“Successive Hurdles” Improves Hiring!

In the typical hiring process of a public agency, there may be a battery of tests (or hurdles) that all applicants must pass through. These include:

1) job application
2) initial interview31
3) physical agility test
4) psych evaluation
5) written test
6) criminal background check
7) reference check
8) credibility assessment (lie detection)
9) medical screening

The applicant must successfully pass each phase to continue the process. In most hiring processes, if an applicant fails one test, that person is disqualified.

Organizations should start with the least expensive hurdle and then add additional hurdles that are more expensive—as the applicant progresses through the hiring process. Each step or hurdle will filter out more and more applicants and less money would be spent initially on less expensive hurdles used in the hiring process.

When considering the cost of each hurdle, applying an EyeDetect test at the beginning of the process would not only be one of the least expensive steps, it would filter out of the group those applicants with disqualifying behaviors so the agency can spend time on more valuable applicants, thus improving the base rate of honesty of the entire applicant pool. In addition, EyeDetect will not be accused of discrimination based on age, sex, race, religion, etc.

EyeDetect mitigates risk and saves money and time.

Here’s how to use EyeDetect + polygraph to identify truthful individuals in your applicant pool with 99% confidence in the test outcome!

Accuracy rates for EyeDetect and polygraph (Source: APA Meta-Analytic review, Table 2, 2012.)
EyeDetect: 0.88 (True Neg.) / 0.83 (True Pos.) — Polygraph: 0.72 (True Neg.) / 0.81 (True Pos.)

First Hurdle – EyeDetect

Let’s test 1,000 candidates with EyeDetect, where we believe 20% are probably guilty of disqualifying behaviors and will lie on the test (deceptive). Those that fail the test will be eliminated from the hiring process. With 1,000 people and a base rate of 20%, there are 800 truthful and 200 deceptive candidates in the group.

For these tests, we will use the EyeDetect and polygraph accuracy rates quoted above.

EyeDetect Test Results
Truthful candidates: 704 pass and 96 fail the test (total 800)
Deceptive candidates: 34 pass and 166 fail the test (total 200)

Confidence: We’re 95% (704 people) confident that those passing the test are ‘truthful’ — but also realize that 5% (or 34 people) are really ‘deceptive.’ And we are only 63% confident that those that failed the test (total 262) are actually ‘deceptive’; this means 37% (or 96 people) of those that failed were actually ‘truthful.’

Summary: In this example of a successive hurdles model, there were 262 that failed the EyeDetect test and they would be eliminated from the hiring process. In that group, 96 were actually truthful. This is the challenge with any test that is not 100% accurate. But, there are 738 individuals remaining that passed the test.

The next test to be administered is polygraph. But the new ‘innocent’ base rate for the remaining candidates is now an impressive 95%. Some agencies or companies might be satisfied with 95% confidence in the test outcome and would conclude the credibility assessment process. If, however, a higher rate of confidence is required, a second test may be conducted.

Second Hurdle – Polygraph

Now the 738 ‘truthful’ candidates will be tested with polygraph; this means we are 95% confident in the test outcome for ‘innocent’ examinees.

Results of Polygraph Test
Truthful candidates: 507 pass and 99 fail (total 606 with 98 inconclusive)
Deceptive candidates: 3 pass the test and 28 fail (total 31 and 3 inconclusive)

Confidence: There is now 99% confidence in the test outcome that the group that passed the test is comprised of truthful candidates; in reality, there would be 507 truthful and 3 deceptive in the group that passed the second test.

Summary: We tested 800 truthful and 200 deceptive candidates. In this example, with a base rate of deception of 20%, we identified 510 ‘truthful’ candidates of which 507 are actually truthful.

This is an impressive 99% confidence in the test outcome!


7 + 3 =