Trend Watch www.shrm.org/trends
Imagine walking into your doctor’s office and being handed a prescription for mercury, the chemical element
found in old-school thermometers
that has long been known to be poisonous and even deadly when contacted
directly. Sounds crazy, right? But if you
lived 100 years ago, mercury was commonly used to treat many illnesses.
Today, we have much higher standards for medicine. Over time, our desire
to introduce more certainty into clinical
decisions gave rise to what is known as
evidence-based medicine. For example,
prescription drugs now undergo rigorous testing before they can be used.
Should human resources follow suit?
Could some of our common practices,
such as annual performance reviews or
unpaid parental leave, one day be considered the mercury of the business world?
Maybe so. Consider, for example,
that many European employers and
some U. S. firms rely on graphology, or
handwriting analysis, as an assessment
tool when hiring, even though studies
show that a person’s handwriting has no
bearing on his or her job performance.
Disconnects like these may be more
common than you think. A U. S. survey
found large discrepancies between what
HR practitioners believe works well and
what research has found to be effective.
In the era of “big data,” many HR
professionals are feeling the pressure
to provide proof that their practices
work. Although analytics are a great
first step on the path to developing
evidence-based HR, they should fit
into a broader framework. The Cen-
ter for Evidence-Based Management,
a nonprofit foundation, describes four
sources you should consider, along
with the quality of the evidence, when
making a decision:
• Findings from empirical studies pub-
lished in academic journals.
• Data, facts and figures gathered from
within the organization.
• The experience and judgment of
• The values and concerns of the
stakeholders who may be affected
by the decision.
Let’s say you want to transform your
organization’s performance management system. Here’s what evidence-based HR might look like in that scenario. First, meet with stakeholders to
understand their pain points and concerns. Then develop a few basic design
principles aligned with their goals, such
as increasing perceptions of fairness and
transparency and reducing administrative burdens.
Review relevant scientific literature
to answer questions related to specific
design features. For example, what
strategies reduce inflated ratings? Are
monetary bonuses effective in improving performance?
Gather organizational data to better
understand the legacy system: How often
do employees challenge their ratings?
How long does each step of the process
take? How have ratings been distributed
over time and across subgroups? And so
forth. Collect additional feedback from
stakeholders, including end-users, on the
When the plan is finalized, test the new
system on a small scale and make changes
as needed before a full implementation.
Finally, prepare a thorough evaluation plan
to assess the extent to which the new system
is meeting its intended objectives.
A new performance management sys-
tem might be a pretty radical step for your
organization, but having a predefined
way to evaluate the outcomes and contin-
uously tweak the program will help you
get it right. How evidence-
based is your HR?
Shonna Waters is the former
vice president of research at
By Shonna Waters
Gather good data to make better decisions.