The Recency and Primacy Effects in the Talent Acquisition Process

In the April crl_masthead 2010 Journal of Corporate Recruiting Leadership, I have an article about two very important bias factors in the hiring process. I’ll talk about them in detail and give you ideas for preventing them.

For now, I wanted to give you just a quick overview.

The two biases are the recency and primacy bias effects.

The recency bias error occurs when an assessor (i.e. recruiter, hiring manager, etc.) is overly affected by information that was presented later (more recently) rather than earlier in any given selection process. In contrast, the primacy bias error occurs when an assessor’s selection is made based on information that was presented earlier (primary information) rather than later in a process. And although the effects appear symmetrically opposing, the research shows that they occur because of different reasons, and that their implications can differ drastically. They are not equal but opposite.

An exemplary candidate who shows up late to an interview but does well on the interview itself may suffer the consequences of the primacy effect. A candidate who shows up on time, does well but says something toward the end of the interview that can be described as intensely negative, might suffer the consequences of the recency effect.

The body of research points to two process models on how decisions are made in the interview process. The first we will describe is the step-by-step (SbS) decision-making model, and the second is the end-of sequence (EoS) decision-making model. We call those models response modes.

The step-by-step decision-making model is where a decision-maker evaluates a candidate incrementally and develops a view of that candidate. That view becomes the “anchor.” Then as more information is presented, the new information is added to the anchor view and a new anchor (hence new view) is formed. This process occurs until all segments of the interview process are completed. With every new anchor and every new set of data about a candidate a new anchor is formed.

The end-of-sequence response mode is self explanatory. Judgment on a candidate is withheld until the final stage of the interview process.

The best example to understanding bias effects comes from a study performed by Scott Highhouse and Andrew Gallor of Bowling Green State University and Indiana University respectively. In their article titled “Order Effects in Making Personnel Decision Making” published in Human Performance in 1997, they performed a study to understand where bias effects take place.

In short they found that in a long interview process, regardless of the complexity of the interview and decision-making type (response mode) of the decision maker, that primacy dominated. In other words, the term we commonly use for our initial cognitive anchor, “first impressions” dominated the views of the interviewers of the applicants.

In the case of a short interview process, the results depended more on whether the decision-making process was EoS (end of sequence) or SbS (step by step). In simple and short interviews where the decision was withheld to the end of the process, primacy dominated. In short and complex interviews the most recent information presented by the applicant dominated when judged by the interviewer. In a short step-by-step decision making process the most recent information always outweighed the first pieces of information.

As a talent acquisition consultant, the results make perfect sense in my practice. In the larger scheme of things human beings are fairly predictable. When we are presented with a large amount of complex information quickly we don’t have time to make step-by-step evaluations and we reserve our decision to the end. In that case, the most recent information seems to be the best, and a bias effect occurs toward the most recent interviews in judgment and possibly the most recent candidates interviewed. When the process takes a long time, we become mentally tired and we simply rely on our first impressions of a given candidate and tend to choose the applicants we interviewed earlier in the process. In the case of a simple and quick interview process, we simply judge quickly, and primacy dominates.

The recruiting leader then needs to implement a system and process to mitigate these effects if they wish to hire effectively and to help their hiring managers judge both accurately and precisely. I’ll talk more about that in the Journal.