By Kathleen
Groll Connolly
Reprinted
by Permission.
If they are mathematically sound, legally defensible,
and predictive of a person's job performance-is it any surprise
that personality tests are more and more a part of hiring? No one
knows for sure how many employers use tests, but the number is growing,
according to professionals in the field of psychological testing.
Pre-employment tests can cover several areas. Basic
intelligence tests, skills tests, and multifaceted observations
are all part of the suite. This article specifically discusses personality
tests. These explore a person's basic motivations, attitudes, and
temperament.
If personality tests are in your hiring future, what can you reasonably
expect? What shouldn't you expect?
Select
the Right Test
First, plan on learning something about testing. Tests
are not all alike and there are many to choose from.
"The biggest misinformation we experience in
the marketplace," says Robert Hogan, Ph.D., a well-known test
designer and psychometrician, "is that a test is a test is
a test. There is very poor awareness of the differences in quality
among tests." Hogan is the author, together with his wife,
Joyce Hogan, Ph.D., of four workplace personality tests. They run
Hogan Assessment Systems in Tulsa, Oklahoma.
A quick look at the history of psychological testing
provides a clue to the confusion. The only tests that you should
consider for a hiring program are those that are mathematically
reliable and valid. This is a rigorous process that can take a test
designer a number of years and hundreds or thousands of measurements.
Yet this field is relatively unregulated, although
is it also highly litigated. There are new tests coming into the
market all the time. They may or may not have the research data
that you should require. It is a case of "buyer beware."
Don't be shy about requiring evidence of the research. A reputable
testing company will have no problem producing it.
A second issue is that some tests are sold for the
workplace but were designed for another purpose. Some of the oldest,
best-known, and best-researched tests were developed to detect pathology
in the general population. While they might be excellent predictors
of tendencies towards violence or dishonesty, they are generally
not predictors of a person's "bright side"-the collection
of visible personality characteristics that make normal people different
from one another. The bright side also has to do with the ways in
which personality predisposes a candidate to be a contributor.
Also, some personality tests have been designed with
questions about a person's religious beliefs and sexual practices,
which run the risk of being viewed as discriminatory. You must be
able to show that your use of testing is fair and does not adversely
impact a protected group, in case an unhappy applicant challenges
you. The surest method for meeting this requirement is a formal
job analysis.
So, three questions you should ask before a test comes
under serious consideration for hiring purposes are these:
1. Is it mathematically reliable and valid?
2. Was it designed specifically for use in the workplace?
3. Has it been fully tested by the test designer for possible adverse
impact? In other words, does the test give the same results independent
of one's race, gender, ethnicity, or other background factor?
Make Tests Part of a Process
Tests are not a stand-alone event. The consensus of
those we consulted is that tests have much potential for increasing
the quality of hiring, assuming you make testing part of a well-designed
and well-managed hiring process. And while you're designing the
hiring process, you'd best install a measurement system for your
company's employment experience as well.
David S. Miller is a principal with a consulting firm that has helped
the management teams of utility companies. During the past five
years, many had to reestablish themselves during deregulation and
restructuring. His firm, Human Resources Strategies and Solutions,
in King of Prussia, Pennsylvania, has worked with numerous downsizing
situations.
Consider the rigor of the hiring process they used.
One utility had 145 candidates for 25 jobs available in engineering
management. Miller and his team instituted a multifaceted process,
of which testing was one part. First, they reviewed the candidates'
technical skills. Miller says of the initial cut, "Skills data
narrowed it down to about 100. We needed to go farther. We were
also recruiting from the outside, to increase the talent pool."
So they required applicants to take an online personality
test. "We needed to level the playing field," says Miller.
"Testing delivered that."
Next, they used behavioral interviews based on the
test results. Two-member teams conducted the interviews. One team
member was always a highly experienced HR professional. "That
was the tipping factor," says Miller. "There is no substitute
for that."
Then they administered a critical thinking test to differentiate
among candidates' problem-solving capacities. By this stage, the
most desirable candidates were becoming obvious and some executives
were now competing with each other for those individuals. Therefore,
no interviewing team was able to make a job offer without consulting
with the rest of the group. "We tried to make decisions that
were good for the whole, not just one of the parts," says Miller.
At this final stage, they also considered diversity issues-both
of people's backgrounds and their relative professional strengths.
He says that tests "knocked out the extremes,"
meaning the people who were very clearly wrong for the jobs. He
used both the Hogan Assessment Systems Personality Inventory and
the Watson-Glaser test.
Measure the Results
Both Hogan and Miller mention that employers have
to keep records of company employment experience-preferably before
and after tests are introduced.
Assuming the test is well selected and correctly used,
employers will be able to prove the value of testing to themselves.
"If they (employers) keep records before and after they begin
testing, they'll see a decrease in negative business indicators-turnover,
accidents, insubordination, drunkenness-and an increase in positive
business indicators-lower turnover, fewer accidents, etc.,"
says Bob Hogan.
Miller thinks measurements need to be developed as
a part of the overall mission. "First, you've got to gain alignment
around a vision," he says. "Then develop key performance
indicators as a part of that effort."
Interviews and Resumes Still
Matter
Can testing produce a clear thumbs-up/thumbs-down
indicator? When the issue is basic honesty for a security guard
or a retail worker, the answer is often yes. When the issue is which
candidate to select among a similarly qualified pool of professionals,
it is far from clear.
"We were interviewing a lot of engineers with MBAs," recalls
David Miller of the diligent management selection process he conducted
for utilities. Testing, he says, "served a winnowing process.
But it's not necessarily going to help you make the final cut."
This problem is greater when the candidate pool is uniformly qualified.
Says Miller, "When you boil people down to paper, they all
begin to look alike."
Some might say that the application of personality
tests gets more complex in higher level hiring. Edward Hoffman,
Ph.D., is the recent author of Psychological Testing at Work (McGraw-Hill,
2002). Hoffman is a psychologist based in Commack, New York. He
thinks that pre-employment testing is most useful for entry-level
positions. "For many kinds of entry-level positions, such as
in sales, pre-employment personality tests are very useful in screening
out potentially undesirable employees. Integrity testing, for example,
has a very high success rate .. Personality tests in the hiring
process that measure extroversion are quite useful, as are those
that measure conscientiousness, anger-proneness, or low stress-tolerance."
So how important are interviews and the resume? According
to Hoffman, who has written extensively on psychological testing
and has designed tests as well, "There is no single rule to
apply as to how to weight various relevant factors in the hiring
process. For certain types of jobs, standardized psychological tests,
including personality tests, are probably more important than how
the person comes across in the interview." As pointed out earlier,
these jobs are often entry-level. At higher levels, more experienced
candidates are likely to appear. Personality testing still has a
role-but that role in the hiring process is not as clear-cut.
He notes that the candidate's relevant work experience
is still the single most significant indicator of suitability for
many positions.