particularly
those with repetitive manual handling that have substantial physical
demands. Ideally, the employer is able and willing to reduce these
demands through ergonomic job re-designs, where possible. Unfortunately,
not all of the demands can be eliminated.
People with insufficient physical ability to meet the
demands are at increased risk of injury when they are placed on these
jobs and they are less likely to stay. As a basic business necessity,
employers need long-term employees who can safely perform the job. Since
some applicants will be denied employment on the basis of physical
ability tests, specific types of validation are required by federal law.
This is particularly true
of physical
ability tests, since some women and individuals over the age of 40, who
are specifically protected from unfair discrimination by law, will be
less likely to pass a physical abilities test if the job demands are
significant, thus causing adverse impact. Validation of the test
battery, as described in the Uniform Guidelines on Employee Selection
Procedures, provides the necessary evidence that any differences in pass
rate for protected groups reflect actual differences in ability to
safely perform the job.
All employment tests must be in accordance with Title
VII of the Civil Rights Act of 1964, the Civil Rights Act of 1991, the
Uniform Guidelines on Employee Selection Procedures (29 CFR
Part 1607), the
Age Discrimination in Employment Act, and the Americans with
Disabilities Act. A fundamental requirement common to virtually all
employment legislation is that any employment decision-making
(Selection) tools must be validated.
The intent of the research reviewed in this article was
to validate the effectiveness of ergonomically-based functional
screening tests for predicting risk of injury.
An ergonomically-based approach can be defined primarily in two ways. First, an ergonomic approach to the job analysis involved directly quantifying the physical demands of the jobs. For example, the heaviest weight routinely handled and how it was handled was quantified, and the aerobic capacity needed to meet the energy expenditure requirement of the job was determined.
One of the strongest study designs for predicting the risk of injury is to give new-hires the test battery, and then place them on the job without regard to their test performance (Rosner, 2000, p. 579-580). The performance on the job is then monitored for those individuals over the course of their employment. Injury rates and retention can then be compared between new-hires who passed the battery and new-hires who failed.
An alternative method for assessing the effectiveness of a testing battery in relation to injury experience is to compare the performance of new-hires who began work before the test battery was implemented to new-hires who began after implementation.
This design is referred to hereafter as a “pre/post-implementation analysis.” The major benefit of this study design is that the testing program can be immediately used for making screening decisions rather than waiting until a sufficient sample of new-hires who fail the battery are brought on the job in order to meet the predictive study
sample size requirements.
The drawback is that the study design involves comparing performance from two different time periods and different pools of employees. Any other changes between those two periods may impact the ability to detect the effectiveness of the screening program.
This issue can be addressed by comparing the groups in relatively tight time periods, such as one year, pre- and post-implementation.
Although the pre/post-implementation analysis design is
not as strong relative to a predictive validation study design in providing evidence to make causal inferences concerning the implementation of the program, obtaining a consistent effect size across different populations, industries, settings, and time does provide a strong inference that the observed effects are due to the program and not to artifacts such as regression to the mean and selection bias. Taken together, the results of this type of study can provide a strong inference that the observed effects are due to the implementation of the testing process.
Opportunities arose to perform predictive validation studies of physical ability tests batteries for warehouse jobs in three different industries. The first study focused on the
selector job in
eleven food distribution warehouses. Over the course of a shift, a
selector manually handled thousands of pounds of product. The second
study focused on loaders in a soft-drink warehouse.
Like the grocery selectors, the loaders built pallets of
soft-drink cases, which were then loaded onto outbound delivery trucks.
Loaders also manually handled thousands of pounds of product over the
course of a shift. The third study focused on de-palletizers and
shippers in three retail distribution warehouses. Again, employees in
these jobs manually handled thousands of pounds of product over the
course of a shift.
After completion of the predictive validation studies,
similar test batteries were implemented in other warehouses with similar
jobs. Comparisons of injury rates pre- and post-implementation were
performed at 175 of these locations.
The process of implementing the
physical ability test battery was the same at all locations included in
these studies.
It consisted of four basic steps.
The first step was to ergonomically analyze the job
requirements for the purpose of quantifying the strength and endurance
demands. The second step consisted of designing a physical ability
battery that measured the significant job demands, as documented with
the ergonomic job analysis. The third step was to determine the cutoff
score for each test.
The fourth step was to gather data on injury experience
and retention for the study groups, and evaluate the effectiveness of
the battery.