Copyright Paul E.
Spector, All rights reserved, March 15, 2005
Determination and Documentation of Individual's
Performance
Should be tied directly to criteria
USES
◦ Administrative decisions (promotion, firing, transfer)
◦ Employee development and feedback (otherwise they will
not know)
◦ For Scientific Research Purposes (i.e. which method works
and which does not)
◦ Whether Training is required or not.
Criterion: a standard, a yardstick, a rule of thumb by
which you can judge or measure something. Ex- for acid
test ratio, criterion is 0.5-1 (good condition of firm)
Single Criterion: Global measure to represent
performance (Universal)
Multidimensional
◦ Each person gets multiple scores that aren’t combined (factor
based)
Theoretical criterion: General measure of
performance. Ex- for teacher: Impart knowledge to
students
Actual criterion: Detailed measure of performance.
Ex- student achievement test scores.
Look at table on Pg 82 of Book.
Relevance: Actual assesses the theoretical
Contamination: Actual measures something other than the theoretical
Deficiency: Actual fails to capture the theoretical
Different jobs have different criterias for
measurement or success
Some jobs even have multiple criterias, all of
which are important
Ex- Quality vs Quantity
Think about the job of a long jump athlete (quantity
focus), a gymnast (quality focus) and a customer
service staff (both quality and quantity focus).
what could be the actual criterion/ criteria for winning
the competition/ job?
should actual criteria focus quality or quantity aspect?
Focus on quantitative elements, Counts of behaviors or
outcomes of behaviors. Ex- absences, accidents,
output/hr, sales targets etc.
Advantages
Consistent standards within jobs
Not biased by judgment
Easily quantified
Disadvantages
Not always applicable (teacher)
Data does not show feelings or special situations
Employees may not like being treated as machines
Qualitative Judgments about performance
Trait based graphic rating scale
Behavior based: Critical incidents
Mixed Standard Scale
Behaviorally Anchored Rating Scale
Behavior Observation Scales
Problems:
Rating errors: Leniency, Severity, Halo
Supervisor subversion of system--leniency as a strategy
Mixed purposes (feedback vs. administrative)
Negative impact of criticism
Let's think for 3 minutes!
Halo errors (when one attribute unfairly
affects other attributes). Ex- Smart, good
looking candidate gets high scores.
Distributional errors
◦ leniency errors (too high marks)
◦ severity errors (too low marks)
◦ central tendency errors (always average marks)
Separate purposes
◦ Raises dealt with separately from feedback
Consistent feedback, everyday
Limit criticism to one item at a time
Praise should be contingent on performance
Supervisors should be coaches
Appraisal should be criterion related, not personal
Employee performance management systems
◦ Web-based
◦ Automated—reminds raters when to rate
◦ Reduces paperwork
◦ Provides feedback
360-degree feedback systems
◦ Ratings provided by different people
Peers
Subordinates
Supervisors
Self
◦ Big clerical task in large organizations to track/process ratings
◦ Web makes 360s easy and feasible
◦ Consulting firms available to conduct 360s
Technology can be helpful for performance appraisal
Ex- 360 degree performance appraisal takes ratings from
various ppl in the organization.
But in large organizations, this can be extremely difficult
to get ratings from all parties. (because there are so many
ppl)
Two ways technology can affect performance appraisal:
◦ Monitoring of objective productivity. Ex- fingerprint recognition
system in BRAC U for admin staff.
◦ Implementation of online performance management systems. Ex-
online faculty evaluation in East West University. Brac still
lagging ?
Two ways to reduce errors:
1. Error resistant rating forms
The key is to develop a quantitative scale for qualitative factors. Ex-
Loyalty Scale, Attitude Scale etc. But hard to develop these.
◦ Behaviorally Anchored Rating Scale, BARS
◦ Behavior Observation Scale, BOS
◦ Mixed Standard Scale, MSS
2. Rater training
◦ Rater error training: instructs raters in how to avoid errors
Reduces halo and leniency error
Less accuracy in some studies
◦ Frame of reference training: Give raters examples of sample
performance and sample ratings to better align their concept
