Performance Management

Objective vs. subjective performance appraisal

(Bommer et al, 1995): objective and subjective measures only correlate ~ .39, and should not be used interchangeably. Emphasis on performance improvement as the ultimate goal in the appraisal process and employee motivation to improve their performance and subordinate perceptions of appraisal fairness (Pulakos et al., 2019)

Dual-process system process of PA

(Feldman, 1981): views PA as a dual-process system of evaluation and decision making. Attention, categorization, recall and information integration are either automatic or controlled. Automatic process usually dominant unless decision is problematic. Categorization and recall are subject to many biases (halo, leniency/stringency, racial)

PA reactions

(Keeping & Levy, 2000) Important to study reactions to PA because a) it is criterion of interest to practitioners and b) are theoretically linked to PA acceptance and success. Reactions include: 1) satisfaction (w/interview, system, or ratings); 2) fairness; 3) perceived utility; and 4) perceived accuracy

Feedforward

(Kluger & Nir, 2010). Articulate what has gone well. Understand the strengths of the target and the context that led to the positive experience. Help target apply those strengths or situational resources to increase likelihood of success in future. Performance review should focus on strengths of employee

Scale types

(Landy & Farr, 1980): graphic rating scales have better cost ratio than BARS. Since courts don't care about complexity of scales, graphic rating scales may be used over BARS. Should choose best one based on utility analysis. Behavioral observation scales (BOS) ask raters to use aids such as diaries to standardize performance observation and recall (Latham & Wexley, 1977)

Ratings errors and use

(Landy & Farr, 1980): potential biases/error in ratings can occur in 1) Roles: rater & rate characteristics; 2) Context: type of org, purpose of rating; 3) Vehicle: type of rating instrument used; 4) Rating Process; and 5) Results of Rating. Ratings should only be used in high-stakes appraisals if the motivation of the raters can be aligned with the goals of accuracy, fairness, and transparency

Reorienting PM

(Murhpy, 2020) Similar to (Aguinis, 2013) with a few tweaks. Focus on behaviors essential to successful leadership and coaching. Cascading goals: translate broad goals into plans and strategies for individuals and teams. Reorient coaching to communicate clear performance objectives and link individual employee behavior with the performance goals of the unit. Focus on why employees are deviating from performance plan and identify ways to help and encourage employees in executing the plan. Forcing supervisors to act as judges prevents them from functioning effectively as leaders. Leaders are most effective when they can inspire, guide, and assist employees (Bass & Bass, 2008)

Assumptions of performance rating research & manager's PA goals

(Pulakos et al., 2019) Everyone has a stable level of true performance that reflects their effectiveness on the job. "true performance". Raters are able to rate others accurately. Raters are motivated to rate others accurately. Managers must balance four competing goals when evaluating employees (Murphy & Cleveland, 1995): 1.Task performance goals using ratings to influence future performance, 2.Interpersonal goals using ratings to maintain relationships with employees, 3.Strategic goals using ratings to increase manager's standing in the organization, 4.Internalized goals which reflect their personal beliefs about evaluating performance

Affect and performance ratings

(Varma et al, 1996): affect was sig related to all performance ratings, though more strongly related to trait-like ratings than task/outcome-like ratings. Affect may be a result of better subordinate performance (i.e., may not be bias)

Performance appraisal legal

1. (Barrett & Kernan, 1987): reviews perf appraisal court cases: 41/51 of cases ruled on the side of the organization. In other cases either 1) clearly racist supervisors or 2) uneven application of performance evaluation standards. 6 recommendations: 1) conduct JA; 2) incorporate JA findings into rating instrument; 3) train supervisors to use instrument appropriately; 4) allow formal appeal rules and review of ratings; 5) document PA evals; 6) provide corrective counseling for poor performers. 2. (Martin et al, 2000): evaluations of performance should be based on results of a PA that incorporates concerns for org justice and fairness to avoid legal concerns. Defense for promotion discrimination: 1) selected person with better credentials; 2) emp not qualified for promotion; 3) shortcomings in present job. Discharge defense: 1) emp fails to meet org expectations for promotion; 2) performance deterioration

Major Performance Management Citations

1. Definitions: appraisal, performance management 2. Legal: Barrett & Kernan (1987) 4. Feedback: Kluger & DeNisi (1996); Aguinis (2013) 5. Goals: Locke & Latham (1990; 2007) 6. PE method: Murphy (2020); Denisi & Murphy (2017); Bommer et al. (1995) 7. Distribution: Murphy (2020); Aguinis et al. (2018) 8. Fairness/reactions/errors: Greenberg (1986); Landy & Farr (1980) 9. 360: Atwater et al. (2007) 10. Assumptions of performance ratings: Pulakos et al. (2019) 11. Purposes: Campbell & Wiernik (2015) 12. Managers: Murphy & Cleveland (1995) 14. Dropping PE: Murphy (2020); DeNisi & Murphy (2017) 15. Org change: Bridges & Bridges (2016) 16. Feedforward: Kluger & Nir (2010) 17. Future: Pulakos et al. (2019); Murphy (2020)

360 ratings

Beehr et al. (2001). 360s are distinct from performance appraisal (generally considered broader) and are typically used for developmental purposes and for assessing OCB. They are not used effectively for administrative purposes; not related to selection. Peer and supervisors provide different but correlated ratings, while self-ratings are not typically correlated with anything else (much more lenient). In contrast, performance appraisals are typically good for assessing task performance.

Purposes for PA

Campbell & Wiernik (2015) 1)research, 2)legal defense for personnel decisions, 3)high-stakes appraisal (e.g. promotion), 4)performance feedback and development, 5)self-managed performance improvement. Work roles are design to accomplish org goals and influence bottom line (Ployhart & Hale, 2014), but individuals should not be held accountable for outcomes they have no control over

Typical vs. maximum performance

Campbell & Wiernik (2015). Are levels of effort typical of the individual or do they reflect specific conditions that keep attention highly focused, increase effort, or maintain higher effort levels for some period of time? Correlation between typical and maximum performance is .42 (Beus & Whitman, 2012)

Performance dynamics

Campbell & Wiernik (2015). Performance changes due to training, goal setting, aging, etc. (Stewart & Nandkeolyar, 2007) and use repeated measurements to capture the dynamic nature (Stokes et al., 2010). Distribution of performance may be normal (Beck et al., 2014) or not (Aguinis & O'Boyle, 2014)

Improving rating construct validity

Construct validity can be improved by concretely specifying performance and incorporating into rater training and rating instruments and raters: Have observed the ratee extensively, Accept the rating goal and performance specifications, Understand the rating instrumentation, Know they are accountable for rating accuracy, Have ample time, Are sensitive to contaminants, such as liking for the ratee

Cross-cultural performance assessment.

Cross-cultural PMS need to accommodate local practices and preferences. Substantial variability in structure and formalization of PMS exist across countries (Peretz & Fried, 2012)

Appraisal and performance management definitions

DeNisi & Murphy (2017) Appraisal: formal process which occurs infrequently by which employees are evaluated by some judge who assesses the employee's performance along a given set of dimensions. Performance management: wide variety of activities, policies, procedures, and interventions designed to help employees to improve their performance, begins with performance appraisal, include feedback, goal setting, and training

Performance appraisal 1980's

DeNisi & Murphy (2017). cognitive processes - how raters acquire, organize, recall, and integrate info. Landy & Farr (1980) both of these studies said to move away from looking at scales and errors and to focus more on rater's cognitive processes in evaluating perf. Murphy et al. (1982) accuracy in evaluating perf not related to accuracy in observing and encoding beh. Conclusion: this work is not as relevant as we thought. Ultimate goal of performance management is to improve firm-level performance. Switch from accuracy of ratings to what drives employees to improve their performance. Focus on how combinations of HR practices make a diff and why some work and some do not in improving firm performance

Performance appraisal 1970-1979

DeNisi & Murphy (2017). focus on scale format and accuracy. Landy & Farr (1980) review of performance appraisal research said that research on scale format was unproductive and variations in scale formats had only modest effects on qual of rating scales. We concluded that scale format is not as important as we thought

Performance appraisal history 1920's - 30's

DeNisi & Murphy (2017). improving scales. Thorndike (1920) - halo error introduction . Assumption that these errors reduce accuracy of ratings and make them less useful. Followed by Rugg (1921) and Remmers (1931) who argued that graphic rating scales were prone to halo. Paterson (1922) graphic rating scales. Cattell (1906) ranking methods. Flanagan (1954) critical incidents technique. Latham & Wexley (1977) behavioral observation scales. Travers (1951) forced choice rating scale. Smith & Kendall (1963) behavioral anchored rating scales BARS - clear definition and consistent FOR

Performance appraisal1980's

DeNisi & Murphy (2017). studies of criteria assessed for evaluating ratings - rater errors, interrater agreement, rating accuracy. Focus on halo error and other rating errors. Assessment of reliability, validity, and accuracy of performance ratings. Make sure you capture multidimensionality in ratings and whether you are measuring typical or maximum perf. 3 rater error measures - halo, leniency/severity, and central tendency/range restriction. Gilford & Jordensen (1938) leniency and central tendency. Error measures became the most common method for evaluating the quality of rating data. Borman (1979) eval rating accuracy by comparing between ratings obtained from a single rater with the average of ratings obtained from multiple experts operating in optimal rating conditions. Rater error and rating accuracy measures can be problematic

Performance appraisal history

DeNisi & Murphy (2017)Scale format and focus on scale accuracy evaluating quality of rating data (criteria to eval ratings), focus on accuracy and rater error, rater training relationship between demographic char of raters and ratees rater training cognitive processes improving performance

New models of perf appraisal

DeNisi & Pritchard (2006) 1.improving individual-level performance and emphasizes key role of employee motivation to improve. Pritchard et al. (2002, 1989) - org productivity using a system of ProMES (productivity measurement and enhancement system). Levy & Williams and Aguinis (2009) - org performance and tying perf mgmt. processes with broader org strategic goals. Murphy & DeNisi (2008) - distal and immediate factors that influence outcomes of appraisal process. 2.Employee reactions. Roberts (2002) employees participate in dev of rating scale they have stronger motivation to improve their perf and this was highly corr with perf improvement (r = .57).

Performance appraisal history

Early history is focused on performance evaluation. Efficiency ratings in US federal civil service in late 1800s and officer performance during WWI (Pulakos et al., 2019). Thorndike (1920) published an article about constant error (now know as halo error). No rating format yielded substantially more accurate or less biased ratings than any others (Landy & Farr, 1980). Forced choice in 2000s: managers choose which behavior is most true of each employee's job performance from a set of equally desirable behaviors and use IRT to place each employee. Rater training to improve ratings (Borman, 1975; Latham et al., 1975). 1980s: more holistic theory needed to understand interactive effects of different factors on ratings using human info processing theories (Landy & Farr, 1980). 1970s and 1980s: legal challenges led to more structured evaluation processes (Pulakos et al., 2019). Forced distribution (top 10%) was popular until ~2010. 360 reports (Borman, 1974)

Limited utility of performance evaluations to orgs

Four purposes for evaluating job performance (Cleveland et al., 1989): Make distinction between individuals (e.g. identify best candidates for salary increases or promotions). Make distinctions within individuals (e.g. identify weaknesses for determining training and development needs). Support HR systems (e.g. validating personnel tests, evaluating training). Documentation (e.g. support decisions for promotion or dismissal). Most common purposes are development and career moves, but put conflicting demands on PMS (Meyer et al., 1965). High ratings on some performance dimensions and low ratings on others are best for identifying development, but useless for separating average ratings of people.

Perceived fairness of evaluation

Greenberg (1986) determinants of perceived fairness of evaluation. Fairness perceptions of appraisal system help determine if employees will accept it. 7 categories of fairness determinants w/ 2 underlying factors: Procedural = a) soliciting input prior to eval and using it; b) two-way communication during interview; c) ability to challenge/rebut eval; d) rater familiarity with ratee's work; e) consistent application of standards. Distributive = a) receipt of rating based on performance achieved; b) recommendations for salary/promotion based on rating.

Feedback intervention theory

Kluger & DeNisi (1996). people are demotivated when they receive feedback that may threaten their self-concept and will have no motivation to continue performing the task if they believe they are too far away from achieving the performance standard or if they repeatedly receive neg feedback. Balance pos and neg feedback and deliver neg feedback in a way that they will accept it. Feedback may not be as effective as we thought it was, effect size was .41 but 1/3 in wrong direction, 1/3 in right direction, and 1/3 no effect. Feedback interventions can improve employee motivation and self-efficacy. Employees have to be motivated to improve perf and motivation is enhanced when feedback is frequent

Key performance management behaviors for managers and employees

MANAGERS (Bryant, 2011): Set clear expectations, success criteria, and standards. Revise expectations in real time, so employees know what to do. Provide informal feedback daily to praise, coach, and course-correct employee performance. Check in regularly with employees to stay in touch and provide guidance. Coach employees and help them solve problems to enable success. EMPLOYEES: Clarify performance expectations to ensure understanding and revise when necessary. Set expectations with peers about who is doing what and by when. Ask for and accept feedback openly and nondefensively. Use feedback to course-correct and continuously improve own performance. PM should be a regular cadence through manager and employee behaviors (Pulakos, 2015).

Performance management overview

More complex PM systems become standard over the last 15-20 years, including cascading goals, expectation setting, and interim feedback review (London & Mone, 2014). Cascading goals: link org's strategic goals down to each employee (Rodgers & Hunter, 1991). Problems w/cascading goals (Pulakos & O'Leary, 2010): take time, cascading goals become disconnected from org goals (e.g. game of telephone), goal setting at beg of year does not account for unexpected events. Individual goal setting: employees perform more effectively when specific goals are set (Locke & Latham, 1990). Shorter time spans are better for driving performance (Latham & Locke, 2007). SMART goals: Specific, Measurable, Aligned, Realistic, and Time-bound. Employees evaluated on behavior and results (Pulakos, 2009). How employees perform and what they deliver are important parts of performance. ROI analyses show millions of $ lost due to costs, time, and no impact on performance (CEB, 2012)

Is it possible to manage performance without evaluation?

Murphy (2020). Current PMS are unreliable and orgs already operating without useful PMS. Switch to providing coaching to those who would benefit most (e.g.. newcomers) (Li et al., 2011). Separate the two purposes (development and career moves). The most serious barrier to success of PMS is that performance evaluation is used to drive career decisions

Four barriers to successful performance evaluation

Murphy (2020). Distribution of performance. Failure to devise reliable and valid methods for obtaining judgments about performance. Limited utility of performance feedback to employees. Limited utility of performance feedback to orgs

Should we evaluate performance?

Murphy (2020). Do people meaningfully differ in their effectiveness in performing their jobs? Is it feasible to evaluate performance (i.e. valid and reliable)?

Methods for performance evaluation

Murphy (2020). Failure to devise reliable and valid methods for obtaining judgments about performance. Objective measures can capture important aspects of job performance but do not capture the full range of behaviors (Landy & Farr, 1983). Subjective ratings of job performance are not reliable (Ones et al., 2008). Denisi & Murphy (2017) suggest that the most common performance measure types (ratings of performance by supervisors, peers, or others) are at best weak indicators of individual job performance.

Limited utility of performance feedback to employees

Murphy (2020). Feedback can be useful when people have little job experience and the feedback has info they might not normally have access to (Li et al., 2011). Results of studies on feedback effects are split into thirds: positive, negative, or no effect according to review of studies (Kluger & DeNisi, 1996). Effects of 360 feedback on behavior and performance is mixed (Atwater et al., 2007). Little evidence that PM feedback improves performance (DeNisi & Smith, 2014). People view the ratings they receive from others as unfairly low (Murphy et al., 2018). Disappointing experiences with PM lead to higher levels of cynicism and disengagement and spirals

Future performance evaluation research

Murphy (2020). How can we better teach raters about performance, goals, dynamics, contamination? How and for what reasons, do raters actually make rating judgments? What information do they use? How can online performance ratings be structured and delivered to avoid careless ratings and instill the same feelings as in-person ratings? What are the best ways to promote transparency and procedural justice for assessors and assessees? How can simulations be used to assess performance on other dimensions of performance? How can they assess adaptive responses to changing requirements? What are the best ways to display and present behavioral performance data to facilitate effective evaluation and decision making?

Distribution of performance

Murphy (2020). Long assumed that performance is normally distributed (Hunter & Hunter, 1984). False. Most indices of performance follow a power law distribution (Aguinis et al., 2018), suggesting that very few people are highly effective at their jobs, the majority of distribution is people who are markedly less effective. Most people are acceptable performers, not star performers. A handful of top performers contribute disproportionately to overall output of org (Aguinis & Bradley, 2015).

Why do PMS refuse to die?

Murphy (2020). Many stakeholders believe performance is beneficial to measure and use to drive decisions. Widely believed that performance feedback is valuable and improves employee motivation and performance. No clear alternative. One assumption: surface features of system can be improved (e.g. rating scales, schedule for feedback, and manager training for implementation)

Should Orgs drop performance evaluation?

Murphy (2020). Orgs should consider dropping regular evaluation of employee performance and focus on the small subset of situations where performance evaluation and feedback are useful. Some performance appraisal systems are built to motivate future performance, some identify poor performers (Welch & Byrne, 2001), and others are built to align performance goals with strategic goals of org with real time feedback (Aguinis, 2013)

Performance evaluation

Murphy (2020). process in which one or more individuals in orgs observe and obtain info about the job performance and effectiveness of individual employees. Some large orgs abandoned them (Cunningham, 2015). Little to no evidence they have any real impact on performance or effectiveness (DeNisi & Murphy, 2017). Survey of 4800 orgs in 24 countries: 82% have a formal PA system (Dewenttick & Remue, 2011)

Successful implementation of PM processes

PM transformation is a major org change effort (Bridges & Bridges, 2016): develop business case, engage stakeholders, gain buy-in, communicate effectively, enable and empower workforce to embed the change, and manage expectations.

Future PM research

Pulakos et al. (2019). Evaluate different approaches to large scale PM behavior change. Does streamline and driving effective PM yield sustained improvements? Make PM more team-based and think about the collection of skills in each team, not individual. What context factors matter most for effective PM processes? Different PM processes may be required for the different contexts within an org. (e.g. work requirements, goals, levels) (Church et al., 2015)

PM Takeaways

Pulakos et al. (2019). Formal PM processes disengage employees, cost millions, and have no impact on performance. Formal systems can be streamlined but should not be eliminated without robust informal processes. Informal day-to-day PM behaviors enable performance but take time and effort to embed

Future research in performance management

Pulakos et al. (2019). Simpler judgments that align with overall judgments raters naturally make are likely sufficient and most practical for the majority of evaluation needs. How to leverage and combine available performance information data into meaningful, valid, and fair performance assessments? What role will humans play in future performance evaluation processes?

New approaches to performance management

Pulakos et al.(2019). Two avenues for value change: 1.Streamline the formal PM system: Clarify purpose of PM, since PM try to serve too many purposes (Pulakos, 2009), Keep ratings? Many big companies have abandoned ratings. Provide some performance info to employees (Adler et al., 2016). 2. Drive more effective PM behavior. Feedback is important for performance and job attitudes (Ilgen et al., 1981). Informal feedback in real time is the most impactful for performance and engagement (Aguinis, 2013)

Samples, simulations, and proxies

Work sample: individual performs an actual job task using real job materials. Simulations: perform a task in a fabricated situation or with facsimiles of task materials. Proxies: Elicit the same performance responses, but do not closely mimic actual job tasks or situations. Pros: Can be more valid assessments of employees' abilities, less contamination. Cons: Expensive, time consuming, construct deficiency, lack construct validity, contamination from not accurately reflecting real performance (Lievens & Patterson, 2011)