Unconscious bias (also called “implicit bias”) has become a trending topic—both in the general media and in the HR world. The topic of unconscious bias is often cited when considering ways companies can improve their diversity and inclusion efforts by recruiting and retaining diverse talent. It’s also critically relevant in the context of conducting workplace investigations because an essential duty of any workplace investigator is to conduct independent, unbiased inquiries about allegations of workplace misconduct.
What Is Unconscious Bias?
Numerous factors influence people’s unconscious thoughts. Whether from the news, movies, or social media, we are bombarded with stimuli that inform our beliefs about certain groups and political issues. In fact, research shows that our brains receive tens of millions of pieces of stimuli at a time, but we can only consciously process about 40 of those pieces. We form beliefs on the basis of this input (including our interactions, experiences, and exposure, or lack of exposure to certain groups). Some of these beliefs represent accurate information and some reflect prejudices and biases.
Unconscious bias refers specifically to those beliefs that reflect attitudes and stereotypes that inform our subconscious information-processing.
The Heightened Risks of Unconscious Bias During Investigations
The research related to unconscious bias at work primarily centers around recruiting, hiring, and retention. Nevertheless, the research related to unconscious bias is equally relevant in the realm of workplace investigations—where, in many cases, the dangers of unconscious bias are even more pronounced than they are in recruiting processes. For example, in the recruiting and selection process, the hiring manager typically does not know the applicant and makes assumptions based on the demographic group to which that applicant belongs.
There are at least two additional issues at play during workplace investigations. First, there is a possibility that an internal investigator may know at least one of the involved parties and might therefore make assumptions based on what he or she believes about that employee (for example, whether the employee is hardworking, a good performer, trustworthy, etc.).
In one example from my own practice, an employee (we’ll call her Angelina) approached an HR employee (we’ll call her Jennifer) to let her know that a colleague (we’ll call him Brad) had been behaving inappropriately toward her for some time. Angelina said she did not report the behavior earlier because she and Brad were equal-ranking coworkers, but because Brad had been recently promoted to a management position and because she would now be reporting directly to him, she wanted HR to know about his behavior. Jennifer’s first response was to tell Angelina that she had observed quite a bit of interaction between Angelina and Brad and that she therefore found it hard to believe that any “harassment” could have occurred. Jennifer felt she was making a fair assessment because it was based on her observation. What she didn’t recognize, however, was that she was making a judgment based not only on very limited observation (she saw them eating lunch together a few times) but also based on an assumption that friendly interaction between the employees negated any possibility that Brad might behave inappropriately toward Angelina in another setting.
Second, workplace investigators run into the danger of judging parties on the basis of their own preconceived notions about people in certain positions, making certain salaries, and with different levels of power at work—and possibly based on the investigator’s assumptions about the credibility of members of certain groups. An investigator may be biased based on outside influences, such as the media’s portrayal of harassment at work, reports of employees receiving disproportionately high money awards because of seemingly mild misconduct, or past experiences with employees whom they believe took advantage of the system. I see this often when it comes to bias related to line employees versus management employees. Such an investigator may have preconceived notions about the validity of all complaints or, alternatively, may believe all employee complaints are valid and exhibits bias toward management employees.
Three Types of Unconscious Bias in Workplace Investigations
There are three relevant ways in which unconscious bias may manifest itself during workplace investigations: (1) confirmation bias; (2) “like me” (or not like me) bias; and (3) priming.
1. Confirmation bias
Confirmation bias refers to an individual’s natural tendency to give more weight to information that tends to confirm his or her preconceived notion (or, conversely, to give less value to information or evidence that contradicts an existing belief). For example, an HR manager who tends to disbelieve employee complaints may have a tendency to collect and analyze evidence supporting that belief. Perhaps the investigator will go over the reporting party’s performance evaluations with a fine-tooth comb seeking to find that he or she had performance issues. An investigator may also opt not to interview certain individuals identified by the reporting party, perhaps dismissing these additional witnesses as also having bones to pick with the accused party. Of course, examining an accusing employee’s evaluations and failing to interview certain parties might be perfectly appropriate depending on the investigation, so long as those things are not done in an effort to substantiate a bias about one of the parties to the conflict.
2. “Like Me” Bias
“Like me” bias occurs when the investigator tends to favor information received from witnesses who are, in some respect, “like” the investigator. This affinity might be based on similar backgrounds, positions in the company, or interests. One famous study examined why female musicians made up such a small percentage of positions on elite orchestras. The researchers observed that when orchestras started auditioning musicians behind a screen, the number of women selected to play in these prestigious orchestras dramatically increased. One elite orchestra, after noticing that even the screened auditions favored male musicians, asked all musicians to remove their shoes before auditioning in case the “clicking” sound of female musicians’ high-heeled shoes was giving away their gender.
This study is noteworthy for at least two reasons. The first is that it shows that unconscious bias can come into play even when the person making the decision—in this case, the conductors and other decision-makers—may have a sincere desire to be fair during the selection process. The other is that there are ways to creatively tackle issues of unconscious bias without negatively affecting outcomes, such as, in this case, modifying the selection process by asking musicians to remove their shoes before auditioning.
Finally, priming refers to a phenomenon in which our reactions to stimulus are affected by our exposure to another stimulus. One classic study provides an example of the priming phenomenon. When researchers asked a group of study participants who were shown video footage depicting an auto accident, “About how fast were the cars going whey they hit each other?” the average response was 34 miles per hour. When the question asked was, “How fast were the cars going when they smashed into each other?” the average response was 40.8 miles per hour. In another experiment, 16 percent of participants reported seeing broken glass when the word “smashed” was used and only 7 percent said they saw broken glass when the phrase “hit” was used—even though there was no broken glass in the video.
In the workplace, an investigator who believes he or she is unbiased might unconsciously use charged words while questioning a witness, thereby influencing responses. For example, an investigator might ask a leading question, such as, “Well you didn’t think that behavior was inappropriate, did you?” Similarly, saying someone has “made allegations of sexual harassment” will likely elicit a different reaction than saying an employee “has raised concerns about inappropriate behavior.” As investigators, what we say and how we say it matters.
How Unconscious Bias Affects Workplace Investigations
These types of unconscious biases can and sometimes do affect investigations at various phases. Here is how unconscious bias can have an effect at different stages of an investigation.
Beginning of an Investigation
Unconscious bias may play a role at the beginning of an investigation when an investigator or management employee is deciding whether to conduct an investigation in the first place. In the Angelina/Jennifer/Brad example, Jennifer may have been wrongly letting her unconscious bias influence her decision to investigate whether Brad’s behavior was appropriate.
Gathering and Analyzing Evidence
Unconscious bias may also be at play while gathering evidence, i.e., conducting interviews, reviewing relevant documents, and analyzing the data gathered. Investigators may want to consider how the three relevant types of unconscious bias could affect the investigation while gathering and analyzing evidence.
1. An investigator’s confirmation bias might come into play when he or she is:
- choosing which witnesses to interview on the basis of a mistaken belief about which individuals can offer relevant evidence;
- choosing which evidence to review and perhaps mistakenly focusing on evidence that confirms his or her preconceived ideas about the accused, the accuser, or the nature of the allegation; and
- analyzing evidence in such a way that confirms the investigator’s preconceived beliefs about the situation.
2. Establishing rapport with witnesses. An investigator might feel more comfortable with some witnesses—i.e., may exhibit a “like me” bias—because the investigator has something in common with the witness. Investigators may, in turn, create a more comfortable or safer environment in which to share information. The investigator may also give more credence to what the “like me” witnesses say.
3. Asking skewed questions. An investigator might be inadvertently priming witnesses by asking them leading or charged questions that may influence the witnesses’ responses.
Finally, unconscious bias might infect decision-making when a determination needs to be made about whether there, in fact, was misconduct, the identity of the wrongdoer, and the appropriate remedial measure to implement assuming there is a finding of some misconduct. For example, an investigator may reach a conclusion that favors the investigator’s preconceived ideas about the conflict and/or the parties involved. The investigator might also recommend a remedial measure that’s more lenient on a “like me” offender.