Hire the Best Candidate Every Time!

One of the most challenging tasks for managersAdditionally, the incremental value of each is 27%
and executives is putting a selection processand 24%, respectively. Averaged across
together that truly selects the best candidate foroccupations, a measure of GMA plus a
a given position. If you polled 100 companies, youwell-designed work sample test will predict 42%
would likely find that each company has its ownof a candidate's performance; GMA combined
distinct process and that the results over timewith a structured interview will predict 40%. A
from that process have been mixed at best. Thefurther advantage of both of these processes is
most important aspect of the hiring process is itstheir versatility in the selection of entry level or
predictive value: the ability to predict the futureexperienced workers. However, work samples
job performance of a given candidate. So whatare easier to create for lower-level workers, who
hiring practices have the highest predictive value?typically have much more tangible results
Decades of research on selection processes hascompared to higher-level managers. For example,
resulted in extremely valuable knowledge aboutit is much easier to use a work sample to count
the utility of various selection methods and thehow many widgets a line operator produces than
superiority of specific selection assessments forit is to measure how well an HR director handles a
making the best hiring decisions.conflict situation.
Research over the past 25 years has shown thatFor this reason, structured interviews typically are
variability in work performance among incumbentused in place of work samples for candidates at
workers is very large. One way to standardizethe manager level or above. Structured interviews
this variability is by measuring the dollar value ofare highly versatile in that they offer a fixed
output. The accepted formula to measure theformat, can be designed through a job analysis,
dollar value of output is a 40% standard deviationand may have an accompanying scoring manual.
of the mean salary of the job (Schmidt &Conventional structured interviews typically consist
Hunter, 1998). For example, if the average salaryof a series of questions focusing on job
for a job is $40,000, then superior workersresponsibilities, knowledge, and achievements in
(those producing in the 84th percentile for theirprevious jobs; however, they generally are not
job, or one standard deviation above thebased on a formal job analysis. Two types of
average) produce $16,000 more than anstructured interview techniques that do rely on
"average" employee and $32,000 more than ajob analysis are the Situational Interview (SI) and
"poor" performer. Therefore, managers canthe Patterned Behavior Description Interview
increase the economic value of candidates hired(PBDI). In the former, the interview questions
by creating a process that reduces the inherentinvolve specific work dilemmas designed to elicit
variability among performance of those hired; i.e.the applicant's intentions as a measure of their
by improving the likelihood of hiring a 'highfuture behavior. The PBDI, in contrast, is based on
performer.'the premise that the best predictor of future
What does the optimal process look like? Almostbehavior is past behavior. Candidates are
100 years of employment research concludespresented with some criterion relevant to the job
that the most valid single predictor of future joband are asked to recall a relevant past incident
performance is General Mental Ability (GMA).and to describe their specific behaviors in that
Among the various assessment procedures usedincident. One added benefit of the PBDI, also
across all types and levels of occupations, GMAreferred to generally as behavioral interviewing, is
has the combined highest validity and lowestits flexibility in allowing candidates to use examples
application cost (Moscoso, 2000). As a stand-alonefrom community engagement, volunteering
assessment technique, a work sample has slightlyexperiences, and educational history in addition to
higher predictive validity but is much more costlyexamples from work.
than a measure of GMA. ForRegardless of the format, both conventional and
professional-managerial jobs, GMA predicts roughlystructured interviews demonstrate significant
34% of performance success in that job. Forpredictive value. In fact, research has shown that
lower level administrative jobs, it predicts 26%.the structured interview is one of the best
For this reason, any selection process should usepredictors of job performance and training
some measure of general mental ability as aproficiency, and it generalizes across occupations
primary differentiator between candidates. Fromand organizations.
there the question is which procedure(s) will addIn economic terms, the gains from increasing the
substantial incremental validity above GMA withoutvalidity of hiring methods can amount over time
prohibitive costs?to literally millions of dollars. Many organizations
Assessment Centers (AC's), a strategy employedrely solely on unstructured interviews for hiring
by numerous selection processes, meet the firstdecisions. In a competitive world, these
criteria but not the second. A 1998 review oforganizations are creating an unnecessary
hundreds of assessment procedure studies founddisadvantage for themselves. By adopting more
that AC's have substantial predictive validityvalid hiring procedures, specifically a combination of
themselves but only add a 2% increase in validitygeneral mental ability and either a work sample or
when combined with a measure of GMA. In othera structured interview, they could drastically
words, applicants who score well on measures ofimprove their competitive advantage.
intelligence typically also perform well in AC's, soOur selection assessments at Roselle Leadership
there is little to no additional value in puttingStrategies, Inc. include four ability tests aimed at
candidates through expensive assessmentcalculating a candidate's true GMA, as well as
centers, which typically cost thousands of dollarspersonality tests, proprietary open-ended
per candidate.sentences and a structured interview aimed at
On the other hand, both work sample tests anddetermining how a candidate is likely to act and
structured interviews offer the same predictiveperform on a day-to-day basis. Our primary
value as GMA, and significant incremental valueobjective is to reduce the variability in
when used in conjunction with GMA. Aperformance of employees hired in order to
well-designed work sample test will predictincrease dramatically the economic value of each
approximately 29% of a candidate's performancehire, specifically, and of the organization, generally.
and a structured interview can predict 26%.