CHAPTER FOUR Employee Characteristics 99 EMPLOYEE CHARACTERISTICS Previous chapters showed how managers gather information about the organization's external environment, its structure, and its work characteristics. This information helps them make better HR decisions. In this chapter, we consider information about the people inside the organization. A huge variety of individual characteristics could be measured, and sometimes the list seems overwhelming. For instance, consider the effect of a Japanese shoe manufacturer's invention. This machine determines the dimensions of a person's foot by using laser beams and then sends those measurements to a computerized production line that cuts, sews, and glues the leather pieces together to make a custom-fitted shoe. When we told human resource managers about this, they said, "I hope my organization doesn't find out, or we'll have to add foot size to the employee files." Sometimes, it's tempting to measure everything, but that's costly and often wasteful. So, what should we measure and how? Prior chapters have already introduced some employee characteristics, such as demographics and competencies. Later chapters will discuss employee attributes relevant to HR activities, such as ability in staffing or motivation in compensation. Here, we focus on two employee attributes that span many activities: performance and attitudes. Performance Performance reflects the organization's success, so it is perhaps the most obvious employee characteristic to measure. Employee performance is fundamental to other HR activities, such as who to hire, promote, lay off, and reward. Thus, this chapter describes performance measurement in some detail. Closely related to performance is employee absence. Even very good performers provide little value if they don't come to work regularly. So, this chapter also discusses how to measure absence and what causes it. Attitudes and Opinions Employee performance reflects mainly the efficiency objectives of the organization, but equity objectives are also key to the diagnostic approach. A key measure of equity is employee work satisfaction and commitment to the organization. Low employee morale may signal future behavior problems, so tracking employee attitudes is also related to efficiency. Employee performance, attendance, attitudes, and opinions are only a few of the employee characteristics that support HR decisions. While focusing on specific HR activities, later chapters describe the characteristics most relevant to each activity. Skills and abilities are relevant to recruiting and hiring. Motivation is relevant to pay. Characteristics indicating family obligations (such as aging parents or small children) are relevant to labor and employee relations. Next, we explore how organizations know if employees are getting the job done. PERFORMANCE Performance assessment tries to give employees the feedback (information about their performance) they need to improve, without diminishing their 100 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.1 Six Key Questions in Performance Assessment Why assess performance? What performance to assess? How to assess performance? Who should assess performance? When to assess performance? How to communicate performance assessments? Performance assessment, or performance appraisal, ^dependence and motivation to do a good is the process that measures employee ->ob- Designmg and implementing perfor- performznee. Employee performance is the degree mance assessment systems requires an-to which employees accomplish work requirements. ^h™^ 1™ qUCStl°nS' aS Sh°Wn m Why Measure Performance, a "Deadly Disease"? Must an organization measure employee performance? Many suggest that performance measurement does little good and may cause great harm. W. Edwards Deming, an expert on quality whose teachings have shaped orgamz^TtonsTn Japan and worldwide, calls performance appraisal one of the seven deadly diseases afflicting American management practice. He views performance ratings as "a lottery, with individual ratings emanating from random factors outside individual control; therefore, performance appraisal is an affliction that should be 'purged from the earth.'"2 Managers in one study perceived no consequences or any practical value in conducting formal performance appraisals.3 Companies Q-1 / such as Ceridian Corporation and Wisconsin Power & Light dropped routine re-^ / views for everyone except the "bad apples."4 Also, performance assessment has costs to develop the system, to have managers and employees carry it out, and to process the results. Developing and implementing a performance assessment system can cost hundreds of thousands of dollars. Clearly, performance assessment should not be undertaken without understanding its added value. ^ Yet, differences in individual performance can have a huge impact. A recent S study calculated the percentage performance difference between high perform-^\w / ers and average performers in different jobs. For routine blue-collar work, high performers performed 15 percent better than average; for routine clerical work, ^ f V 17 percent; for crafts, 25 percent; for clerical decision makers, 28 percent; for ^> \ ntVYF^OC-lrtTVolc Af\ r\£»»V£»rit" 4rr\r nAninonron^fi color A O t-v^i• r-\ t ■ -i 1-1 /\ 4-V-,-»- iiirnrn-inn professionals, 46 percent; for noninsurance sales, 42 percent; and for insurance ^ ^ ^ *C sales, 97 percent.5 This means that the best-performing insurance salespeople <^. J Valdez ran aground on March 24, 1989, causing the largest oil spill in Alaska, the difference between good and poor tanker captain performance became all too apparent. Value Added: Integrating HR Activities Performance information can serve four general purposes: (1) providing feedback about strengths and weaknesses: (2) distinguishing between individuals to CHAPTER FOUR Employee Characteristics 101 USE ?J3 allocate rewards; (3) evaluating and maintaining the human resource systems of the organization; and (4) creating a paper trail to document the reason for certain actions, such as dismissing an employee.6 Thus, performance information can support virtually every decision managers make about people. Perhaps appraisals are so disparaged precisely because they are constructed to serve so many purposes. Even when the explicit purpose of appraisals is the same, managers perceive the purpose differently, and they may change their appraisals to fit the purpose—for example, giving more lenient appraisals if the purpose is to inform employees about their performance. However, when performance information is matched to the situation, it can help to integrate HR activities, so that they support each other. Performance, like other HR activity, must be tailored to achieve the right goals.8 Performance appraisal, like all measurements, should also be flexible. As perpetual change becomes the rule, the standards and methods used must also change. As you read this chapter, keep in mind that organizations constantly combine and recombine methods, and constantly work to fit appraisal information to organization goals. Potential Conflict Unfortunately, the goals of performance measurement often conflict. You have probably experienced this. Suppose you and a classmate agree to share lecture notes when one of you misses class, but your classmate's notes are incomplete. How do you provide feedback on this performance? You want your classmate to see the problem and to improve but not to become alienated. This is the dilemma—how to provide positive feedback th;>t accurately fits the performance. Exhibit 4.2 shows how this conflict affects organizations. 9 Conflict within an organization reflects tension between the organization's policy to give the highest rewards to the most deserving employees (paying for high performance, promoting the most qualified, separating the clearly unacceptable performers) and the organization's desire to help employees improve by providing honest feedback. Linking rewards to performance requires comparing employees to each other (or to a standard) and focusing on the past. Helping employees improve requires considering each employee independently and focusing on the future. Coach or Judge? The Question of Leadership INK? Suppose you held stock shares in GE Corporation, and you were told that managers may be fired by Jack Welch because they don't exhibit certain values, even if they reach their profit and sales goals. Do you think such a policy is good for GE shareholders? Why or why not? Should today's managers be leaders? "We need to function more as coaches and less as dictators," says one senior executive.10 The Center for Creative Leadership identifies 22 attributes of leaders, such as being daring, dynamic, considerate, empowering, entertaining, friendly, thrifty, optimistic and trusting. However, it's difficult to score highly on all of these, especially when discussing poor performance. How do you entertain and seem friendly when being "thrifty" requires telling someone they will not receive a salary increase due to poor performance? How important is leadership? In his letter to shareholders of General Electric, CEO Jack Welch declared that "GE cannot afford management styles that suppress and intimidate" subordinates. Leaders were assessed as high or low on their business results, but also on how well they exhibited these values. High on both values and results is good performance, an "easy decision." Low on both is poor performance, 102 PART ONE Assess Conditions. Plan and Set Objectives EXHIBIT 4.2 Performance Assessment Can Cause Contiict Purpose Organization Goals Individual Goals Allocating rewards and opportunities Providing feedback and counseling Pay for performance Promote most qualified Weed out unsatisfactory workers High pay Career progress Job security Within-organization conflict Help empoyees improve Identify training needs Improve communication Individual-organization conflict Within-organization conflict Discuss difficulties Locate assistance Share ideas an "easy decision." Miss your results but share the values, and you get a second chance. If you make your results, but do so by force or intimidation, you may be ousted." Honesty versus a Good Impression Exhibit 4.2 shows that within-individual conflict affects the person being evaluated. People want to obtain rewards that come with favorable appraisals, so they may use impression management, either by promoting themselves to appear more competent or by ingratiation to make the appraiser like them more. Impression management, especially ingratiation, does raise performance evaluations.12 However, if employees need help with problems, they must honestly share their difficulties. Painting too rosy a picture may prevent getting needed assistance, but sharing too many problems may give the impression that they can't do their jobs. Accuracy versus Inflation The final conflict shown in Exhibit 4.2 exists between the individual and the organization, because the individual's goals of obtaining high rewards conflict with both of the organization's goals. Accurate information is needed to allocate rewards and provide feedback, but employees may share only the most positive information out of fear they won't be rewarded. This is especially true for poor performers and may explain why poor performance appraisals are rare and difficult. If you have ever had to tell team members they failed to pull their weight, you know how unpleasant giving a poor appraisal can be. Despite the cost and conflict, current practice supported by research suggests that managers believe the value of performance assessment outweighs its costs. The next step is deciding what performance to measure. CHAPTER HOUR Employee Characteristics 103 iat Performance to Measure A manager in Melbourne, Australia, tells of a new clerk in a hardware store who was confronted by a customer carrying a tire. The customer said, "This tire is defective. Look at how the tread is separating from the body. I paid $50 for this tire, and I want my money back." The clerk had been told by her supervisor to be customer-friendly, and to make every effort to satisfy the customer. So, the clerk cheerfully gave the customer a refund in exchange for the obviously defective tire. Only later did the clerk realize that the store did not sell tires. Is this an example of very good or very bad performance? What is performance anyway? If a salesperson sells lots of merchandise, shouldn't that be enough? Not necessarily. Organizations also value long-term customer satisfaction, helping other employees, and paying attention to customers before and after the sale. In the film Miracle on 34th Street, Kris Kringle was applauded for sending Macy's customers to other stores if Macy's didn't have exactly what the customer wanted. PepsiCo truck driver Buck Robuck ignored the computer-generated map supplied by PepsiCo, bypassing the map's interstates for back roads and side streets that save time. Was this insubordination? PepsiCo president Roger King praises this as an example of the "ownership mentality" critical to business success.13 Many things could be measured to determine performance, as shown in Exhibit 4.3. On the left are individual characteristics. Although not often directly observed, these combine with tasks and organization factors to produce behaviors that can be observed. Appropriate behaviors lead to results reflecting the combined efforts of many individuals. Multiple performance measures are needed. Behaviors reflect a person's attempts to perform; individual characteristics signal the causes of the behaviors. Results ensure a link between individual behaviors and broader goals. Citizenship or Counterproductivity? Performance reflects more than simply carrying out work tasks. A broad spectrum of behaviors determines the value of people to organizations. On the positive side, such extra behaviors have been called organization citizenship. Sometimes, citizenship may be controversial, as in the case of whistle-blowers who support the organizational value of honesty by revealing unethical or dangerous practices at work. Employees with higher work satisfaction (discussed later) may engage in more citizenship behaviors.15 On the other hand, counterproductive behaviors include a wide variety of harmful actions, such as theft, injuries, or violence. Employee theft costs U.S. organizations billions of dollars each year, and injuries cause over 90 lost work days per year, per 100 full-time workers. Like citizenship, theft is related to employees' satisfaction and perceptions of fairness. Organizations also monitor smoking, drug use, telephone conversations, and even excess wiggling by computer operators.16 A Bureau of National Affairs study of how white-collar performance was measured found that 93 percent of firms used quality of work, 90 percent used quantity of work, 87 percent used initiative, 87 percent used cooperation, 86 f\rganization citizenship reflects helpful and I ^cooperative behaviors that go beyond the specific tasks of the job, including (1) helping )thers, (2) sharing and creating new ideas, (3) being iependable and reliable, (4) defending and promoting le organization's goals.14 104 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.3 Examples of Performance Criteria Skills/Abilities/Needs/Traits Behaviors Results Job knowledge Perform tasks Sales Strength Obey instructions Production levels Eye-hand coordination Report problems Production quality Licenses Maintain equipment Wastage/scrap Business knowledge Maintain records Accidents Desire to achieve Follow rules Equipment repairs Social needs Attend regularly Customers served Dependability Submit suggestions Customer satisfaction Loyalty Smoking abstinence Honesty Drug abstinence Creativity Leadership WHAT DO YOU THINK? How can an HR manager help an organization celebrate "mistakes" that indicate pursuit of organization goals, but at the same time correct mistakes made due to ignorance or wrong decisions? How would you help a supervisor see the difference? percent used dependability, 85 percent used job knowledge, 79 percent used attendance, and 67 percent used need for supervision.17 Of course, sometimes, it's hard to tell if an activity is good or bad for the organization. Scientific studies have suggested that listening to stereo headsets and filing formal grievances about work may associate with higher performance. At Motorola they celebrate mistakes "in pursuit of business goals" as learning opportunities for an empowered work force.18 Thus, there is no single best measure of performance, but there are guidelines. Performance ratings should be goal related, observable, understandable, and controllable. A core competency is a basic capability essential for the organization to compete and grow. Will the Measure Contribute to Achieving Goals and Good Decisions? The overriding goal is to improve decisions. For allocating rewards, the most valuable performance measures might reflect results. For assigning employees to training or developing their careers, individual characteristics such as knowledge may be best. To weed out the least valuable employees, identifying _ counterproductive behaviors or results may be useful. Many managers believe that future decisions will reflect increasingly rapid change. Rather than focusing on jobs, organizations "identify, cultivate, and exploit the core competencies that make growth possible."19 Although Honda, for example, makes many products, its core competencies are designing and building engines regardless of how jobs are designed to do that. The key decisions for managers and employees involve how best to identify and nurture these core competencies. Here, performance assessment might focus either on the skills, abilities, needs, or traits thought to reflect those competencies or on behaviors that use those competencies. The key to supporting decisions is to ask, Who uses the appraisal information and what do they do with it? CHAPTER FOUR Employee Characteristics 105 Can the Performance Be Observed? Performance assessment supports important decisions such as promotions, layoffs, and rewards. As you saw in Chapter Two, the United States and other countries have employment laws that protect certain groups from discrimination, so performance information often becomes a key piece of evidence. For this and other reasons, good performance criteria are observable and objective. Even flying like an eagle can be observable. PepsiCo wants managers who "act like owners, run lean, and get big results.*' CEO D.W. Calloway says. "We take eagles and teach them to fly in formation.'* so the performance assessment process for managers includes an annual performance review between a superior and each subordinate manager, focusing on "what the manager actually did this year to make a big difference in the business, not whether he's a nice guy or wears the right color socks. Did he make his sales target? Did he develop a successful new taco chip or soda commercial?" PepsiCo combines these outcome-oriented appraisals with subordinate evaluations of their bosses, in confidential reports. The emphasis is on outcomes and behaviors, not just traits.20 Are Measures Understandable? Obviously, raters need to understand what they are looking for, but so should the person being rated. For example, meeting the production quota seems to be a very understandable performance measure. Yet, a popular cartoon in the former Soviet Union showed two people standing in front of one huge nail. One person says, "I know it's useless, but we attained our quota of producing 100 kilograms of nails this month." Understanding the meaning of good performance is even more difficult when it includes providing good customer service, achieving high quality, or displaying creativity. Not all performance measures must be based on simple behaviors, but when both raters and those being rated mutually understand what is required, they are more likely to aim for common goals. A study of financial analysts found that the more employees felt they understood the appraisal system the more likely they were to agree with their supervisor's ratings.21 Can the Performer Control Performance? It makes sense to measure performance factors that the employee can control. For example, using a customer satisfaction survey as the sole performance indicator for airline maintenance engineers probably doesn't make sense. Such surveys might be appropriate for the flight attendant who directly contacts virtually every airline passenger. Raters and those being rated may disagree about how much performance can really be controlled. For example, college students working as supervisors believed that poor performance was more likely to be caused by subordinates' motivation or ability while students working as subordinates believed that poor performance was more likely to be caused by tools and equipment, task preparation, personal constraints, and scheduling.22 If subordinates and managers disagree about how much the performer can control performance, the rating system may fail to motivate much improvement. to Measure Performance How something is measured may be as important as what is measured. Suppose grade point averages (GPA) determine which students are on the dean's list. 106 PART ONE Assess Conditions. Plan and Set Objectives GPA could be measured by ranking all students and awarding dean's list positions to the top 10 percent. If 100 students compete, only the top ten get on the list, even if the llth-highest student's GPA is very close to the 10th. Or, we could compare every student's GPA to a fixed standard, such as 3.8 on a 4.0 scale. The 3.8 GPA cutoff might be chosen because it usually separates the top 10 percent, but the dean's list award would go to everyone who achieves 3.8 or better. Some years, this may be the top 9 percent; in other years, the top 12 percent. Both methods measure GPA, so would it make any difference? Students say it does. Ranking sends the message that students are competing with each other. Competition can get so fierce that books with key exam material mysteriously disappear from the library just before an exam, only to return afterward. The second approach, rating, allows potentially all students to be on the dean's list if they achieve the GPA standard. Rating sends the message that students don't have to beat other students to excel and may even foster cooperation among them. Similarly, after an organization decides what performance to measure, it decides how to observe and record the information. In Exhibit 4.4 we list performance assessment approaches as comparisons to agreed objectives, to job standards, and between individuals. This exhibit also contains the most common methods for each comparison. Organizations often combine performance assessment techniques. Next, we describe the most widely used performance assessment techniques and discuss their advantages and disadvantages. Comparing to Objectives: MBO Management by objectives (MBO) is like writing a contract. The person being rated and the rater set objectives to achieve by a certain date. The objectives should be measurable and observable. Performance is assessed by jointly re-viewing how well the objectives are achieved. A salesperson might set a quarterly goal of TOO sales calls, selling 1,000urutsj)f a certain product, and^om-pleting"4 sales courses. At the end of the quarter, that personand the perforrnlmlJe^ppraisermigfa that~the salesperson madeJUO calls, soldjOO units, and completed 3 courses, and the reasons for the deviations would be discussed. Some companies encourage supervisors and employees to set goals that are SMART: Specific results are obtained. Measurable in quantity, quality, and impact. Attainable, challenging yet within view. Relevant to the work unit, organization, career, and so forth. Time-specific, with deadlines to expect a result. Research shows that MBO increases productivity for both individuals and groups, especially when the goals are specific, challenging, and accepted.23 Exhibit 4.5 shows the guidelines used by General Electric Corporation for setting "stretch goals." It's also j important to keep culture in mind, as AT&T learned when it required production plants in the Czech Republic to prominently display banners reading, among other things, "Work for the Team, not for the Individual." Though successful in the United States, such i WHAT DO YOU THINK? Admonishing and appraising workers based on teamwork may disturb workers in cultures where "teamwork" was often a code word for following the party line, and for a reward system based on loyalty to a political party, not work performance. If you wanted to encourage teamwork in such a culture, how would you gain the trust of workers? CHAPTER FOUR Employee Characteristics 107 [HIBIT 4.4 Performance Assessment Comparisons Comparison to Agreed Objectives Comparison to Job Standards Comparison between Individuals Management by objectives Physical observation Ranking (MBO) Checklists Forced distribution Rating scales Critical incidents Behaviorally anchored rating scale (BARS) Essays/diaries EXHIBIT 4.5 Guidelines for Setting "Stretch Goals" at General Electric Corporation • By definition, stretch goals are very difficult to meet. Don't punish people for not hitting them. • Don't set goals that stretch your employees crazily. • Understand that stretch targets can unexpectedly affect other parts of the organization. • Don't give tough stretch goals to those people already pushing themselves to the limit. • Share the wealth generated by reaching stretch goals. Source: Strat Sherman, "Stretch Goals: The Dark Side of Asking for Miracles," Fortune, Nov. 13, 1995, p. 231, © Time. Inc. All rights reserved. banners disturbed Czech workers because they were similar to the propaganda under the old Soviet-style manufacturing systems. MBO is attractive in fast-changing situations because it does not rely on a fixed set of behaviors or results. Computer software can help make change even quicker by guiding managers to establish unit goals, assigning these goals to individuals and teams, tracking progress (e.g., green is on schedule, yellow is behind, red is critical), reminding managers when to track performance, and suggesting rewards such as sports tickets or dinner for two.24 Of course the flexibility of MBO may also be its greatest flaw because performance standards can change so much that they eventually bear little resemblance to the original purpose. Like all human resource processes, MBO must be monitored to achieve its goals. Comparing Performance to Established Standards These methods examine the work and set standards describing desirable or undesirable performance. The characteristics of the person being rated are compared to these standards. 108 PART ONE Assess Conditions. Plan and Set Objectives Physical Observation Watching people work might seem rather old-fashioned, suited mostly for assessing performance of manual laborers or athletes, but physical observation is indeed used for some of the world's most high-technology jobs. As many recent movies have shown, computer users can be monitored by recording keystrokes, reading their electronic mail, and even installing devices in their chairs to detect wiggling. (So far, we haven't seen a "student wiggling" method for assessing professors' classroom performance.) The World Wide Web has made "superfluous surfing" (visiting sites for games, sports, or seamier fare) cause for "disciplinary action or termination" at companies such as Pepsi-Cola and Lockheed. Employees at Olivetti Research Laboratories wear badges that constantly report their locations. Just like Star Trek, managers can call up information on employees' locations at any time.25 One of the most controversial physical observations is the lie detector, or polygraph. These tests were severely restricted in 1988 by the Employee Polygraph Protection Act (EPPA).26 An equally controversial physical performance measure is testing for drug or other substance abuse. A survey of 1,200 U.S. companies revealed that drug testing increased tenfold between 1987 and 1993, driven in large part by government mandates. Ninety-one percent used urinalysis, while 13 percent used blood tests. The average cost per person was $41 in 1991. Apparently, this form of appraisal has an effect because the percentage testing positive dropped from 4.2 percent to 2.7 percent in just one year. Companies that prefer not to use physical drug tests use video games to test for drug impairment. Research suggests that substance abusers report more withdrawal behaviors (daydreaming, absence, low effort, theft) and antagonistic behaviors (filing complaints, arguing, disobeying instructions, gossiping). The legality and value of substance abuse tests continues to be debated.27 Checklists A checklist is a set of behaviors, adjectives, or descriptive statements. A rater who believes the statement describes a person's performance checks the item; if it does not, the rater leaves it blank. Each statement is scored to reflect its positive or negative impact on job performance. The performance rating is the sum of the scores for the items checked. Rating Scales One of the oldest and most widely used assessment techniques is the rating scale, sometimes a graphic rating scale when it appears as a graph or line or series of boxes along which performance levels are marked. A graphic rating scale in Exhibit 4.6 shows some of the criteria discussed earlier. The evaluator marks the box that best describes the evaluated person's performance on each criterion. If desired, each level could be assigned a number, perhaps ranging from 5 for outstanding to 1 for unsatisfactory. When each criterion is assigned a weight according to its importance, every person's evaluation can be expressed as the sum of the importance weights times the level of each criterion achieved. Summated scales use more sophisticated statistical methods for assigning such weights to each item.28 Mixed standard scales give the rater several statements, and ask how typical they are of the person's performance. Exhibit 4.7 shows such a scale for professors. CHAPTER FOUR Employee Characteristics 109 EXHIBIT 4.6 Typical Graphical Rating Scale Name Dept. Date Outstanding Good Satisfactory Fair Unsatisfactory Knowledge of job □ □ □ □ □ Clear understanding of the facts or factors pertinent to the job Comments: Personal qualities □ □ □ □ □ Personality, appearance. sociability, leadership, integrity Comments: Cooperation □ □ □ □ □ Ability and willingness to work with associates, supervisors, and subordinates toward common goals Comments: 4.7 Mixed Standard Scale for Rating Professor Performance For each item, indicate whether the professor's performance is better than, equal to, or worse than the description: 1. Professor's response repeats a point in the lecture. 2. Professor stands behind the podium. 3. Professor misses words when reading from notes. 4. Professor insults or verbally attacks a questioner. 5. Professor uses concrete examples to clarify answers. 6. Professor varies pitch and tone of voice to emphasize points. Note: Items 1,4. and 5 relate to "Responding to Questions," and Items 2, 3, and 6 relate to "Speaking Style." Source: Kevin R. Murphy and Jeannette N. Cleveland, Understanding Performance Appraisal (Thousand Oaks. CA: Sage, 1995), p. 437. Critical Incidents Critical incidents are statements describing very effective and every ineffective behaviors critical to performance.29 Critical incidents can be included in almost any performance assessment technique, such as behaviorally anchored rating scales. Behaviorally Anchored Rating Scales (BARS) The BARS approach starts with a rating scale as described earlier and adds critical incidents that provide anchors for different points on the scale.30 These 110 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.8 Behaviorally Anchored Rating Scale (BARS) for a Residence Hall Adviser Performance Dimension Concern for individual dorm residents: attempts to get to know individual dorm residents and responds to their individual needs with genuine interest. This resident adviser could be expected to: Rating Scale Good (1) (2) (3) (4) (5) Poor Recognize when a floor member appears depressed and ask if person has problem he or she wants to discuss. Offer floor members tips on how to study for a course he or she has already taken. See person and recognize him/ her as a floor member and say "Hi." Be friendly with a floor member; get into discussion on problems, but fail to follow up on the problem later on with student. Criticize a floor member for not being able to solve his or her own problems. anchors make the scale more job-specific and, hopefully, less subjective and less error prone. The example in Exhibit 4.8 shows a BARS for a residence hall adviser. The steps in developing a BARS are: 1. Supervisors identify performance dimensions or categories of activities that make up the job. 2. Supervisors write a set of critical incidents for each dimension. 3. An independent group of supervisors sorts the incidents into dimensions and rates each incident on a good-to-bad performance scale. 4. Incidents consistently placed in one dimension are kept and scales are constructed for each dimension, anchored with incidents showing good and bad performance. Researchers have developed variations on BARS, including behavioral observation scales (BOS) that assess the frequency of behavior, and behavior discrimination scales (BDS) that compare the actual frequency of behavior to the opportunity and the expected frequency of the behavior.31 These variations offer some advantages, but no evidence shows the general superiority of BARS or its variations over more easily developed appraisal techniques.32 Essays/Diaries Evaluators can write essays describing strong and weak aspects of the employee's behavior over time. Essays can be completely open-ended; however, evaluators usually follow guidelines indicating topics and purposes. Essays can be constructed from diaries in which evaluators have recorded and observed critical incidents during the assessment period. Essays can also be used with rating scales or BARS to document and elaborate on the ratings. CHAPTER FOUR Employee Characteristics 111 Comparison between Individuals Did you ever tell someone you wanted to break up because you had met someone more interesting or have you ever heard this from someone else? Such comparisons are unpleasant and most people avoid them. Likewise, in organizations it is often easier to compare one person's performance to a goal or standard than to another's performance. Yet, some decisions require comparisons. Just as most of can't date two people on the same night, organizations can neither promote everyone nor give everyone maximum pay increases. Such decisions require dividing a fixed set of opportunities among employees, so individuals must be compared to each other. Comparisons often reflect only a subjective general impression, but they can reflect more objective data, using the techniques mentioned earlier. Next, we discuss several comparison methods. Ranking Ordering performers from highest to lowest is simple, fast, easy to understand, and inexpensive, but it becomes complicated for large numbers of employees. Alternation ranking and paired comparisons simplify the process. With alternation ranking, the ranker first determines who is the highest performer, then who is the lowest performer, then who is the second-highest, the second-lowest, and so on. Paired-comparison ranking compares every possible pair of employees, deciding which of the pair is better. After judging all possible pairs, the person with the most better-of-the-pair choices is ranked highest, and so on. In 1994. AT&T eliminated ratings completely for 37,000 U.S. managers because the fostered rivalry and discouraged teamwork. Some complain, however, that the stars don't stand out as well and that feedback is less effective, using the simple narrative descriptions.33 What*s the Best Way? . . . It All Depends It would be simple if there were one best way to measure performance but the most effective way depends on the situation. There is no right answer. The three-dimensional box in Exhibit 4.9 illustrates this. From right to left, the work gets less routine; from bottom to top, the employees require more independence; and from front to back the environment gets more unstable. The lower-left-front corner represents very routine, stable situations with low employee independence, where techniques that compare to behaviors and standards, such as rating scales, are appropriate. In the middle of the box, with moderate routinization, stability, and independence, it is appropriate to focus on achieving objectives (MBO), allowing employees latitude in how they do it. The upper-right-back corner shows very unstable and nonroutine situations, with a high degree of employee independence: this calls for very unstructured comparisons, perhaps using diaries or essays.34 Your tests and examinations reflect similar choices. Very stable, routine material might be tested with multiple-choice questions or mathematical problems. Moderately stable, changeable material might be tested with short-answer questions that give students some leeway in the language they use. Finally, less structured topics such as ethics, values, and opinions might best be tested using essays or papers, and imposing little structure on students. In organizations, a mix of different methods may be used to get a more complete picture. 112 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.9 How Characteristics of the Work, the Employees, and the Environment Affect Performance Assessment Methods Unstructured Volatile Routine Nonroutine Work routinization Who Should Judge Performance? What if you could decide your own grade in this course? Many students would leap at the chance to give themselves As regardless of their learning or test performance. On further reflection, however, some students realize that an A means little if everyone gets one. Others realize that they would honestly assign themselves a lower grade than the instructor would. Still others simply wouldn't want to deal with the decision. We could ask the same question about having the entire class vote on each person's grade, or having everyone take a standardized test administered by an independent testing agency. Along with the usual approach, where your instructor assigns grades, all of these methods have advantages and disadvantages. 360-Degree Appraisal: Breakthrough or Going in Circles? "360 Feedback Can Change Your Life" proclaims Fortune magazine: Everyone from the office screwup to your boss, including your crackerjack assistant and your rival across the hall, will fill out lengthy, anonymous questionnaires about CHAPTER FOUR Employee Characteristics 113 EXHIBIT 4.10 360-Degree Performance Feedback Top management (Internal customer) Source: John F. Millman. Robert A. Zawacki, Carol Norman. Lynda Powell, and Jay Kirksey. "Companies Evaluate Employees from All Perspectives," Personnel Journal. November, 1994, p. 100. you. You'll complete one, too. A week or two later you'll get the results, all crunched and graphed by a computer. Ideally, all this will be explained by someone from your human resources department or the company that handled the questionnaires, a person who can break bad news gently.35 Exhibit 4.10 shows the person being assessed ("ratee" in current parlance) in the middle of many different possible assessors. Each assessor can be perceived as a "customer," in the sense that they all need something from the ratee. Having peers or subordinates judge performance can produce insights. There is evidence that managers who receive subordinate feedback may improve over time, that leaders adjust their own self-ratings in line with follower ratings, and that managers from cultures that value uncertainty avoidance were perceived as more controlling and less likely to delegate. The 360-degree appraisal holds promise, but the evidence is still sketchy. Most performance systems still rely solely on the supervisor or higher-level manager.36 The trick is striking a balance between gathering every bit of available information and keeping the process manageable. No matter who is involved in appraising performance, they should have three characteristics: (1) opportunity to observe performance, (2) ability to translate observations into useful assessments, and (3) motivation to provide useful assessments. 114 PART ONE Assess Conditions, Plan and Set Objectives Being in the Right Place and Time: Opportunity to Observe Performance Although organizations typically rely on supervisor to observe performance, changing work arrangements give supervisors less opportunity to see all the important performance behaviors. You read earlier that many organizations now emphasize teams. Should team members assess each other's performance? W. L. Gore & Associates and Quaker Oats Company have used work teams for more than 20 years. At the Quaker Oats pet food plant, peer review makes up 100 percent of performance appraisal, with teams making all promotion and compensation decisions.37 At Chrysler Corporation, subordinates evaluate their bosses on teamwork, communication, quality, leadership, planning, and development of the workforce. The boss summarizes the anonymous reports and then meets with subordinates to discuss how to improve.38 In China, employment is guaranteed, so the Xian Department Store needed an alternative to dismissal as a way to motivate clerks who performed poorly. They created an award for the 40 worst employees and found shoppers eager to report instances of clerks throwing merchandise, being rude, and leaving their posts.39 Though less dramatic, many employers make employees accountable to a customer or client, whether it is an actual purchaser, a coworker who depends on that employee's performance, a team, or a supervisor. Subordinates and peers can provide more information than supervisors because they see different behaviors.40 Ready and Able: Ability to Make Useful Assessments Observing performance is just the first step. Assessors must understand the purposes of appraisal, and gather and report the information in a useful way. Psychologists list three phases of the appraisal process: (1) paying attention and observing, (2) storing information in memory, and (3) recalling and making an evaluation. Attention and observation depend on the "set" or framework of the observer, which is affected by stereotypes, knowledge of the job or rating scale, and the purpose of the rating. Storage and memory appears to be less objective than once believed; people using general impressions ("he's a good worker") to guide what they observe and remember. Recalling and evaluating are affected by the age, sex, attractiveness or similarity between ratees and raters. More intelligent raters may be more accurate, long delays decrease recall and accuracy, and ratings made for pay purposes may be less accurate than those made solely for research or for providing feedback. Psychologists have focused mostly on rating scales and rater thought processes, but ratings may also be affected by organizational pressures on raters and the degree of trust between raters and ratees, where we have much less information.41 Appraisal training (discussed later) may also improve ability, but it's difficult to train some appraisers, such as customers. Can a computer appraise performance? Not yet, but it can help. On the web, you can view a description of performance review software at http://www.austin-hayne.com/ea3BMfct.htm NO THE WEB Personal computer "review writers" offer advice, store notes, provide suggested descriptions of performance (even inserting the ratee's first name to personalize it), check for "inappropriate" words, and provide a box warning that if you give a low rating, you should include supporting details in case of a legal challenge.42 CHAPTER FOUR Employee Characteristics 115 Politics and Social Influence: Motivation to Provide Useful Assessments Rater motivation may often be more important that the design of the appraisal system itself. Has your instructor ever asked you to tell him or her how you "really" feel the class is going? Was your first thought to give the most accurate and complete information that you could? More likely, you thought about what would happen to you. If you want to avoid unpleasant confrontations and increase the chance that you will receive a good grade, it may be best to avoid saying anything very negative—and this is exactly what researchers have found. The more accountable you are for your ratings and the more your rewards depend on what your boss thinks, the more incentive to inflate your appraisal. There's even an economic theory to explain it. Bosses may wonder why they don't get much feedback when they say, "I don't want yes-men around me. Tell me what you think, even if it costs you your job!"43 This affects not only appraisals of bosses by subordinates, but also works in the other direction. Seventy-six percent of U.S. managers admitted they softened their ratings to avoid discouraging someone, to avoid conflicts, to encourage subordinates when pay raises were low, or to protect someone facing personal difficulties such as divorce. It also appears that this pattern may get worse at higher organization levels.44 Exhibit 4.11 summarizes some of the effects of political appraisals. The reality of appraisal is social. Relationships matter, and appraiser motivation reflects not just the stated purposes of the process, but its effect on those relationships.45 The key is to create a system where the purposes of the appraisers, appraisees, and the organization are aligned as much as possible. Errors in Performance Assessment Errors and biases in performance ratings have been widely studied. Extreme errors that lead to wrong decisions can reduce the value of an appraisal system. The most frequently studied errors are halo, leniency, severity, and central tendency. Halo error is when a general impression causes similar ratings on different dimensions, such as rating someone high on job knowledge because they have good social skills.46 (When this kind of error lowers ratings, some call it horns error.) Leniency error is giving overly favorable ratings to an entire group. Author Garrison Keillor tells of Lake Woebegone, where "all the women are strong, all the men are good looking, and all the children are above average." Severity error is the opposite. One study found that subordinates gave harsher ratings to managers who they thought were pushing them harder than their peers 47 Central tendency error is incorrectly giving all ratings near the middle of the scale (e.g., when an instructor gives grades only between B- and B+), despite large differences in student performance. Age, attractiveness, race, and similarity to the person being appraised also affect ratings. However, despite all the attention given to rating "errors," they may often reflect the actual situation. Reality isn't always free of central tendency.48 Can Ability and Motivation Be Improved? Training and Rewarding Appraisers Most large U.S. organizations train their performance appraisers. Rater training might involve error training, which helps raters become aware of the errors WHAT DO YOU THINK? Performance appraisals are often raised or lowered for quite understandable reasons, even though this lowers their accuracy. If you were advising a manager or employee, how would you help them understand the value of being accurate in their appraisal? How would you create a situation to reduce the pressure to "adjust" appraisals for expediency? 116 PART ONE Assess Conditions. Plan and Set Objectives EXHIBIT 4.11 Politics in Executive Appraisals: Five Main Themes 1. The higher one rises in the organization, the more political the appraisal process becomes. 2. Because of the dynamic, ambiguous nature of managerial work, appraisals are susceptible to political manipulation. 3. Performance is not necessarily the bottom line in the executive appraisal process; ratings are affected by: • The boss's agenda. • The "reputation" factor. • The organization's current political climate. 4. Senior executives have extraordinary latitude in evaluating subordinate executives' performance; this can lead to pitfalls: • Superiors fail to set meaningful performance goals and standards. • No communication about the desired style and means of accomplishing goals. • The "good, but not good enough" syndrome, as an attempt to foster continuous improvement. 5. Appraisals used as a "'political tool" to control people and resources. Source: Dennis A. Gioia and Clinton O. Longenecker, "The Politics of Executive Appraisal," Organizational Dynamics 22, Winter 1994. p. 50. Reprinted by permission of the publisher, from Organization Dynamics. Winter 1994. American Management Association, New York. All rights reserved. discussed above; performance dimension training, which helps raters become aware of different performance dimensions, so they avoid making only one global rating; frame-of-reference training, which not only defines dimensions, but also provides critical incidents, practice and feedback on how closely the rater's use of the dimensions and standards fits with expert raters; and behavioral observation training, which strives to improve the way behaviors are observed, not how they are evaluated. One research review found that error training did reduce errors, but an emphasis on "one best rating distribution" reduced accuracy. A combination of training approaches was most effective.49 Considering the importance of the performance feedback, it may make sense to train ra-tees how best to use performance information. Timing Is Everything: When to Assess Performance Performance appraisals are usually conducted once a year, often all at once, or on the anniversary of the date the employee was hired. More frequent appraisals may improve accuracy by reducing memory lapses and linking appraisals to key events. Students usually prefer feedback throughout the course, rather than one final grade. But too frequent appraisals are time-consuming and bothersome, especially if there's nothing new to discuss. Performance changes over time, according to research on baseball pitchers, batters, and sewing-machine operators; high performers can become low performers and vice versa. Baseball players also bat better the year before they sign new contracts (when their performance can affect their pay) than in the year after.50 Experienced workers seem to perform better. Also, those who do best when they strive for maximum performance may not be the best at typical performance, and vice versa.51 Perhaps the best timing is at the end of important tasks or projects, or when key results are supposed to have been completed. CHAPTER FOUR Employee Characteristics 117 EXHIBIT 4.12 Tactics Used to Influence Others Tactic Definition Rational persuasion Using logical arguments and factual evidence to persuade a person that a proposal is likely to attain objectives. Inspirational appeal Making a request that arouses enthusiasm by appealing to the person's values, ideas, and aspirations or by increasing the person's self-confidence. Consultation Seeking the person's participation in planning a strategy, activity, or change, or offering to modify a proposal to deal with that person's concerns and suggestions. Ingratiation Trying to get the person in a good mood or to think favorably of the influencer before asking to do something. Exchange Offering an exchange of favors, indicating willingness to reciprocate at a later time, or promising to share the benefits if the person helps. Personal appeal Appealing to feelings of loyalty and friendship before asking for something. Coalition Seeking the aid of others to persuade the person to do something, or using the support of others as a reason for the person to agree and go along. Legitimating Claiming the authority or right to make a request because it is consistent with organization policies, rules, practices, or traditions. Pressure Using demands, threats, or persistent reminders to get a person to do something. Source: Adapted from Gary Yukl and J. Bruce Tracey, "Consequences of Influence Tactics Used with Subordinates, Peers, and the Boss," Journal of Applied Psychology 11, 1992, p. 526. It's All in How You Say It: How to Communicate Performance Assessment Suppose you've been getting Bs on your class projects, and after lots of hard work you get an A. You are delighted until you notice the instructor's comments: "Congratulations, you finally got a decent grade." Even the best-designed performance measurement system can be ruined by poor communication. Communicating results based on specific and challenging goals can improve performance.52 Those being appraised react more favorably to the process and are more motivated to improve when they believe they have participated in the process, when the message is positive, and when the source of the assessment is viewed as expert, reliable, and attractive.53 Performance appraisal strives to influence people. The appraiser influences the employee to try to improve. The person being appraised influences the appraiser to give good ratings. Several influence tactics are listed in Exhibit 4.12. Which tactics do you use? Are they effective? A recent study showed that managers use different tactics with different groups and purposes. To get subordinates committed to a task, inspirational appeal and consultation had positive effects, while pressure had a negative effect. To get peers to commit, inspirational appeal, consultation, rational persuasion, and exchange had positive effects, but 118 PART ONE Assess Conditions. Plan and Set Objectives coalition and legitimating had negative effects. To get superiors to commit, rational persuasion and inspirational appeal had positive effect. Yet, all three group's ratings of the manager's effectiveness were influenced by rational persuasion.54 To appear competent, be rational. To get commitment, inspire. Firing Up Underperformers When someone isn't performing up to standard, there is usually a progressive set of steps to try to correct the problem. Later chapters will discuss discipline procedures and the termination process. The performance discussion is often the first opportunity to deal with such problems. There is no one best method, but a diagnostic approach can help. Identify whether the problem is with the individual, the job, the organization, or some other factor. Then, focus on changing the behaviors that are causing the problem. If someone says, "My boss is too stubborn," ask, "What does your boss do that seems stubborn to you?" If someone says, "I just can't do it," ask, "What would happen if you tried?" If someone says, "That other group just doesn't understand us," ask, "What has the other group done to make you think that?"55 Sometimes underperformance is due to destructive feedback, which tears down relationships and reduces motivation. Exhibit 4.13 shows how organizations might reduce the tendency to give destructive feedback. Notice that there are two main targets: the feedback givers' self-control, to help them become more aware and responsible for their behaviors; and the feedback givers' self-esteem, to help them feel better about themselves and their role in the feedback process. Paired with each of these targets are two general strategies: Change the "outcomes," so the results of supportive feedback are more desirable; or change the "contingencies," so there is a clearer link between supportive feedback and valuable outcomes. Finally, notice that the strategy differs depending on the type of relationship—the rows of Exhibit 4.13. "Control-dominated" relationships include supervisor, teacher, or leader, and the goal is to instruct or evaluate. "Reward-dominated" relationships include customer, supplier, or competitor, and the goal is to impress or meet expectations. "Affiliation-dominated" relationships include friends, partners, and co-workers, and the goal is to advise, persuade, and build trust. The next lime you find yourself or others engaging in destructive feedback, you might try one of these strategies. How to Evaluate Whether Performance Assessment Works Which techniques work best? Exhibit 4.14 shows how different techniques compare on several dimensions. Managers in the largest U.S. organizations listed the most important performance assessment goals as (1) employees being rated accept the system, (2) employees have a sense of being treated fairly, and (3) employees believe the results are fair.56 We don't have enough evidence to say which techniques are regarded as fairest. A study of college recruiting teams in a large corporation found that leaders' reactions to subordinate ratings were not affected by any agreement of the leaders' self-ratings with the team ratings. Leaders who saw ratings from their own team members instead of average ratings across all teams, however, had higher intentions to discuss the ratings with the team, and found the process more useful and satisfying57 In an investigative division of a government agency, subordinates appraising their managers felt the system was more useful if they believed they had sufficient CHAPTER FOUR Employee Characteristics 119 EXHIBIT 4.13 How Organizations Can Reduce Destructive Feedback Focus on the Source's Self-Control Focus on the Source's Self-Esteem Change Outcomes and Outcome Values Change Feedback Outcome Contingencies Change Outcomes and Outcome Values Change Feedback Outcome Contingencies Control-Dominated Relationships Demonstrate that leadership is enhanced by giving constructive feedback Emphasize source's accountability for the outcome Reward source for effective leadership; model leadership; demonstrate reward power Chance to role-play and receive feedback on leadership skills Reward-Dominated Relationships Enhance rewards to source from constructive feedback and from recipient's success (rather than failure) Source pre-commits to a goal, tracking mechanism, and reward; model benefits of constructive feedback; demonstrate behavior outcome contingencies Link successful behaviors to valued extrinsic rewards; recognize achievements publicly; set goals for giving feedback Attribute successes to source; ensure a close link between desired behavior and positive outcomes Affiliation-Domination Relationships Make giving feedback a required part of group process; give feedback in group forum; model empathy and respect for recipient's feelings Clarify likely outcomes delay— how long it can take before the recipient reacts positively to constructive feedback Demonstrate the value of being empathetic and respectful of others' feelings; learn constructive responses to visceral reactions to stress (e.g., count to 10) Reward team outcomes and cooperative behavior ource: Manual London, "Giving Feedback: Source-Centered Antecedents and Consequences and Destructive Feedback." Human (source Management Review 5, 1995. p. 179. ability to rate performance, and that the manager would use the ratings to improve. The raters' motivations, the trust between raters and those being rated, and whether performance is limited by special circumstances also affect acceptability.58 Does Performance Assessment Affect the Bottom Line? Research linking performance assessment to organizational results is rare, but the research we have suggests it makes a big difference. One study estimated that introducing performance feedback for 500 managers might produce returns of $5.3 million per year.59 At a more modest level, think about workers in a frozen-yogurt or ice-cream store. It's easy for them to make mistakes, especially by filling cups or cones too full. College-age counter workers in a Florida frozen-yogurt shop were given performance information (a scale on the 120 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.14 Evaluations of Performance Assessment Techniques Providing Feedback Allocating Rewards Avoiding Technique and Counseling and Opportunities Minimizing Costs Rating Errors Management by Excellent: Poor: Poor: Good: objectives (MBO) Specific problems, Nonstandard objectives Expensive to Tied to deficiencies, and across employees develop. observations, plans are identified and units make Time consuming reflects job content, comparisons difficult. to use. low errors. Checklist Average: Good-Ave rage: Average: Good: Genera] problems Comparative scores Expensive Techniques available identified, but little available, and development, but to increase job- specific guidance dimensions can inexpensive to use. relatedness and for improvement. be weighted. reduce errors. Graphic rating scale Average: Average: Good: Average: Identifies problem Comparative scores Inexpensive to Substantial areas, and some available but not develop and use. opportunity for information on easily documented errors, though they behaviors/outcomes and defended. can be linked to needing improvement. specific dimensions. Behaviorally Good: Good: Average: Good: anchored rating Identifies specific Scores available, Expensive Based on job scales (BARS) behaviors leading documented, and development, but behaviors, can to problems. behavior-based. inexpensive to use. reduce errors. Essay Unknown: Poor: Average: Unknown: Depends on essay No overall score Inexpensive Good observation topics chosen by available, not development, but can reduce errors, evaluators. comparable across expensive to use. but lack of structure employees. poses a danger. Comparing Poor: Poor-Average: Good: Average: individuals Based on general Overall score available. Inexpensive to Usually consistent. (ranking, forced factors, with few but difficult to defend. develop and to use. but subject to distribution) specifics. halo error and artificiality. WHA1 [INK? Evidence suggests that firms with good performance management do better financially than those without such performance management. What are some possible connections between performance management, productivity, customer satisfaction, and ultimately, financial returns? counter to weigh the yogurt servings), and the store doubled yearly profits by reducing waste.60 Imagine the value of effective performance appraisal where employees handle materials even more valuable than frozen yogurt! For example, one study compared 205 publicly traded companies with "performance management" to 232 such companies without "performance management." Performance management was indicated by setting explicit expectations, coaching and feedback, feedback, and performance-based rewards. Exhibit 4.15 shows a comparison between the two sets of companies on key financial and productivity measures. While encouraging, such results don't necessarily mean that performance management caused the returns. It could be the other way around. Still, the associations are suggestive. CHAPTER FOUR Employee Characteristics 121 EXHIBIT 4.15 Financial Performance Differences Between Companies With Performance Management And Without 10% 8% 6% 4% 2% 0% Financial measures r- CM Total Return on Return on Cash flow Real growth shareholder equity assets return on in sales return investment Productivity measures — © Sales per Income per Real growth employee employee in employees (dollars in (dollars in (Percent) thousands) thousands) □« Companies without Performance Management Q Companies with Performance Management OUTCe: Danielle MCDonald and Abbie Smith, "Performance Management and Business Results", Compensation and Benefits Review, anuary-February, 1995. p. 61. Reprinted by permission of the publisher, from Compensation and Benefits Review, January-February. 995. American Management Association, New York. All rights reserved. Two Sides to Every Story: Handling Disagreements Chapter Sixteen on employee relations discusses discipline and conflict in the workplace. Most organizations have methods to resolve disagreements about performance ratings. Sixty-four percent of large U.S. companies have at least an informal system, while 25 percent have a formal process. Only 10 percent have no appeals process at all.61 A useful dispute resolution process provides ample opportunity for all parties to express their views, involve parties with many different views (such as employees, managers, and union representatives), and clearly states the duties and possible outcomes at each stage of the process. Some research suggests that the performance assessment process is based more on perceptions of "due process" or fairness, than the quest for "truth." There may not be a single objective performance level. Rather, the important thing may be the fairness of the system, and whether people feel that the process treated them properly. Exhibit 4.16 shows the features of a due-process appraisal system that was implemented in a government agency. Compared with employees and managers working in a system without these features, the employees in the due-process system reported the system was more fair, accurate, and satisfying. Managers reported decreased work problems, greater satisfaction with the appraisal system, greater job satisfaction, and less tendency to distort their appraisal ratings. Complying with the Law and Equal Employment Opportunity No law requires employers to assess performance or compels them to conduct assessments in a particular way. As you saw in Chapter Two, however, when 122 PART ONE Assess Conditions, Plan and Set Objectives EXHIBIT 4.16 Improving Due Process in Appraisal Systems Content of Due Process Appraisal System* System feature Employees Managers Adequate notice Training to develop standards X X Training to communicate/clarify standards X X Expectation setting meeting X X Training to negotiate standards X X Training to give feedback o X Training to participate in midstudy feedback session X X Midstudy feedback session X X Fair hearing Training to encourage two-way communication in final review o X Training on how to use form X X Training to conduct self-appraisal X o Conduct self-appraisal X 0 Judgment based on evidence Training to sample representative performance o X Training to keep performance diary X X Training to solicit performance information from employees o X Training to make supervisor aware of accomplishments X o Appraisal form fitted to job X X Manual on appraisal process X X *X = present, O = absent. Source: M. Susan Taylor; Kay B. Tracy; Monika K. Renard; J. Kline Harrison; and Stephen J. Carroll. "Due Process in Performance Appraisal: A Quasi-Experiment in Procedural Justice," Administrative Science Quarterly 40. 1995. p. 509. decisions such as pay, promotion, hiring, and layoff adversely affect protected-group members, an employer may defend the decisions as being based on merit, rather than the protected characteristics of the individual. The courts examine whether merit is defined in a biased way. John Chamberlain, a 23-year employee and manufacturing manager at Bissell, was denied a wage increase because of performance problems, and his poor performance and attitude were discussed with him. Later, he was fired. He successfully sued Bissel for more than $61,000, in part because no one stated that he was in danger of being fired during the performance interview.62 Several reviews of court cases suggest the following guidelines: 1. Base performance appraisals on job analysis. 2. Provide raters with written performance standards based on job analysis. 3. Train raters to use the rating instrument properly. 4. Provide a formal appeal mechanism and review of ratings by upper-level personnel. 5. Document evaluations and instances of poor performance. 6. Provide counseling and guidance to assist poor performers to improve.63 CHAPTER FOUR Employee Characteristics 123 EXHIBIT 4.17 Intel's Performance Assessment System Intel Inc., a microprocessor company in Santa Clara, California, has a comprehensive appraisal system. Sandra Price, a manager of strategic planning and projects, describes it: "In addition to getting a formal, annual review written to their file, managers meet to rank workers against peers performing similar jobs. A good ranking group is 10 to 30 people, and each individual is ranked on his or her contribution to the department, the division, and organization. A 'ranking manager' runs the meetings where this takes place, keeps people on track, maintains objectivity, and drives the output—a ranked list. "Before managers go into the meeting, they fill out a short evaluation form to make their approaches similar. It captures what happened during the year—the individual's accomplishments. "As part of the ranking process, workers go through a rating that measures how they did against the requirements of their job. After surveys found that a five-level rating system equated too closely to an A, B, C, D, E system, in which C isn't as positive, we changed it to three simple ratings—outstanding, successful, or improvement required. "We also evaluate performance trends—how an individual is trending against his or her peers. Are individuals moving faster? Taking new responsibility? Equal to peers? About as expected? Slower? This trend evaluation is a new element, designed to give people as early a notice as possible when their performance starts to slow down. "Intel also conducts 'focals'—a point in time when we evaluate an employee's individual performance, contribution to the group, development needs for future growth and appropriate compensation. "We use past performance to project the future, and ask the employee to decide with the manager the areas he or she needs to improve in or wants to concentrate on over the next year. Improvement relates to defined objectives—we're an MBO company. Do employees have the skills and knowledge that enable them to meet objectives and grow as they desire? The worker and the manager agree to the skills that are required, and together devise a plan that will use both formal and informal learning systems to help the worker succeed. "We leave it up to the manager to determine how often they meet and discuss progress, and we place the responsibility for improvement on the employee. They own that." Price, an attorney, believes the comprehensiveness of Intel's evaluation process is one reason why the company has "for its size, been relatively untouched by litigation related to wrongful termination." Source: Katherine G. Hauck, "Failing Evaluations." Reprinted with permission of Human Resource Executive, Sept. 1992, p. 52. For subscription information, call (215) 789-0860. Putting It All Together: The Role of the Human Resource Manager Organizations rely on human resource managers to help them make the choices described in this chapter and to design a complete performance appraisal system. There is no one best way as our discussion has shown. Every performance assessment situation is different, and organizations must decide how much time and effort are worth the payoff. Human resource professionals usually assist with the design, administration, and evaluation of appraisal systems. See Exhibit 4.17 for the appraisal process at Intel, one of the world's most innovative and successful makers of microprocessors for personal computers. Notice how this system integrates the decisions we have discussed in this chapter. ORE THAN JUST PERFORMANCE: EMPLOYEE WITHDRAWAL Performance is important, but a person has to show up and stay at work for performance to have real value. Even the best-performing employees are not valuable if they miss work frequently or leave soon after they are hired, so organizations assess withdrawal behaviors. Work withdrawal takes place when employees stay in the organization, but they minimize their work time by being absent, shirking their duties, and socializing with others. Job withdrawal occurs 124 PART ONE Assess Conditions. Plan and Set Objectives when employees leave the job or organization. The most visible withdrawal behaviors are absenteeism and separations. Today's absent employees are more likely to leave in the future, though the relationship is weaker when jobs are scarce.64 Absenteeism Absenteeism is often punished by discipline and dismissal because it is such an observable and obviously detrimental behavior. The formula for absence used by the Bureau of National Affairs (BNA) is: Worker days lost through job absence during the month Average number of employees x Number of work days in month This formula reflects both the number of absent employees and the duration of their absence. BNA does not include absences for jury duty, scheduled disciplinary time off, long-term disabilities (after the first four days), or excused absences scheduled in advance. Absence is higher in larger and nonprofit organizations.65 However, some decisions require measuring absenteeism differently. - To capture the total cost of absence, orga- Absenteeism is the frequency and/or duration of nizations might include all missed work work time lost when employees do not come to days, regardless of the reason. Or, to cap-work. Attendance, the opposite of absenteeism, ture intentionally poor attendance, they is how often an employee is available for work. might include only absence the employee can control.66 Psychologists have long studied the causes and consequences of absence; they have found that it is affected by the employee's ability and motivation to attend and by on-the-job and off-the-job factors.67 Study the diagnostic model of employee attendance in Exhibit 4.18. Note that perceived ability to attend can be reduced by attendance barriers such as illness and accidents, family responsibilities, and transportation problems. Attendance motivation is affected by organizational practices (such as attendance penalties or rewards); absence culture (whether absence is considered acceptable or unacceptable); and employee attitudes, values, and goals. Absence frequency and duration are related to work satisfaction, although the relationship is not extremely strong. The greater the proportion of females, the stronger the relationship between absence and satisfaction.68 Notice how information on several employee characteristics is needed to effectively manage HR outcomes. Absence-control programs must address the broad causes of absenteeism. Organizations influence absence motivation by disciplining absent employees, verifying excuses, communicating absence rules, and rewarding good attendance records. Dealing with other absence causes, such as family responsibilities, is much more difficult, but very important. Indeed, such tactics can backfire, as employees prohibited from one form of absence may gravitate to another. 69 This may explain why surveyed employers felt that the best control mechanism was a "paid-leave bank," where employees have a bank of paid days to use for any purpose. Once the bank is used up, the employee loses pay for every absent day. If all the days are not used, the employee receives the equivalent of the unused days' pay at the end of the year. The Commerce Clearing House estimated the 1995 cost of absenteeism to have increased to $668