Three tips for checking references
An obvious challenge related to checking references is that candidate-provided references are likely to show a positive-bias toward the candidate (although there may be exceptions). Below, I outline three simple steps Human Resource professionals, hiring managers, and recruiters can take to improve the quality and accuracy (i.e., overall validity) of the information gained through reference-checks.
Step 1: Identify key attributes (hard & soft) required for job success
This first step may seem like a given and, in reality, many actually take it for granted but it is the most critical step of the three. Taking this step lightly is akin to ordering a water-pump for a car without providing any other information (model, make, or year) about the car; there’s a high probability that one will end up with the wrong part. I’ll save the detailed procedure for accurately identifying key attributes for another time. For now, simply keep in mind that a job should be carefully examined (a full job analysis is always recommended for mission-critical positions) to identify a limited number of personal qualities (KSAOs – knowledge, skills, abilities, and other personality and value characteristics) necessary for success. I specifically say a “limited number” because while it’s tempting to list two-dozen ideal qualities, doing so will not improve hiring success; rather, it is more likely to make the search that much more confusing and challenging. For most positions, top three hard skills and top three soft skills should suffice. I will call these core competencies (CC) in contrast to non-core competencies (NCC) or those competencies, while useful or nice-to-have, are not essential to the job.
Step 2: Identify 3 additional NCCs for hard and soft skill areas (each)
You may be asking, why would you want to identify three NCCs for each hard and soft area? There’s actually a logical reason for this. The NCCs should be related to the job; just not important enough to be counted as CCs. In other words, they are secondary or “nice-to-haves” but not essential to job success. Take a software engineer position, a hard CC might be expert-level understanding of C# and Java. Ruby on Rails, on the other hand, might be a plus but not essential for job success; hence, Ruby would be considered a NCC.
Do the same for soft skills (e.g., dependability might be a CC critical for job success while having “high energy” might be a nice-to-have but not essential; thus, a NCC). Once you have identified 3 hard CCs, 3 hard NCCs, 3 soft CCs, and 3 NCCs, you’re ready to move onto the actual call. **Note: The actual number of CCs and NCCs can vary depending on the complexity of a given position.
Step 3: Use question formats designed to yield more objective responding
Rather than the typical “Would you recommend this person for the job at our company?” your first question should be something like: “Keeping in mind the candidate, can you name top 2 strengths and top 2 areas for improvement?”
This open-question is designed to zero-in on those qualities that are viewed as key strengths and weaknesses of the candidate and not necessarily related to the CCs and NCCs you have identified for the position in question. Nonetheless, response to this question should be compared to the CC and NCC list. For the ideal candidate, one or both strengths (identified by the reference) should correspond “roughly” with the qualities in the CC list that you have in front of you.
It is very possible that neither of the candidate’s top two strengths is found in the CC list. Hence, a second question should follow. Have the list of CCs and NCCs in front of you and ask the referring person to have a pen ready. Then, ask “How would you rank the candidate on the following list of hard skills from strongest to weakest?”
What you’re looking for here is to see how many of the qualities from your list of CCs end up on top (i.e., most strong) and how many end up at the bottom (least strong). The more of the CCs that end up on top (those viewed by the reference to be key strengths of the candidate), the higher the candidate’s score should be – say on a scale of 1 to 5 (the scale you use is entirely up to you). On the flip side, if more of the NCCs end up on top or more of the CCs end up on the bottom, the lower you should rate the candidate.
One final question is to pick out the quality that you deem most important for the job (i.e., one of the qualities on the CC list). Let’s say this is “reliability.” You may want to ask “How would you rate the candidate on a scale of 1 to 10 on reliability?” Be sure to define the scale as well as your conceptualization of “reliability.” Any response below 9 should be considered a red flag – meaning that the reference does not find the candidate exceptionally reliable. You’re looking for either a 9 or a 10. This technique is backed by research on faking (e.g., for personality tests) where the inflation or faking rate is around 15%. This means that on average, while a reference source may inflate his/her rating of a candidate, he/she is less likely to give a rating of 9 or a 10 if he/she truly feels that the candidate is not reliable.
I hope you find this information useful. There’s a lot more to be said here but I wanted to keep it to roughly a page. Happy reference-checking!