Page 230 - Fundamentals of Management Myths Debunked (2017)_Flat

Exhibit 7–5 Selection Decision Outcomes CHAPTER 7 • Managing Human Resources 229

Later Job Performance Successful Unsuccessful decision decision
Reject
Correct
error

Correct
Accept
error

Reject Accept
Selection Decision

A decision is correct when (1) the applicant who was predicted to be successful (was
accepted) later proved to be successful on the job, or (2) the applicant who was predicted reliability
The degree to which a selection device measures
to be unsuccessful (was rejected) would not have been able to do the job if hired. In the the same thing consistently
former case, we have successfully accepted; in the latter case, we have successfully rejected. validity
Problems occur, however, when we reject applicants who, if hired, would have performed suc- The proven relationship between a selection device
cessfully on the job (called reject errors) or accept those who subsequently perform poorly and some relevant criterion
(accept errors). These problems are, unfortunately, far from insignificant. A generation ago,
reject errors only meant increased selection costs because more applicants would have to be
screened. Today, selection techniques that result in reject errors can open the organization to
charges of employment discrimination, especially if applicants from protected groups are dis-
proportionately rejected. Accept errors, on the other hand, have obvious costs to the organiza-
tion, including the cost of training the employee, the costs generated or profits forgone because
of the employee’s incompetence, and the cost of severance and the subsequent costs of addi-
tional recruiting and selection screening. The major intent of any selection activity is to reduce Employees who apply for management
positions at this Sarku Japan fast-food
the probability of making reject errors or accept errors while increasing the probability of restaurant must take written tests as part of
making correct decisions. How? By using selection procedures that are both reliable and valid. the company’s selection process. Properly
designed written tests for intelligence,
integrity, personality, and interests are
What iS reLiabiLity? Reliability addresses whether a selection device measures the popular selection devices that often are valid
same characteristic consistently. For example, if a test is reliable, any individual’s score and reliable in helping to predict which
applicants will be successful on the job.
should remain fairly stable over
time, assuming that the characteris-
tics it’s measuring are also stable.
The importance of reliability should
be self-evident. No selection device
can be effective if it’s low in reli-
ability. Using such a device would
be the equivalent of weighing your-
self every day on an erratic scale. If
the scale is unreliable—randomly
fluctuating, say, 10 to 15 pounds
every time you step on it—the re-
sults will not mean much. To be ef-
fective predictors, selection devices
must possess an acceptable level of
consistency.

What iS VaLidity? Any selection
device that a manager uses—such as
application forms, tests, interviews,
or physical examinations—must
also demonstrate validity. Validity
Michael S. Williamson/The Washington Post/Getty Images