week4
mayrob56
chapter4.docx
The Importance of Performance Appraisals
One-panel comic of a woman reading to her daughter before bed. The girl says to her mother, "I think the Little Engine was probably worried about his performance reviews."
Throughout your life, people will make life-changing evaluations of your performance. From elementary school through college, on the playing field and in your community, from your first part-time job to your adult career, others will give you tests and evaluate and compare your performance, the results of which will determine your advancement (or failure to advance) to the next phase of life.
Within organizations, assessment of employees' performance tends to be perceived as a necessary evil that neither managers nor staff particularly like. Many employees fear that even one low performance rating could affect their pay or damage their career. Even more frightening is the prospect of receiving low ratings from a manager who doesn't ever directly observe or work with the employee but uses secondhand information or personal biases to make his or her evaluations. Sadly, this frequently happens.
Consider This: How Do You Feel About Being Evaluated?
•Think about one or more occasions in which you were being evaluated. It could be at work, school, a playing field, or elsewhere.
•Describe your feelings and thoughts before receiving these evaluations. Were you anxious? Were you looking forward to the evaluations?
•Describe your feelings and thoughts while receiving these evaluations. Were you surprised? Upbeat? Interested in receiving feedback? Actively involved? Passively receiving the information? Feeling under attack?
•Describe your feelings and thoughts immediately after these evaluations. Were you excited? Flattered? Humiliated? Angry? Defensive?
•What effects did these evaluations have on your personal, social, or professional life? Did they make you a better person in any way? Explain.
Managers also suffer anxiety when completing performance appraisals. Most often, they worry that criticisms, no matter how small, might provoke negative reactions, ranging from disappointment and frustration to anger and hostility. These emotions can put strain on the manager-employee relationship or cause the employee to become less motivated or even to quit. As a result, managers tend to shy away from providing negative performance feedback, which of course negates accuracy.
Consider This: How Do You Feel About Evaluating Others?
•Think about one or more occasions in which you had to evaluate or give feedback to someone. Again, it can be at work, school, or a playing field. Personal and social settings can also be used for this exercise.
•Describe your feelings and thoughts before you gave your evaluation or feedback. Were you anxious? Hesitant? Excited?
•What were your primary concerns? The fairness of your evaluations? The reactions of the people you were evaluating? The repercussions of your evaluation for yourself and/or the person you were evaluating?
•Describe the settings in which you had to communicate your evaluations. Was it face to face? On the phone? Through e-mail? In a written report?
•Describe the content of your feedback. Was it positive, neutral, or negative?
•What were the reactions of the people you were evaluating? Was your feedback appreciated? Tolerated? Rejected?
•How did you manage or leverage the reactions of the people you were evaluating? Did you involve them? Did you ask for their input?
•Describe your feelings and thoughts after giving your evaluations. Were you stressed? Drained? Relieved? More confident about your feedback communication skills and your ability to accurately assess others' performance?
•What effects did your evaluations have on others' personal, social, or professional lives? Did they make them better? Did they help them advance their careers? How did they affect your relationships with the individuals you evaluated? Explain.
So then, if everyone dislikes performance appraisals, why keep doing them? For one thing, unmanaged performance is chaotic and random. Employees' work needs to be aligned with the organization's overall goals, and clear performance feedback helps everyone to know if this is indeed happening. In fact, a well-designed performance appraisal system should not only provide employees with rich feedback but should also communicate clear performance expectations and include information that will help them perform at the highest level possible (Pulakos & O'Leary, 2011). Appraisals that meet each of these concerns will enable the organization to further its mission to succeed. The question, then, is not whether to keep doing performance appraisals but how to make them most effective.
Uses of Performance Appraisal
A performance appraisal is the formal process through which employee performance is assessed, feedback is provided to the employee, and corrective action plans are designed. Organizations conduct performance appraisals for the following reasons:
Photo of an evolution form. The evolution check boxes read unsatisfactory, marginal, satisfactory, very good, and outstanding.
Performance appraisals have several benefits, including increased worker motivation. Employees are more productive and satisfied when they know what is expected of them and how they can achieve these results.
1.To evaluate performance objectively. Organizations need some sort of system to measure the value of each employee's performance. These measures must be objective and must allow for consistent comparison of performance of people with the same job function.
2.To increase worker motivation. Appraisals provide employees with specific feedback regarding their strengths and weaknesses. When workers know what they should be doing, how they actually are doing, and how they can improve, they are often motivated to perform better.
3.To make administrative decisions. Managers rely heavily on data from performance appraisals when making decisions about employee raises and bonuses, promotions, demotions, or even terminations. Employees must perceive these decisions as fair and free from bias; a good performance appraisal will facilitate those favorable perceptions.
4.To improve organizational performance. Performance appraisals are essential to improving organizational performance (DeNisi & Sonesh, 2010). They pinpoint skill deficiencies in specific parts of the organization, helping managers to focus their training and selection efforts. Appraisals also enhance an organization's opportunities for success by identifying poor performers, which not only helps weed out subpar personnel but also motivates top performers to keep their performance levels high.
5.To establish training requirements. Appraisal data provide insight into workers' knowledge, skill, and deficiency levels. This information helps managers establish specific training objectives, update or redesign training programs, and provide appropriate retraining for specific employees.
6.To enhance selection and testing processes and outcomes. An important use of performance- appraisal data is to establish the criterion-related validity of selection tests. Recall from chapter 3 that criterion-related validity establishes a predictive, empirical (number-based) link between test scores and actual job performance by correlating applicants' employment test scores with their subsequent performance on the job. How can the predictive capacity of a test be determined if the organization does not design and implement objective performance measures and procedures? It would have no accurate data to correlate test scores with, which would hinder its ability to design and use accurate tests and implement effective selection processes.
4.2 Approaches to Measuring Performance
I/O psychologists have identified a number of techniques to measure employee performance. These measures can be either objective or subjective. Generally, what is measured and how it is done depends on the type of work an employee performs. Some jobs, such as sales and assembly-line work, have objective outcome measures (sales revenue, number of pieces assembled), whereas others are more subjective (wait staff performance, art design work).
Objective Performance Measures
Concepts in Motion:
How to Evaluate Business Performance
<span id="w39039" class="werd"> </span>
Objective performance measures are quantitative measures of performance that can be found in unbiased organizational records. They are generally easy to gather and include two types of data: production measures (such as sales volume, number of units made, number of error occurrences) and personnel measures (such as absenteeism, tardiness, theft, and safety behaviors). Both measures are also usually evaluated according to the quality of performance.
However, objective measures can be deceivingly simple. Consider the performance of two sales professionals in an insurance company. Over the course of a year, Salesperson A sold 500 policies and Salesperson B sold 1,000. According to this data, Salesperson B appears to be the better salesperson—twice as good, in fact. However, if we examine the quality of each worker's performance, we might learn that Salesperson B sold unprofitable policies, resulting in a $1 million loss for the company. Salesperson A, on the other hand, sold very profitable policies, resulting in a $1 million profit.
Alternatively, Salesperson B may have focused on selling more policies while cutting corners on after-sale service and follow-up, which could have resulted in dissatisfied customers. On the other hand, Salesperson A may have invested more time per sold policy on such interactions. Although after-sale service and follow up may not be directly measured or rewarded by the organization, these customer interactions can help build the reputation of the organization and are known to result in more satisfied customers returning for additional products and referring others. Repeat business from and referrals by satisfied customers are significantly less costly for an organization to generate than is building new clientele. However, these additional sales may not be easily attributable to Salesperson A if the returning or referred customers are assigned to another salesperson. As you can see, evaluating worker performance by quantity alone without also adding in the quality component is not a wise course of action.
Photo of a man pointing to his wristwatch.
Types of personal data like absenteeism and tardiness are used to evaluate workers' performance and provide an objective measure of an employee's success or failure on the job.
Unfortunately, even after controlling for performance quality, objective data may not provide an accurate or complete picture of an employee's performance. Many factors beyond workers' control can limit their ability to perform their best. Looking more closely at our two insurance salespeople, we might discover that the difference in sales volume could be attributable to the location of each employee's branch office. Salesperson A could work in a small Midwestern town while Salesperson B works in Manhattan. Thus, Salesperson A could have captured a larger market share of his designated region than Salesperson B, even though he sold fewer policies. Alternatively, perhaps Salesperson B was assigned an easyto- sell policy because she was a new employee, while Salesperson A, as a veteran employee, was assigned a hard-to-sell but very lucrative policy. As you can see, accurate performance evaluations require more than a cursory look at sales and production numbers, although adding manager interpretation into the mix does make objective performance measures more subjective.
Personnel data, another objective measure, includes such components as theft, tardiness, absenteeism, safety behaviors, and rate of advancement. Though not typically related to a worker's ability to do the job, these elements do indicate job success or failure. Many jobs, such as teachers, customer service representatives, and bank tellers, require consistent daily attendance. Thus, absenteeism and tardiness are often used to evaluate these workers' performance. Other jobs, such as machine operators, assembly-line workers, and truck drivers, have serious safety risks. With jobs such as these, it makes sense to keep count of employees' accidents and safety incidents and use them as objective measures of performance.
As with any type of objective data, however, taken on its own, personnel data can be misleading. Once again, circumstances outside the worker's control could affect performance. Sick children, a death in the family, or transportation troubles may affect an otherwise superior employee's ability to come to work. Similarly, a workplace accident could have been caused by faulty company equipment, not worker error. Because you now understand the limits of objective data, let's turn to subjective performance measures, their limitations, and ways to keep this data fair and free from bias.
Subjective Performance Measures
The allure of objective performance measures has to do with their ability to provide biasfree information on all workers across a specific job. Of course, we now know that objective data can still be misleading. Further, most jobs require much more than simply looking at sales or production numbers, because most jobs are composed of a complex web of tasks, not all of which can be measured objectively. A teacher's performance must be made up of more than his or her students' test scores, just as a police officer cannot be evaluated solely on the number of arrests he or she makes each month. To address these issues, I/O psychologists created subjective performance measures, which rely on human judgment and are thus exposed to some degree of subjectivity. To reduce bias, which is always a factor in subjective measurements, evaluators must base their ratings on observations of worker behaviors that are critical for successful job performance. Further, these behaviors must be identified through an accurate job analysis.
Interestingly, research shows only a small correlation between objective and subjective performance measures, which suggests that they measure different aspects of worker performance (Bommer, Johnson, Rich, Podsakoff, & McKenzie, 1995). Thus, the two sets of measures are complementary and should be utilized in conjunction whenever possible.
Organizations use many types of subjective performance measures, ranging from manager-composed performance narratives to numerically oriented rating scales. Each method differs in complexity as well as the amount of time required to create and implement it. The next section provides a brief review of some common subjective performance measures.
Written Narratives
Photo of a businessman writing with a pen.
Reference letters are one type of written narrative used by managers to highlight an employee's performance over a certain period of time. Written narratives can also be used for performance evaluation, highlighting the worker's strengths and weaknesses.
With the written narrative, one of the easiest performance measures to develop, the manager writes a paragraph or two summarizing an employee's performance over a certain length of time. An example of a written narrative is a reference letter written by a supervisor for an intern at the end of an internship. When used for performance measurement, managers often share specific examples about the worker's strengths and weaknesses, which the worker can then use to help improve his or her performance during the next appraisal cycle.
Although written narratives are quick and easy, they have a number of drawbacks. First, every manager will set different evaluation standards, making it impossible to compare workers with different managers. As a result, the written narrative should not be used to make decisions about compensation, promotions, or layoffs. Second, managers vary in their level of written communication skills. Some may use ambiguous, incomplete, or misleading language, which can mean that the employee could misinterpret or not understand the manager's feedback. Finally, managers are often reluctant to address poor performance in a straightforward manner and sometimes deliberately write the narrative to cast a positive light on negative behavior.
The drawbacks of the written narrative have prompted I/O psychologists to develop a number of techniques both to improve the objectivity of subjective performance measures and to reduce managerial biases.
Rank Ordering
Cartoon of a man balancing on top of a chair which is placed on top of a desk. There are three judges seated at another desk holding up sings with the number "5" on each to judge his balancing act.
Rank ordering requires no forms or instruments and is the easiest way to evaluate workers. Managers simply rank their employees from best to worst. Some managers have the tendency to evaluate employees similarly. Rank ordering provides muchneeded differentiation, even though the small differences between median employees still make ranking employees a challenge for managers. Rankings also do not provide workers with performance feedback, which means they are not useful as tools for self-improvement or training guidance. Because of their limitations, rankings should be used only during periods of downsizing, reorganization, or any other situation in which understanding a worker's relative standing to other workers would be valuable.
Paired Comparison
As with rank ordering, the paired comparison technique requires the manager to evaluate a worker's performance in comparison to the other workers on the team. In a systematic fashion, the manager compares one pair of workers at a time and then judges which of the two demonstrates superior performance. After comparing all the workers in all possible pairings, the manager then creates a rank ordering based on the number of times each worker is the better performer of a pair. For example, using the formula (N(N − 1)/2)N to determine the number of discrete pairings in a group, a manager with 10 employees would need to make 45 paired comparisons. A manager with a team of 20 employees would need to make 190 comparisons. As you can see, the number of pairs goes up quite quickly as the size of the team increases. For this reason, paired comparisons are advantageous only for smaller groups.
Like general rank orderings, paired comparisons do not provide performance feedback. However, they are generally simpler to use because managers need only compare one employee pair at a time instead of the entire work team. Organizational leaders should keep in mind that rankings are not standard across the entire workplace. The lowest ranked member of a high-performing team might, for example, actually perform better than the highest-ranked member of a poorly performing team.
Forced Distribution
When an organization needs to evaluate a large number of employees, forced distribution is a viable option. With this technique, managers place employees into categories based on pre-established proportions. A typical performance distribution uses the following performance categories and proportions:
Superior 10%
Above average 20%
Average 40%
Below average 20%
Poor 10%
Using this distribution for a team of 100 workers, a manager would identify the top 10 employees (10%) and the bottom 10 employees (10%) and place them in the superior and poor categories, respectively. From the remaining 80 workers, the manager would then select the next 20 highest performers (20%) for the above-average category and the next 20 lowest performers (20%) for the below-average category. The final 40 workers (40% of the original 100) would fall into the average category. The lowest performance group would then be reprimanded, put on probation, or terminated. This approach is most commonly associated with Jack Welch, former CEO of General Electric. GE eliminated the lowest 10% of performers every year using this method.
Obviously, one of the major drawbacks of forced distribution is that it assumes that worker performance follows a normal distribution (some high performers, some low, most somewhere in the middle). This method makes no concessions for teams filled with superior performers or, conversely, teams fraught with poor performers. Further, it makes no distinctions among workers in a category; all average workers, for example, are simply considered average. Finally, as with rank ordering and paired comparisons, forced distribution can add artificial luster to "superior" members of poor-performing teams, or unfairly tarnish "poor" members of a high-performing team.
Graphic Rating Scale
Graphic rating scales are the most commonly used method for rating worker performance. Managers observe specific employee behaviors or job duties along a number of predetermined performance dimensions, such as quality of work, teamwork, initiative, leadership ability, and judgment. Then, the manager rates the quality of performance for each dimension on a scale ranging from high to low performance. Looking at Figure 4.1, you can see that this employee received a below-average rating on teamwork.
Figure 4.1: General graphic rating scale
Each point on a graphic rating scale, called an anchor, is defined along the continuum. Anchors can vary in number, description, and depth of detail and can be stated in numbers, words, longer phrases, or a combination of these forms. Typically, the manager rates workers' performance for each anchor using the five-point rating scale, although sevenor even nine-point scales are not uncommon.
Graphic rating scales are versatile, inexpensive, and quickly made. However, in order for the manager to make clear, accurate distinctions in worker performance across different dimensions, care must be taken to create specific and unambiguous anchor descriptions. For example, in Figure 4.2, scale A uses only qualitative anchors and requires the rater to place a check mark at the point that represents the worker's current performance level. This is a poorly designed rating scale because the rating anchors are left undefined. Similarly, scales B, C, and D include both verbal and numerical anchors, but ratings rely solely on manager judgment. Of course, this can problematic, because one manager might judge his or her employees more or less stringently than another. Standard measures allow for clear comparisons across workers, even if they have different managers.
Figure 4.2: Examples of different graphic rating scales
Some organizations ask managers to provide written examples that support their ratings of employees for each performance dimension and/or for their overall level of performance. By combining both their rating scores and written feedback, employees learn both how their performance compares to the company's expectations and what their current strengths and weaknesses are. This allows the company to set goals or devise training strategies and the employee to seek self-improvement or educational resources to improve his or her performance.
Behaviorally Anchored Rating Scale
First proposed by Smith and Kendall in 1963, the behaviorally anchored rating scale, also called BARS, attempts to evaluate workers' performance of very specific behaviors critical for job success. These behaviors are established using the critical-incidents job analysis technique discussed in chapter 2.
Developing a BARS can be a long and difficult process. To begin, a group of supervisors familiar with the job identifies both the performance dimensions (quality of work, teamwork, initiative, etc.) that should be measured and observe critical incidents that exemplify both effective and ineffective job performance. Another group of subject matter experts transforms the list of critical incidents into behavioral statements that describe different levels of performance. A final group evaluates the behavioral statements and assigns a numerical value scale to each. See Figure 4.3 for an example.
Figure 4.3: Example of a behaviorally anchored rating scale (BARS)
One positive feature of the BARS approach is that the behaviorally defined anchors are very explicit as to what performance criteria are being measured. This makes it much easier for managers to distinguish between high and low performers. Additionally, because the rating scale is standardized, managers can compare BARS performance ratings across individuals and teams. Further, workers perceive BARS to have high face validity, which reduces negative reactions to low ratings.
Despite the advantages, the significant time investment needed to develop BARS means that most organizations do not employ this technique. Additionally, this method's overall rating quality is still dependent upon each manager's observational skills (or lack thereof). Finally, research shows that the BARS is no more valid or reliable than any other rating method, nor is it more successful at decreasing rater error (Landy & Farr, 1980).
Behavioral Observation Scale
The behavioral observation scale (BOS) is similar to BARS in that both use critical incidents to identify worker behaviors observed on the job. The biggest difference between BOS and BARS is the rating format. Instead of quality, BOS rates the frequency with which a worker is observed to perform a critical job behavior (see Figure 4.4 for an example). Frequency is typically measured on a five-point rating scale, comprising numerical (0–10% of the time, 11–20% of the time, etc.) or verbal assignments (sometimes, always, never, etc.) or a combination. The ratings are aggregated across all behavioral statements to establish a total score. Some researchers have tried to determine which method, BARS or BOS, is superior, but the research has been mixed and inconclusive.
Figure 4.4: Example of a behavioral observation scale (BOS) for a bank teller
4.3 Sources of Performance Appraisal
Concepts in Motion:
How Can Business Performance Be Evaluated?
<span id="w41779" class="werd"> </span>
The goal of performance appraisal is to accurately measure employees' performance. To do so, raters must directly observe an employee's actions and behaviors. In many cases, managers cannot directly observe their employees often. A police chief, for example, cannot accompany all of his or her officers as they perform their daily patrols, nor can a school principal sit in classrooms all day long. Who, then, should evaluate such workers? For many jobs, input from a variety of sources (for a police officer, his or her partner or members of the community; for a professor, students or fellow faculty members) helps to create a more accurate, well-rounded performance appraisal. The following section describes the most common sources of input.
Supervisor Evaluation
Supervisors are the most common source of input for a performance appraisal, and rightly so. After all, managers are in the best position to evaluate employees' performance as it relates to the organization's objectives. Further, because managers are responsible for recommending rewards and punishments, they must be able to tie their evaluations to employees' performance. Without this link between performance and rewards, employees can become less motivated, resulting in poorer performance. Indeed, research shows that supervisors' performance ratings are more strongly correlated to employees' actual performance than any other rating source (Becker & Klimoski, 1989).
Peer Evaluations
Peer evaluations, or evaluations made by one worker about a coworker, are common in jobs that require employees to work as part of a team. Peer feedback can be especially insightful. Coworkers often understand the job in greater depth than managers and can analyze how team members' behaviors affect each other and contribute to the team's success. Similarly, peer ratings on the dimension of leadership effectiveness provide a valuable perspective into a worker's leadership skills and abilities.
Photo of a businessman sitting in his office and analyzing a printout.
Supervisors are in the best position to evaluate employees' performance as it relates to the organization's objectives. Evaluations completed by managers are typically more accurate than any other rating source.
How do employees respond to peer evaluations? Generally, reactions are mixed. In some situations, workers are appreciative because their peers are the only people who ever directly observe their performance and are therefore the only ones who can provide accurate evaluations. Further, because peer ratings are nearly as accurate as supervisory ratings, they are excellent guides for self-improvement (Harris & Schaubroeck, 1988). On the other hand, workers may question the validity of a negative peer review, something which can detrimentally affect their future performance. DeNisi, Randolph, and Blencoe (1983) found that workers who received negative peer-rating feedback went on to hold more negative perceptions of group performance, group cohesion, and overall satisfaction during the subsequent team task. Conversely, positive peer feedback did not significantly affect any of these variables on the next task. Peer ratings, therefore, should serve as a supplemental, not the only, source of a worker's performance evaluation.
Subordinate Evaluations
Subordinates are uniquely capable of assessing their manager's actions and effectiveness across a broad range of dimensions, including delegation, coaching, communication, leadership, and goal setting. The process of subordinate evaluation, also called upward feedback, involves performance evaluation of a superior by his or her subordinates.
Some research supports the notion that upward feedback can lead to improved management performance. In one study, subordinates rated their managers on a number of different dimensions (quality, fairness, support, and communication). Superiors who received low to moderate ratings showed significantly more rating improvements six months later than those who received high ratings (Smither et al., 1995). Further research shows that managers can improve their ratings even more by discussing upward feedback with their subordinates (Walker & Smither, 1999).
Confidentiality is critical to ensure accurate subordinate evaluations. Many employees fear the potential repercussions of providing managers with negative feedback and will artificially inflate their evaluations if they know that the manager will be able to identify them (Antonioni, 1994).
Self-Evaluation
Close-up of a self evolution form and a pen.
Although self-ratings show more leniency and greater bias than ratings by supervisors or peers, employees who complete self-evaluations feel more engaged in the appraisal process, which can decrease defensiveness regarding the final ratings.
Employees who complete self-evaluations, or evaluations of their own performance, feel as though they have a voice in the appraisal process, which in turn increases their acceptance of and decreases their potential defensiveness about the final ratings. Although typically used as supplemental evaluative data, self-ratings are especially useful with workers who work alone or independently.
Generally, self-ratings show more leniency, less variability, greater bias, and less agreement with those provided by supervisors, peers, or subordinates (Harris & Schaubroeck, 1988). These differences could stem from the worker's use of a different evaluative standard (Schrader & Steiner, 1996), but research has identified several ways to make selfratings more accurate. First, workers can be told that their self-ratings will be compared against objective performance criteria. Second, the organization can make it clear that the self-evaluation will be used only for self-developmental purposes (Meyer, 1991). Third, educating both superiors and subordinates on the rating criteria and appraisal process leads to greater agreement between supervisor ratings and self-ratings (Williams & Levy, 1992).
360° Appraisals
360° appraisal is a multisource evaluative process; it utilizes performance input from many different viewpoints. In this process, a manager might, for example, receive feedback from his or her supervisor, peers, subordinates, and internal or external customers as well as doing a self-evaluation. Normally, each source uses a rating scale to evaluate the manager's current proficiency level for a predetermined set of job duties and/or leadership dimensions (coaching, delegating, communicating, etc.). After all ratings are complete, they are compiled in a report and shared with the manager.
Over the last 20 years, 360° appraisals have grown significantly in popularity. In the mid- 1980s, fewer than 10% of companies used this method to evaluate managers (Bernardin, 1986). Today, even though there is no exact percentage, it is likely that every Fortune 1000 company has had some experience with conducting 360° appraisals. Interestingly, there is no consensus on how exactly to use these evaluations. Some believe they are appropriately used in making administrative decisions (Church & Bracken, 1997), but most disagree, suggesting they be used only for management development purposes (Antonioni, 1996; Coates, 1996).
I/O psychologists recommend a number of practices to increase the effectiveness of 360° appraisals. First, both raters and the manager should receive instructions on how to interpret the different performance dimensions and the rating scale. Second, all participants must be explicitly told that feedback will only be used for the purpose of manager development. Further, maintaining rater anonymity tends to prompt subordinates to view 360° appraisals more positively, although, interestingly, managers tend to prefer that employees be held accountable for their ratings (Antonioni, 1994). Finally, to ensure the highest quality, a 360° appraisal program must include skilled coaches to help managers interpret and use their feedback to create goals and courses of developmental action (Coates, 1996; Crystal, 1994).
Consider This: Seeking Feedback
•Most of us often seek feedback on our performance in various life domains, whether at work, at school, at home, in sports, or at church. Where do you normally seek or get feedback on your performance in each domain of your life?
•What are the advantages and disadvantages of each source of feedback?
Find Out for Yourself: Performance Measures and Sources
•Visit the following websites of the human resources offices at University of California, Berkeley and University of California, Davis for templates of performance measures that utilize feedback from different sources as well as examples of performance evaluations using these measures: ◦http://hrweb.berkeley.edu/performance-management/forms (http://hrweb.berkeley.edu/performance-management/forms)
◦http://www.hr.ucdavis.edu/forms/Perf_Eval/000 (http://www.hr.ucdavis.edu/forms/Perf_Eval/000)
4.4 Sources of Rating Error and Bias in Performance Evaluation
Performance appraisal relies on the assumption that human judgment is capable of some degree of accuracy. However, humans are not objective observers. We can never be completely certain that our judgments are free from error or personal biases. Often, of course, errors are unintentional. In the workplace, rating errors can occur if managers do not observe workers' performance or if they do not use the rating scale correctly. More insidious, managers can also harbor unacknowledged biases against certain types of workers. At other times, rating errors are intentional. Managers can deliberately inflate a poorly performing employee's ratings because they don't want to jeopardize their relationship with the employee or because they don't want to cause negative reactions. I/O psychologists have worked hard to identify, understand, and correct sources of rating error.
Types of Rating Error
In order to improve accuracy, one must first identify and understand error. With performance appraisals, rater errors typically fall into three major categories: observational errors, distributional errors, and rating scale errors. In this section, we review the specific rating errors associated with each category and discuss how these errors reduce performance appraisal accuracy.
Observational Errors
As stressed throughout this chapter, appraisals must be based on thorough observation of an employee's performance in order to be accurate. Without direct observation, managers may rate employees based on such unreliable sources of information as general impressions, observations from past ratings, or hearsay.
Photo of a businesswoman looking through binoculars.
To increase accuracy, appraisals must be based on thorough observation of an employee's performance. Otherwise supervisors may incorrectly rate employees based on general impressions, past ratings, or hearsay.
Even if managers are able to observe workers' performance, they are often unable to remember more than an employee's most memorable performance accomplishments (or failures). An average appraisal cycle lasts up to twelve months, and, as you might expect, a manager will remember and thus more strongly emphasize recent performance, an effect called recency error. Of course, when a recency error occurs, a worker's ratings do not accurately represent his or her overall performance throughout the entire appraisal cycle.
The best way to overcome recency error is simple: Reduce the amount of performance information the manager needs to remember. One practical way for managers to do this is to shorten the appraisal cycle by conducting more frequent appraisals throughout the year, instead of once annually. Additionally, managers can improve recall by keeping a detailed performance log for each employee, especially at the beginning of the rating cycle.
Distributional Errors
Within an organization, evaluation standards tend to differ from manager to manager. Significant error occurs if managers inaccurately distribute rating scores along the rating scale. For example, some managers may clump everyone together with average scores. Others may be overly positive or afraid to give anyone negative scores. A third group of managers may be too stringent and thus give most of their employees low scores. These faulty judgments are called distributional errors.
First, one of the most common forms of rating error in general is leniency error. In this situation, managers have a low performance standard and rate their employees higher than their performance deserves. A graphic representation of a lenient manager's rating scores will show that they tend to cluster on the positive end of the distribution. Even though the high scores could be a reflection of a truly high-performing team, research tells us that such a result is unlikely. Normal worker performance distribution tends to follow a bellcurve pattern, with some workers falling at the high and low ends but most clustering somewhere in the middle. Leniency error is very obvious when it occurs. For example, one study found that out of 12,000 federal employees, 85% received ratings at the superior level on their performance appraisals, whereas only 1% received ratings below the fully successful level (Marrelli & Tsugawa, 2009). This distribution is extremely unlikely to be accurate.
Research shows that rater personality characteristics such as agreeableness and conscientiousness could be linked to leniency error. In an experimental lab study, people with low conscientiousness and high agreeableness rated their peers more positively, regardless of actual performance (Bernardin, Cooke, & Villanova, 2000). As previously discussed, a manager's reluctance to give negative feedback is another common impetus for rating leniency. Most managers want to develop and maintain positive relationships with their subordinates, and offering positive feedback during a performance discussion is certainly more enjoyable than the alternative. Unfortunately, being lenient not only fails to challenge and improve workers' future performance, but also makes terminating a poor performer legally difficult.
The second form of distributional error is central tendency. Central-tendency error occurs when a manager is reluctant to rate employees as either superior or inferior. As a result, ratings scores cluster around the middle of the performance scale. Third is severity error, in which a manager holds excessively high standards and rates employee performance as lower than it actually is. A severe manager's ratings scores will cluster around the low end of the performance scale. Although these two distribution errors are less common than leniency error, they are just as problematic. Because they fail to address the real differences among employees' performance, scores tainted by severity and central tendency error are worthless for making employee decisions.
Find Out for Yourself: Distributional Errors and Biases
•The next time you participate in a group activity or team project, rate each of the group members (excluding yourself) on his or her performance and contribution to the project on a scale of 1–10.
•Ask each team member to evaluate each of the other members (excluding himself or herself).
•Are your evaluations consistently more lenient, more stringent, or comparable to your team members' evaluations?
Rating Scale Error
Sometimes, performance appraisal errors occur because the rater does not know how to use the rating scale correctly. In other cases, a manager's general opinion about a specific employee can color his or her ratings of all performance dimensions for that employee (Lance, LaPointe, & Stewart, 1994). This tendency, called the halo effect, is the most common form of rating scale error and can either artificially inflate or deflate ratings. For example, if a manager believes one of his or her subordinates is extremely smart, he or she might transfer that positive opinion to evaluations of other performance areas, such as collaboration, ethics, and loyalty. Basically, then, managers who rate workers high (or low) on one significant dimension will go on to score them high (or low) on all other dimensions on the appraisal, especially if the other dimensions are not well-defined or directly observed.
One way to counteract the halo effect is for managers to rate all employees on the same dimension before moving on to the next one. This helps managers keep employee performance in perspective. Another option is to use more than one source to rate employees.
Although most researchers believe that the halo effect is present in almost all ratings and settings, some studies suggest that it is less prevalent—and less of a concern—than previously thought. In a review of past research, Murphy, Jako, and Anhalt (1993) concluded that, in the studies they examined, halo was not nearly as common as traditionally believed and, even when it did occur, did not negatively affect rating accuracy. Surprisingly, when organizations consciously try to control halo, they end up with less accurate ratings (Murphy & Reynolds, 1988). Organizations must therefore not become overzealous in their attempts to eliminate halo. Indeed, some employees really are very strong (or very weak) across all performance dimensions and their consistent ratings reflect an accurate evaluation of their performance.
Rater Biases
Ratings can also be influenced by a worker's personal relationship with the evaluator. Similar-to-me-error, for example, occurs when evaluators give higher ratings to workers whom they perceive to be like them (Wexley, Alexander, Greenawalt, & Couch, 1980). A study of 104 Air Force officers showed that familiarity between the officers and the aviators they debriefed, if it existed, positively affected the aviators' ratings of the officers (Scotter, Moustafa, Burnett, & Michael, 2007). Other research has examined personal characteristics such as personal attractiveness and demographic characteristics, each of which influences performance ratings.
Photo of two young businessmen in the background and one older businessman in the foreground.
Performance ratings are often based on the evaluator's personal biases relating to such things as gender, age, and race. For instance older workers typically receive lower performance ratings than their younger coworkers.
Three demographic characteristics require particular attention: race, gender, and age. As you recall from chapter 2, any tool organizations use to make employee decisions (hiring, promotion, placement, compensation, termination, etc.) must not discriminate against protected classes. In most organizations, performance appraisals provide important data used in those decisions. You can easily understand, then, that any personal biases based on race, gender, and age are especially problematic for an organization if they significantly influence performance appraisal ratings.
Research on race, gender, and age biases has produced mixed results. In the category of race, research shows that overall, black employees receive slightly lower ratings than white employees (McKay & McDaniel, 2006). However, because closer examination shows that all managers tend to give higher ratings for individuals of their own race, the overall results seem to be due to a combination of similar-to-me-error and the higher proportion of whites in managerial roles.
Likewise, gender bias occurs in some situations, but it tends to be limited to situations of gender-role incongruence. Specifically, male employees' performance tends to be rated higher than female employees' performance in traditionally masculine roles but lower in traditionally feminine roles, and vice versa for women (Pulakos, White, Oppler, & Borman, 1989). A more recent study of 448 upper-level managers examined the equity of promotions for men and women and found that women needed to show significantly higher performance appraisal ratings than men in order to be considered for promotion (Lynes & Heilman, 2006).
Finally, research shows that managers may favor younger workers. In general, older workers receive lower ratings than their younger coworkers, and the bias increases when older employees work for younger managers (Shore, Cleveland, & Goldberg, 2003).
Improving Rater Accuracy
Clearly, performance appraisals are susceptible to numerous types of rating errors, which affect the appraisals' usefulness in employee decision-making and self-improvement. To thwart the negative effects of error, I/O psychologists have identified several ways to increase rating accuracy. Let's take a moment to review some of these techniques.
Rating-Error Training
One way to deal with rating errors is to train raters not to make them. Rating-error training involves teaching raters about the different types of rating errors along with practical strategies for avoiding them. When dealing with leniency error, for example, raters will first learn about the type of error (in this case, inflated ratings with little variability), then its possible causes (e.g., raters wish to maintain a positive relationship with the employee), and finally strategies to overcome it (e.g., rank-order employees before rating each one). This process assumes that educating raters about potential errors and possible solutions can help them overcome their biases.
Unfortunately, the perceptual biases that lead to rating errors (stereotypes, the need to please, high expectations) are ingrained in our cognitive processes and are difficult to change. Rating-error training thus requires significant time and effort, with sessions typically lasting from six to eight hours (Latham, Wexley, & Pursell, 1975). Yet even with extensive training, improved rating accuracy is short-lived (Fay & Latham, 1982).
Rating-Accuracy Training
Like rating error training, frame-of-reference (FOR) training aims to help raters improve their rating accuracy. With this method, raters learn about the type of employee performance their organization wishes to see. They are given a frame of reference, or model, against which to compare their employees. If raters know the organization's performance standard for a specific job position, they will be less likely to use their own idiosyncratic standards to judge a worker in that position.
Designing FOR training involves a number of important steps. First, the raters receive a clear and concrete definition of various performance dimensions and rating scale included on the appraisal they will be using. Next, the trainer discusses work behavior that illustrates appropriate performance at each rating-scale level. After the raters have a basic understanding of the different performance expectations, they practice using the rating scale, usually by watching video scenarios of people working at various levels of performance and then attempting to rate them appropriately. Finally, the trainer goes over each scenario, discussing at which performance level each sample employee should have been rated.
Photo of business people seated at a table during a meeting.
Calibration meetings increase accountability because managers must discuss the ratings of employees in similar jobs, thus requiring managers to justify their rating of each employee to their peers.
Research has found that FOR training does indeed lead to more accurate performance appraisal ratings (Woehr & Huffcutt, 1994). As an added bonus, research by Davis and Mount (1984) found that managers who received FOR training became more effective at creating development plans for their direct reports, because they more clearly understood what the company expected of its employees.
Rater Accountability
Raters need to be motivated to make accurate ratings, but organizations often do not provide much incentive to do so. If your boss never underwent formal performance appraisal training, or if he or she was never really held accountable for the quality of his or her performance ratings, how motivated would he or she be to take performance appraisal seriously? In turn, how likely will you be to take it seriously when it is time for you to evaluate your own staff? Unfortunately, the poor behavior modeled by some executives often provokes similarly cavalier attitudes among the company's entire management team. If, however, managers are held accountable, they will take the appraisal process seriously and will be more accurate in their ratings (Harris, Ipsas, & Schmidt, 2008; Mero & Motowidlo, 1995). One way to increase accountability is for organizations to hold calibration meetings, at which managers discuss the performance and ratings of workers in similar jobs. As you might surmise, because they force managers to justify their ratings to their peers, these meetings increase rating consistency across employees (McIntyre, Smith, & Hassett, 1984).
4.5 Performance Management Systems
So far in this chapter, we have focused on the need for and strategies used to improve the accuracy of performance appraisals. However, even the most accurate appraisal can fuel negative worker reactions if it does not include quality feedback. This knowledge has prompted some I/O psychologists to suggest that worker self-improvement rather than rating accuracy should be the focus of a performance appraisal (Ilgen, 1993). In order to build a successful self-improvement plan, an employee needs to have the following information: first, clear performance expectations; second, an understanding of current performance; and third, suggestions from his or her manager on how to improve. The appraisal, then, is really part of a larger performance management system, which includes not only the appraisal but also the setting of performance expectations and continuous feedback. Of course, if he or she truly wishes to improve, the employee must be willing to listen to and address constructive performance feedback, something that will not happen unless the manager and employee are able to establish a significant amount of trust.
In this next section, we will review the three major components of the performance management system: building trust, continuous feedback, and expectation setting.
Building Trust
Think about the last time you worked with a peer or for a manager you did not trust. How did you feel? Chances are, at one time or another, you experienced frustration, confusion, apathy, or even anger toward that person. Now imagine that the person took some time to offer you constructive performance feedback. How would you respond? Most of us would question this untrustworthy person's intentions, withdraw from the situation, or even become hostile, ignoring the feedback or interpreting it in a negative way. Further, most people will stay quiet and not question the feedback because they fear repercussions for disagreement. Actual negative feedback is even more detrimental, especially when little trust exists between the manager and employee. Employees can become unmotivated and lose self-esteem (Kluger & DeNisi, 1996). When trust exists, however, workers are more likely to accept criticism openly, believing their manager has their best interests in mind.
As you can see, a positive manager-employee relationship is necessary for a positive performance appraisal experience, and trust is a prerequisite for an effective performance management system (Peterson & Hicks, 1996). Therefore, it is important for organizations to train managers how to build trust with their staff by doing such things as keeping their commitments, displaying integrity, providing timely feedback, and showing an interest in their subordinates.
Continuous Feedback
Photo of a man and a woman having a meeting.
When managers provide continuous feedback, employees feel less threatened and can immediately adjust their performance to solve existing problems and account for current demands.
During a typical annual post-appraisal meeting, a manager will hold a one-sided conversation with an employee, reviewing the employee's successes and failures from the most recent performance review cycle. As you may have determined from the opening exercises of this chapter, the meeting can quickly become tense, even hostile, if the feedback includes criticism. Employees tend to instinctively deflect criticism, blaming it on forces outside their control. Just think for a moment, though: If this meeting were a once-a-year occurrence, how desirable or useful would criticism be if it referred to something that happened ten months ago? Could the employee even do anything now to fix the situation? Probably not. How might an employee react differently if, instead of annually, she or he received informal feedback on a more regular basis?
Both scientific researchers and in-the-field practitioners advocate the habit of providing continuous informal feedback to employees (Corporate Leadership Council, 2002; Gregory, Levy, & Jeffers, 2008). More specifically, continuous feedback should be given in addition to, not to the exclusion of, the annual performance appraisal, and it must be provided immediately after any instance of effective or ineffective employee performance. Once-ayear feedback has never been very effective, but is especially unhelpful for today's users of instant-communication systems such as Facebook and Twitter. Increasing its frequency makes feedback not only less threatening but also more helpful, because employees can actually use the information to solve existing problems or adjust performance to meet current demands.
In addition to increasing frequency, I/O psychologists have identified a number of other conditions under which employees will be more likely to adopt managers' feedback suggestions:
1.Feedback must address specific employee behaviors or actions, not personal characteristics. It is best to use facts, data, statistics, or direct observations to support positive or negative feedback.
2.Feedback should involve two-way communication between managers and employees. If employees feel they can express their views, they are more satisfied with the feedback process (Diboye & Pontbriand, 1981).
3.Managers should provide constructive feedback only on behaviors the employee can control.
4.If feedback is negative, it should be constructive. Further, managers should provide support for the employees' self-improvement.
Setting Expectations
Most of us truly wish to do our best at our jobs, but it is hard to do so if we do not understand what is expected of us. Managers can facilitate effective goal-setting sessions by following these guidelines:
1.Managers should not simply assign performance goals. Rather, they should collaborate with employees to create mutually agreed-upon expectations. Participation in the goal-setting process increases an employee's goal-aspiration motivation (his or her desire to meet the goal) and leads to the creation of more challenging goals (Latham, Mitchell, & Dossett, 1978).
2.Goals that are specific and challenging yet achievable more successfully motivate workers to perform their best (Locke & Latham, 1990).
3.Workers perceive open-ended goals as less urgent than those with deadlines and are therefore less likely to achieve them. All goals should have specific achievement deadlines.
4.Employees tend to follow the adage, What gets rewarded, gets done. Employees should always understand how achieving (or not achieving) goals will impact them and how accomplishments will be rewarded.
5.Not all goals hold the same importance. Managers should help employees prioritize and order goals according to their importance.
4.6 The R in ROI: Linking Performance Evaluations to Financial Results—Friend or Foe?
One-panel comic showing one man sitting at a desk with all of his employees surrounding him. The caption reads, "Pay for how we perform? What kind of nutty idea is that?"
Performance measurement is extremely important, for the numerous reasons discussed earlier in this chapter. However, many times performance appraisals become a laborious exercise that I/O psychologists or the human resources department push on the rest of the organization to serve other, less important goals. For example, managing the performance appraisal system can become a goal in itself for those whose job is to see to it that the system is functional and well-maintained. It justifies the jobs they are holding and the salaries they are paid. When that is the case, the managers performing the evaluations start to view the performance-appraisal system as a formality and do not take it very seriously. They may fill in the necessary forms, but the accuracy of their ratings will likely be questionable.
Another common but often less effective approach to performance appraisals is limiting their outcomes to the determination of annual salary increases. Especially in a tight economy when overall payroll allocations are frozen, relatively stable, or even decreased, linking performance appraisals only to annual raises is unlikely to be conducive to performance outcomes. For example, when the best performers get a raise of only 4% and the worst performers get a raise of 2%, you can see why the difference is unlikely to translate into any constructive performance feedback for either employee. In cases where there is a limited pool of resources to distribute, high performers often feel guilty about getting a raise at the expense of their colleagues, rather than feeling appreciated or rewarded. Conducting performance appraisals only to determine annual raises also ignores their purpose of providing continuous constructive performance feedback. As discussed earlier, employees need to receive feedback on their performance more often. This feedback should be provided to help them better align their performance with the organization's goals. When appraisals are always linked to pay, appraisal sessions become mostly about money rather than about performance improvement.
Performance appraisals are also often used in conjunction with layoff decisions. Even though this is a legitimate use, if they are only used to justify these decisions, they become perceived as a way for the organization to provide a legally defensible paper trail, which diminishes their value, and they become resented and distrusted by managers and employees alike. Again, although the above uses of performance appraisals are important and legitimate, the primary use of performance appraisals should be as a tool to objectively measure performance and facilitate its improvement.
So how do organizations, managers, I/O psychologists, and human resource departments design truly effective performance appraisals systems? It is critical to choose the correct measures. As discussed earlier, the correct measures should be readily linked to the goals of the organization and its success. This chapter offers numerous ways to enhance the quality of performance measures. However, in the words of sociologist William Bruce Cameron (1963) in his book Informal Sociology: A Casual Introduction to Sociological Thinking: "not everything that can be counted counts, and not everything that counts can be counted" (p. 13). The Pareto efficiency principle, also known as the 80-20 rule, posits that in most situations, 80% of outcomes are caused by 20% of the inputs. In performance appraisal, this means that it would be most effective for managers to focus on their employees' most critical behaviors, which constitute about 20% of everything that their employees do on a daily basis, because these critical behaviors cause 80% of the outcomes that truly have an impact on the organization's success and effectiveness.
Although not the only approach, performance measures that are directly linked to financial results tend to be perceived as objective and fair because they reflect the true value of an employee to an organization. They are also critical for resource allocation decisions, because they put human resource decisions on a par with other investments. For example, the financial value of an employee's performance can justify the costs of hiring or retaining that employee over outsourcing the position or investing in a piece of equipment that would allow for the automation of that employee's job.
Should an appraisal system then attempt to capture the financial value of every aspect of an employee's performance? Absolutely not! The concept of opportunity cost, introduced in earlier chapters, implies that only performance dimensions where the benefits of measurement exceed the costs should be captured. For example, even though it is easy to quantify stationery consumption, the cost of policing employees into being less wasteful in their use of inexpensive stationery items can be higher than the cost savings that may accrue from these initiatives. Similarly, many organizations install expensive equipment or have their managers waste numerous hours managing their employees' attendance when the costs of the few minutes that an employee may come in late or leave early far exceed the costs of monitoring.
What, then, should an appraisal system track? The Pareto efficiency principle implies that it should focus on performance dimensions that would be of high enough financial value to justify the cost. The key is not always the absolute cost or benefit of a performance dimension but rather the variation of that cost or benefit. For example, the difference between the most and least conservative use of stationery is not substantial. In measuring the performance dimensions that would have the most substantial effects on financial outcomes, emphasis should be on the dimensions with the highest variability and the ones that are "pivotal" to performance (Cascio & Boudreau, 2008). For example, at an upscale restaurant, the most pivotal performance dimension for cooks is their cooking skills. On the other hand, the most pivotal skills of the wait staff are their social skills. Although it is necessary for cooks to have some social skills in order to effectively deal with the wait staff and restaurant management, social skills are not pivotal for cooks. They can still be subjectively evaluated (e.g., using a narrative or a rating scale), but attempting to place a financial value on them is both impractical and unnecessary.
Consider This: Appraising Performance Dimensions That Really Matter
•Choose a job that would be of interest to you. It can be your current job, a job you held in the past, a job you hope to have in the future, or just a job that you have come across in a jobopening announcement or advertisement.
•Describe the job in detail. For more information, you may search for job descriptions of similar jobs.
•How would you measure the performance of the incumbent of this job? What are the most important dimensions to measure according to the Pareto efficiency principle?
•What are the performance dimensions that will likely exhibit the most variability (the most pivotal dimensions)?
•Which dimensions will be most readily linked to financial results? Explain.
•Which dimensions should be subjectively evaluated?
•Which dimensions should be ignored and not evaluated? Note that this is an important decision because it can significantly affect the efficiency and effectiveness of a performance appraisal system. It is also a decision that is often neglected or inadequately addressed.
4.7 Crossing the Positivity Threshold: Toward More Positive Appraisal Sessions
As humans, we have a tendency to overemphasize and amplify the magnitude of negativity in our lives (Baumeister, Bratslavsky, Finkenauer, & Vohs 2001). Negative stimuli tend to receive more of our attention and energy. For example, threatening personal relationships have been shown to receive more of our thought time than supportive ones, and blocked goals tend to receive more thought time than those with open and available options (Klinger, Barta, & Maxeiner, 1980). Performance appraisal is no exception. It is much easier to dwell on our own or others' faults than to acknowledge talents, strengths, and positive performance attributes. Doing the latter requires intention.
One-panel comic showing one man speaking to his employee. The manager says to his emplyoee, "We're giving you a 6% increase in appriciation. So... nice haircut, Andy."
So why do humans in general tend to focus on negativity? The tendency to overemphasize negativity has been attributed to primitive survival mechanisms in reaction to perceived physical danger. In civilized societies, overemphasis on negativity has been attributed to four psychological factors that are comparable to these survival mechanisms: intensity, urgency, novelty, and singularity (Cameron, 2008). The first factor is the intensity of negative stimuli. Because negative events are perceived as threatening, they are experienced more intensely. Second is the sense of urgency that negative stimuli place on our perceptions and action tendencies, because something is wrong and needs to be fixed. Positive stimuli do not pose the same sense of urgency, because ignoring positive stimuli does not pose as much risk as ignoring negative stimuli. Third is the perceived novelty of negative events. Believe it or not, a lot of what is going on in most people's lives is positive. That's why it tends to go unnoticed. Negativity is the exception. That's why it gets more attention.
Fourth, one of the unique characteristics of negativity is what is referred to as singularity. Imagine a system with one defective component, a body with one ailing organ, a team with one counterproductive employee, or a family with one dysfunctional member. A single negative component is capable of tainting the performance of the collective, which causes that single negative component to really stand out and alarm the rest for the need to somehow remedy the problem. On the other hand, positivity tends to be more general and global. One positive component alone does not necessarily make a system better. One good employee alone usually cannot make an organization successful. One healthy organ alone cannot make the whole body healthy. This singularity makes the effect of negativity more pronounced and far-reaching.
Photo of an executive and an architect at a construction site. The businessman is patting the architect on the back.
Supervisors should focus on the positives of each employee's performance rather than dwelling on the negatives. If the supervisor makes three positive comments for every negative one, the worker likely will be more receptive to the constructive criticism offered.
Paradoxically, humans also have a natural tendency, referred to as the heliotrophic tendency, to gravitate toward what is pleasurable (i.e., positive) and away from painful or uncomfortable stimuli. However, this tendency tends to be overwhelmed by the intensity, urgency, novelty, and singularity of negativity and needs to be brought out through intentional decisions and actions. That is why although most managers recognize their tendency to overemphasize their employees' weaknesses, faults, and mistakes and wish they could be more positive, they cannot. For example, they may get overwhelmed by the urgency of addressing the dysfunctional behaviors of their worst employees that they have no time to interact with and praise their better ones for their consistent positive behaviors. Moreover, those consistent positive behaviors may no longer stand out; they may be taken for granted and a manager may forget to recognize them when appraising these employees' performances.
So how can managers overcome their negative tendencies and lead more positive performance appraisal sessions, which can lead to positive relationships with their employees that can be conducive to higher subsequent performance and a better-functioning organizational culture? First, a manager needs to recognize the important concept of the positivity ratio, which was introduced in chapter 3. Although extreme, Pollyannaish positivity is unnecessary and can even be dysfunctional, research supports the existence of a tipping point or threshold for positivity at which humans go beyond just being average or functional and start to thrive and flourish (Keyes, 2002). This tipping point tends to take place at a positivity-tonegativity ratio of about 3:1.
So, managers need to intentionally create about three positive interactions with their employees for every negative interaction. In performance appraisal sessions, managers should really put in the effort to find and comment on three positive aspects of their employees' performance for every negative aspect they want to bring to an employee's attention. This requires the art of catching employees doing something right instead of the common practice of focusing on problems and mistakes. Interestingly, research shows ratios of 2:1 or 1:1 are not significantly different. They are almost equally counterproductive. Interactions that fall below the 3:1 threshold will likely be perceived by the employee to be excessively negative, regardless of how negative.
You might think that this hand-holding is more necessary for new or inexperienced employees and that more mature employees or more established relationships can tolerate lower positivity ratios. However, research shows the tipping points in those situations are actually higher. For example, the threshold is about 5:1 in more complex settings such as top management teams and as high as 6:1 in marital relationships (Fredrickson & Losada, 2005; Losada & Heaphy, 2004; Gottman, 1994).
Consider This: Positive Performance Appraisal Sessions
In order to conduct more positive performance appraisal sessions, managers need to be more positive in collecting and sharing performance information. Below are two examples of positively oriented practices that can be used to replace the negative practices often used by managers.
•Negative: Reprimand workers when late. Positive: Praise and reward workers who are consistently on time. Rationale: Workers who are on time will know that their positive behavior is noticed and appreciated instead of ignored. Late workers will start coming on time to get the manager's attention and receive rewards.
•Negative: Criticize an employee for weaknesses (e.g., poor people skills or leadership skills). Positive: Find and acknowledge the employee's strengths that parallel those weaknesses (e.g., independent thinking or willingness and ability to follow directions). Suggest changes in role to better fit employee strengths and/or training opportunities to develop lacking skills. Rationale: Weaknesses may be based on stable personality traits that cannot be readily changed (e.g., introversion). In these cases, role changes are more likely to lead to performance improvements. In others, training and development are more likely to be perceived as opportunities. Criticism is more likely to be perceived as a threat.
Find Out for Yourself: Do Your Relationships Cross the Threshold?
•Choose an individual you interact with on a regular basis in a close relationship. It can be a spouse or significant other, a friend, a coworker, a parent, a sibling or a classmate. (Do not share any information about positivity ratios or this exercise with that individual yet.)
•Keep track of your interactions with this person for a week. Jot down brief notes about your interactions and evaluate whether you consider each interaction to be positive or negative.
•At the end of the week, calculate the positivity ratio of your relationship by dividing the number of positive interactions by the number of negative interactions.
•Did your relationship cross the positivity threshold (generally 3:1 for work relationships and 6:1 for personal relationships)?
•Now ask that individual to honestly assess whether your relationship is positive, neutral, or negative. Compare the individual's perceptions with the positivity ratio you calculated. Do not be discouraged if the perception is too negative or if your relationship did not cross the threshold. Eighty percent of relationships and interactions do not.
•Here is the hardest part of this exercise! If your relationship did not meet the threshold, then over the next few weeks, intentionally try to increase the positivity ratio, keeping track of your interactions and calculating your ratio on a weekly basis until you reach the threshold. Describe (and enjoy) the noticeable increase in the relationship quality.
