such as organizational behavior modification and rating scales can be very effective. These methods can link the company's goals to the specific behavior required to achieve those goals. Behavioral methods also can generate specific feedback, along with guidance in areas requiring improvements. As a result, these methods tend to be valid. The people to be measured often help in developing the measures, so acceptance tends to be high as well. When raters are well trained, reliability also tends to be high. However, behavioral methods do not work as well for complex jobs in which it is difficult to see a link between behavior and results or there is more than one good way to achieve success. 12

Measuring Results Performance measurement can focus on managing the objective, measurable results of a job or work group. Results might include sales, costs, or productivity (output per worker or per dollar spent on production), among many possible measures. Two of the most popular methods for measuring results are measurement of productivity and management by objectives. Productivity is an important measure of success, because getting more done with a smaller amount of resources (money or people) increases the company's profits. Productivity usually refers to the output of production workers, but it can be used more generally as a performance measure. To do this, the organization identifies the products—set of activities or objectives—it expects a group or individual to accomplish. At a repair shop, for instance, a product might be something like "quality of repair." The next step is to define how to measure production of these products. For quality of repair, the repair shop could track the percentage of items returned because they still do not work after a repair and the percentage of quality-control inspections passed. For each measure, the organization decides what level of performance is desired. Finally, the organization sets up a system for tracking these measures and giving employees feedback about their performance in terms of these measures. This type of performance measurement can be time consuming to set up, but research suggests it can improve productivity. 13 Management by objectives (MBO) is a system in which people at each level of the organization set goals in a process that flows from top to bottom, so employees at all levels are contributing to the organization's overall goals. These goals become the standards for evaluating each employee's performance. An MBO system has three components: 14 1. Goals are specific, difficult, and objective. The goals listed in the second column of Table 8.2 provide two examples for a bank. 2. Managers and their employees work together to set the goals. 3. The manager gives objective feedback through the rating period to monitor progress toward the goals. The two right-hand columns in Table 8.2 are examples of feedback given after one year. MBO can have a very positive effect on an organization's performance. In 70 studies of MBO's performance, 68 showed that productivity improved. 15 The productivity gains tended to be greatest when top management was highly committed to MBO. Also, because staff members are involved in setting goals, it is likely that MBO systems effectively link individual employees' performance with the organization's overall MBO can have a very positive effect on an organization's performance. In 70 studies of MBO's performance, 68 showed that productivity improved. 15 The productivity gains tended to be greatest when top management was highly committed to MBO. Also, because staff members are involved in setting goals, it is likely that MBO systems effectively link individual employees' performance with the organization's overall goals. In general, evaluation of results can be less subjective than other kinds of performance measurement. This makes measuring results highly acceptable to employees and managers alike. Results-oriented performance measurement is also relatively easy to link to the organization's goals. However, measuring results has problems with validity, because results may be affected by circumstances beyond each employee's performance. Also, if the organization measures only final results, it may fail to measure significant aspects of performance that are not directly related to those results. If individuals focus only on aspects of performance that are measured, they may neglect significant skills or behaviors. For example, if the organization measures only productivity, employees may not be concerned enough with customer service. The outcome may be high efficiency (costs are low) but low effectiveness (sales are low, too). 16 Finally, focusing strictly on results does not provide guidance on how to improve.

The principles of total quality management, introduced in Chapter 2, provide methods for performance measurement and management. Total quality management (TQM) differs from traditional performance measurement in that it assesses both individual performance and the system within which the individual works. This assessment is a process through which employees and their customers work together to set standards and measure performance, with the overall goal being to improve customer satisfaction. In this sense, an employee's customers may be inside or outside the organization; a "customer" is whoever uses the goods or services produced by the employee. The feedback aims at helping employees continuously improve the satisfaction of their customers. The focus on continuously improving customer satisfaction is intended to avoid the pitfall of rating individuals on outcomes, such as sales or profits, over which they do not have complete control. With TQM, performance measurement essentially combines measurements of attributes and results. The feedback in TQM is of two kinds: (1) subjective feedback from managers, peers, and customers about the employee's personal qualities such as cooperation and initiative; and (2) objective feedback based on the work process. The second kind of feedback comes from a variety of methods called statistical quality control. These methods use charts to detail causes of problems, measures of performance, or relationships between work-related variables. Employees are responsible for tracking these measures to identify areas where they can avoid or correct problems. Because of the focus on systems, this feedback may result in changes to a work process, rather than assuming that a performance problem is the fault of an employee. The TQM system's focus has practical benefits, but it does not serve as well to support decisions about work assignments, training, or compensation.

Sources of Performance Information All the methods of performance measurement require decisions about who will collect and analyze the performance information. To qualify for this task, a person should have an understanding of the job requirements and the opportunity to see the employee doing the job. The traditional approach is for managers to gather information about their employees' performance and arrive at performance ratings. However, many sources are possible. Possibilities of information sources include managers, peers, subordinates, self, and customers. Using just one person as a source of information poses certain problems. People tend to like some people more than others, and those feelings can bias how an employee's efforts are perceived. Also, one person is likely to see an employee in a limited number of situations. A supervisor, for example, cannot see how an employee behaves when the supervisor is not watching—for example, when a service technician is at the customer's facility. To get as complete an assessment as possible, some organizations combine information from most or all of the possible sources, in what is called a 360-degree performance appraisal.

Managers The most-used source of performance information is the employee's manager. For example, at YMCA of Greater Rochester, New York, managers rate the performance of the organization's 2,900 employees. The YMCA also reviews the managers' performance in evaluating employees. The vice president of human resources and the chief operating officer go over each performance appraisal together. When they identify reports in which feedback is vague or seems to be the first conversation a manager and employee have had about an issue, they work with the manager to improve the reviewing process. 17 It is usually safe for organizations to assume that supervisors have extensive knowledge of the job requirements and that they have enough opportunity to observe their employees. In other words, managers possess the basic qualifications for this responsibility. Another advantage of using managers to evaluate performance is that they have an incentive to provide accurate and helpful feedback, because their own success depends so much on their employees' performance. 18 Finally, when managers try to observe employee behavior or discuss performance issues in the feedback session, their feedback can improve performance, and employees tend to perceive the appraisal as accurate. 19 Still, in some situations, problems can occur with using supervisors as the source of performance information. For employees in some jobs, the supervisor does not have enough opportunity to observe the employee performing job duties. A sales manager with many outside salespeople cannot be with the salespeople on many visits to customers. Even if the sales manager does make a point of traveling with salespeople for a few days, they are likely to be on their best behavior while the manager is there. The manager cannot observe how they perform at other times.

Peers Another source of performance information is the employee's peers or co-workers. Peers are an excellent source of information about performance in a job where the supervisor does not often observe the employee. Examples include law enforcement and sales. For these and other jobs, peers may have the most opportunity to observe the employee in day-to-day activities. Peers have expert knowledge of job requirements. They also bring a different perspective to the evaluation and can provide extremely valid assessments of performance. 20 Peer evaluations obviously have some potential disadvantages. Friendships (or rivalries) have the potential to bias ratings. Research, however, has provided little evidence that this is a problem. 21 Another disadvantage is that when the evaluations are done to support administrative decisions, peers are uncomfortable with rating employees for decisions that may affect themselves. Generally, peers are more favorable toward participating in reviews to be used for employee development

Subordinates For evaluating the performance of managers, subordinates are an especially valuable source of information. Subordinates—the people reporting to the manager— often have the best chance to see how well a manager treats employees. Dell, for example, asks employees to rate their manager in terms of measures such as whether the employee receives ongoing performance feedback and whether the supervisor "is effective at managing people." 23 Subordinate evaluations have some potential problems because of the power relationships involved. Subordinates are reluctant to say negative things about the person to whom they report; they prefer to provide feedback anonymously. Managers, however, have a more positive reaction to this type of feedback when the subordinates are identified. When feedback forms require that the subordinates identify themselves, they tend to give the manager higher ratings. 24 Another problem is that when managers receive ratings from their subordinates, the employees have more power, so managers tend to emphasize employee satisfaction, even at the expense of productivity. This issue arises primarily when the evaluations are used for administrative decisions. Therefore, as with peer evaluations, subordinate evaluations are most appropriate for developmental purposes. To protect employees, the process should be anonymous and use at least three employees to rate each manager.

Self No one has a greater chance to observe the employee's behavior on the job than does the employee himself or herself. Self-ratings are rarely used alone, but they can contribute valuable information. A common approach is to have employees evaluate their own performance before the feedback session. This activity gets employees thinking about their performance. Areas of disagreement between the self-appraisal and other evaluations can be fruitful topics for the feedback session. YMCA of Greater Rochester introduced self-appraisals in response to complaints that ratings by the managers weren't an effective tool for employee development. Employees report that the opportunity to give examples of their successes and request training has sparked more helpful conversations with their managers. Managers, in turn, feel that the employee-provided information makes the evaluation process easier. 25 The obvious problem with self-ratings is that individuals have a tendency to inflate assessments of their performance. Especially if the ratings will be used for administrative decisions, exaggerating one's contributions has practical benefits. Also, social psychologists have found that, in general, people tend to blame outside circumstances for their failures while taking a large part of the credit for their successes. Supervisors can soften this tendency by providing frequent feedback, but because people tend to perceive situations this way, self-appraisals are not appropriate as the basis for administrative decisions. 26

Customers Services are often produced and consumed on the spot, so the customer is often the only person who directly observes the service performance and may be the best source of performance information. Many companies in service industries have introduced customer evaluations of employee performance. Marriott Corporation provides a customer satisfaction card in every room and mails surveys to a random sample of its hotel customers. Whirlpool's Consumer Services Division conducts mail and telephone surveys of customers after factory technicians have serviced their appliances. These surveys allow the company to evaluate an individual technician's customerservice behaviors while in the customer's home. The "Best Practices" box provides another example of a company that effectively uses customer feedback to support better employee performance. Using customer evaluations of employee performance is appropriate in two situations. 27 The first is when an employee's job requires direct service to the customer or linking the customer to other services within the organization. Second, customer evaluations are appropriate when the organization is interested in gathering information to determine what products and services the customer wants. That is, customer evaluations contribute to the organization's goals by enabling HRM to support the organization's marketing activities. In this regard, customer evaluations are useful both for evaluating an employee's performance and for helping to determine whether the organization can improve customer service by making changes in HRM activities such as training or compensation. The weakness of customer surveys for performance measurement is their expense. The expenses of a traditional survey can add up to hundreds of dollars to evaluate one individual. Many organizations therefore limit the information gathering to short periods once a year.

As we noted in the previous section, one reason for gathering information from several sources is that performance measurements are not completely objective, and errors can occur. People observe behavior, and they have no practical way of knowing all the circumstances, intentions, and outcomes related to that behavior, so they interpret what they see. In doing so, observers make a number of judgment calls, and in some situations may even distort information on purpose. Therefore, fairness in rating performance and interpreting performance appraisals requires that managers understand the kinds of distortions that commonly occur

Several kinds of errors and biases commonly influence performance measurements: • People often tend to give a higher evaluation to people they consider similar to themselves. Most of us think of ourselves as effective, so if others are like us, they must be effective, too. Research has demonstrated that this effect is strong. Unfortunately, it is sometimes wrong, and when similarity is based on characteristics such as race or sex, the decisions may be discriminatory. 28 • If the rater compares an individual, not against an objective standard, but against other employees, contrast errors occur. A competent performer who works with exceptional people may be rated lower than competent, simply because of the contrast. • Raters make distributional errors when they tend to use only one part of a rating scale. The error is called leniency when the reviewer rates everyone near the top, strictness when the rater favors lower rankings, and central tendency when the rater puts everyone near the middle of the scale. Distributional errors make it difficult to compare employees rated by the same person. Also, if different raters make different kinds of distributional errors, scores by these raters cannot be compared. • Raters often let their opinion of one quality color their opinion of others. For example, someone who speaks well might be seen as helpful or talented in other areas, simply because of the overall good impression created by this one quality. Or someone who is occasionally tardy might be seen as lacking in motivation. When the bias is in a favorable direction, this is called the halo error. When it involves negative ratings, it is called the horns error. Halo error can mistakenly tell employees they don't need to improve in any area, while horns error can cause employees to feel frustrated and defensive.

Ways to Reduce Errors Usually people make these errors unintentionally, especially when the criteria for measuring performance are not very specific. Raters can be trained how to avoid rating errors. 29 Prospective raters watch videos whose scripts or storylines are designed to lead them to make specific rating errors. After rating the fictional employees in the videos, raters discuss their rating decisions and how such errors affected their rating decisions. Training programs offer tips for avoiding the errors in the future. Another training method for raters focuses on the complex nature of employee performance. 30 Raters learn to look at many aspects of performance that deserve their attention. Actual examples of performance are studied to bring out various performance dimensions and the standards for those dimensions. This training aims to help raters evaluate employees' performance more thoroughly and accurately.

Political Behavior in Performance Appraisals Unintentional errors are not the only cause of inaccurate performance measurement. Sometimes the people rating performance distort an evaluation on purpose to advance their personal goals. This kind of appraisal politics is unhealthy especially because the resulting feedback does not focus on helping employees contribute to the organization's goals. High-performing employees who are rated unfairly will become frustrated, and low-performing employees who are overrated will be rewarded rather than encouraged to improve. Therefore, organizations try to identify and discourage appraisal politics. Several characteristics of appraisal systems and company culture tend to encourage appraisal politics. Appraisal politics are most likely to occur when raters are accountable to the employee being rated, the goals of rating are not compatible with one another, performance appraisal is directly linked to highly desirable rewards, top executives tolerate or ignore distorted ratings, and senior employees tell newcomers company "folklore" that includes stories about distorted ratings. Political behavior occurs in every organization. Organizations can minimize appraisal politics by establishing an appraisal system that is fair. One technique is to hold a calibration meeting, a gathering at which managers discuss employee performance ratings and provide evidence supporting their ratings with the goal of eliminating the influence of rating errors. As they discuss ratings and the ways they arrive at ratings, managers may identify undervalued employees, notice whether they are much harsher or more lenient than other managers, and help each other focus on how well ratings are associated with relevant performance outcomes. For example, when consultant Dick Grote leads calibration meetings for his clients, he often displays flip charts, one for each rating on a scale, and gives each manager a differentcolored Post-it Note pad. On their Post-It Notes, the managers write the names of each employee they rate, and they attach a note for the rating they would give that employee. The distribution of colors on the flip charts provides visually strong information about how the different managers think about their employees. A cluster of green notes on "outstanding" and yellow notes on "meets expectations" would suggest that one manager is a much tougher rater than others, and they could then discuss how they arrive at these different conclusions. 31 The organization can also help managers give accurate and fair appraisals by training them to use the appraisal process, encouraging them to recognize accomplishments that the employees themselves have not identified, and fostering a climate of openness in which employees feel they can be honest about their weaknesses. 32

Scheduling Performance Feedback Performance feedback should be a regular, expected management activity. The custom or policy at many organizations is to give formal performance feedback once a year. But annual feedback is not enough. One reason is that managers are responsible for correcting performance deficiencies as soon as they occur. If the manager notices a problem with an employee's behavior in June, but the annual appraisal is scheduled for November, the employee will miss months of opportunities for improvement. Another reason for frequent performance feedback is that feedback is most effective when the information does not surprise the employee. If an employee has to wait for up to a year to learn what the manager thinks of his work, the employee will wonder whether he is meeting expectations. Employees should instead receive feedback so often that they know what the manager will say during their annual performance review. Finally, employees have indicated that they are motivated and directed by regular feedback; they want to know if they are on the right track. Managers have found that young employees in particular are looking for frequent and candid performance feedback from their managers. 33 In response, Ernst & Young created an online "Feedback Zone," where employees can request or submit performance feedback at any time beyond the formal evaluations required twice a year.

Preparing for a Feedback Session Managers should be well prepared for each formal feedback session. The manager should create the right context for the meeting. The location should be neutral. If the manager's office is the site of unpleasant conversations, a conference room may be more appropriate. In announcing the meeting to an employee, the manager should describe it as a chance to discuss the role of the employee, the role of the manager, and the relationship between them. Managers should also say (and believe) that they would like the meeting to be an open dialogue. As discussed in the "HR How To" box, the content of the feedback session and the type of language used can determine the success of this meeting. Managers should also enable the employee to be well prepared. The manager should ask the employee to complete a self-assessment ahead of time. The self-assessment requires employees to think about their performance over the past rating period and to be aware of their strengths and weaknesses, so they can participate more fully in the discussion. Even though employees may tend to overstate their accomplishments, the self-assessment can help the manager and employee identify areas for discussion. When the purpose of the assessment is to define areas for development, employees may actually understate their performance. Also, differences between the manager's and the employee's rating may be fruitful areas for discussion

Conducting the Feedback Session During the feedback session, managers can take any of three approaches. In the "tell-and-sell" approach, managers tell the employees their ratings and then justify those ratings. In the "tell-and-listen" approach, managers tell employees their ratings and then let the employees explain their side of the story. In the "problem-solving" approach, managers and employees work together to solve performance problems in an atmosphere of respect and encouragement. Not surprisingly, research demonstrates that the problem-solving approach is superior. Perhaps surprisingly, most managers rely on the tell-and-sell approach. 34 Managers can improve employee satisfaction with the feedback process by letting employees voice their opinions and discuss performance goals. The content of the feedback should emphasize behavior, not personalities. For example, "You did not meet the deadline" can open a conversation about what needs to change, but "You're not motivated" may make the employee feel defensive and angry. As the "HR Oops!" box shows, even employees who are told they are meeting performance goals may not see this as a compliment. The feedback session should end with goal setting and a decision about when to follow up.

Finding Solutions to Performance Problems When performance evaluation indicates that an employee's performance is below standard, the feedback process should launch an effort to correct the problem. Even when the employee is meeting current standards, the feedback session may identify areas in which the employee can improve in order to contribute more to the organization in a current or future job. In sum, the final, feedback stage of performance management involves identifying areas for improvement and ways to improve performance in those areas. As shown in Figure 8.7 , the most effective way to improve performance varies according to the employee's ability and motivation. In general, when employees have high levels of ability and motivation, they perform at or above standards. But when they lack ability, motivation, or both, corrective action is needed. The type of action called for depends on what the employee lacks: • Lack of ability —When a motivated employee lacks knowledge, skills, or abilities in some area, the manager may offer coaching, training, and more detailed feedback. Sometimes it is appropriate to restructure the job so the employee can handle it. • Lack of motivation —Managers with an unmotivated employee can explore ways to demonstrate that the employee is being treated fairly and rewarded adequately. The solution may be as simple as more positive feedback (praise). Employees may need a referral for counseling or help with stress management. • Lack of both —Performance may improve if the manager directs the employee's attention to the significance of the problem by withholding rewards or providing specific feedback. If the employee does not respond, the manager may have to demote or terminate the employee. As a rule, employees who combine high ability with high motivation are solid performers. As Figure 8.7 indicates, managers should by no means ignore these employees on the grounds of leaving well enough alone. Rather, such employees are likely to appreciate opportunities for further development. Rewards and direct feedback help to maintain these employees' high motivation levels