Abstract

Harvard University’s Implicit Association Test (IAT) has been lauded as a tool for assessing implicit bias in a number of areas (e.g., race, gender, religion). Despite the creators’ assertions that “instructors…should offer an alternative assignment” to the IAT, textbooks (e.g., Shaw & Lee, 2020) and pedagogical practices (e.g., Crutcher Williams & Wright, 2020) demonstrate a different reality. In exploring the IAT with students, problems arose. In some cases, the IAT triggered students who found that it counterintuitively promoted racial and gender stereotypes. Further, question phrasing and the design of the IAT left some frustrated by its inflexibility and student results. In redeveloping curriculum to be more inclusive, I consider ways to achieve the goals the IAT proposes and avoid negative outcomes, making a preliminary case for substituting it with other bias exercises and/or practices. My study offers an analysis of the test and documents the potential benefits associated with replacing it as a primary source, as demonstrated in data collected from three undergraduate sections of the course Introduction to Women’s, Gender, and Sexuality Studies from Fall 2020, Spring 2021, and Fall 2021. This proposed substitution allows students to grasp the impact of intersectional identity and hidden bias on their own terms while meeting the goal of personal reflection, thereby underscoring feminist pedagogical commitment to bias education.

Keywords

Diversity, Ethics, Feminism, Implicit Bias, Praxis

Introduction, Background, and Study Rationale

Since the IAT was released in 2011 by researchers at Harvard University and supported and furthered by other elite American universities and colleges, it has been widely used in a number of governmental (Staats et al., 2017) and educational contexts. It also continues to be used as a tool in diversity training and workshopping in private industry. In higher education and professional training the IAT is recommended and implemented across disciplines, from the communication classroom (Crutcher Williams & Wright, 2020) to the medical school (Hernandez et al., 2013), to the WGSS classroom and textbook (Shaw & Lee, 2020). Using the IAT, Project Implicit’s team offers education sessions and programming and consultation services in diverse sectors, including “academia, business, education, health care, law, military, and public policy” (Project Implicit, 2011).

Given that the IAT has gained ground as a tool since its launch and is used in so many settings, much research has investigated both its effects and its effectiveness—e.g., research reviewed by Staats et al. (2017)—in achieving the project’s admirable stated goal of “challenging organizational and institutional disparities” (Project Implicit, 2011). The research has delivered some positive assessments of the test, such as the early findings that provide data that suggest that “faking,” or falsifying and/or manipulating results on individual IAT “score” is unlikely to impact data significantly (Steffens, 2004).

As described, the IAT has been promoted as a good way to expose individuals to the concept of hidden bias and facilitate discussion of bias and is included in a number of reputable academic resources and studies. It is perhaps obvious that the prestige of the institution that developed the IAT influenced its inclusion in a textbook from a prestigious academic publisher (i.e., Shaw & Lee, 2020). Like some previously cited instructors, the inclusion and recommendation of the IAT for use in the educational setting by prestigious sources in multiple disciplines led me to craft a writing assignment that relied on the student’s participation in IAT as one of two primary sources. Inclusion in the assignment, described in more detail below, had some unforeseen negative outcomes, first documented by the second author. These negative effects partly inspired the current project’s inquiry into student perceptions about the IAT’s effectiveness in bias education and its greater implications.

What is revealed both by some student accounts and the inclusion of the IAT in prestigious (re)sources elides the ethical considerations laid out as words of caution by the creators of the IAT: that no one should be compelled or asked to take the IAT in a formal setting, where they are in a vulnerable position relative to the people or institutions asking them to take the IAT. Because these ethical considerations are hardly ever mentioned in sources like textbooks and peer-reviewed pedagogical studies, one can assume that their elusion is iterative. In other words, in practice, text-takers are often compelled to take the IAT in situations that are ethically questionable. Of course, the next logical question, outside of the confines of the present study, is where the IAT could be administered without alternative activities (and how opting into alternative activities is potentially stigmatizing for those who are offered the choice). Outside of the data the IAT gathers for Project Implicit’s own purposes, the service goals of the project (i.e., informing the examinee about their own implicit bias toward positive change) are what take center stage in most contexts.

It is my contention that, instead, the ethical considerations of asking people to complete the IAT in any formal context are what should take center stage. Informing this assertion is not only my research assistant Valaree Ford’s initial observations and the results presented below, but the way the IAT is most often used in practice. I believe this in-practice use is a direct result—despite controversies (outlined in Howell & Ratliff, 2017; Nadan & Stark, 2016; Clark & Zygmunt, 2014)—of the IAT’s status as a recognized tool created by elite researchers at prestigious institutions. One of the primary goals of any Women’s, Gender, and Sexuality course should be to question the assumptions built into historical notions of prestige as part of the larger feminist ethos of working toward creating a more equitable and inclusive society. Hence, these ethical considerations are especially significant in the WGSS classroom context.

Context of the Use of the IAT

A writing assignment offered in three regular (non-Honors College) 25-26-person sections of an introduction to WGSS course, from Fall of 2020 to Fall of 2021, asked learners to take several IATs as well as read Toni Morrison’s short story “Recitatif” to analyze and arrive at conclusions about implicit bias. Students were at no time asked to disclose IAT results. However, students were asked to take the test as part of a graded assignment. As stated, this is, in part, the motivation for the current study: to perhaps remedy a common oversight in practice as regards IAT participation.

The textbook employed by the majority of instructors of this course offers a learning activity toward the beginning of Chapter 2, titled “Systems of Privilege and Inequality,” that directs students to take IAT to facilitate education about implicit bias (Shaw & Lee, 2020). Further, as cited, other documented practices supported the use of the IAT in assigned and graded material and in formal workshop and training settings. At the outset, the purpose of the original assignment was to educate learners about implicit bias and spark self-reflection and discussion among a largely white and affluent group of students at a predominantly white institution of higher learning, in keep with the WGSS disciplinary definition that “emphasizes the fundamental intersection of gender and sexuality with other vital categories of human identity and experience, including race, ethnicity, (dis)ability and class” (James Madison University, 2021). Shaw and Lee (2020) state in their learning activity that “these tests can determine if perhaps you hold hidden biases concerning race, sexual identity, age, gender, or body image” (p. 45). So, the integration of the IAT via the textbook’s learning activity into an assignment about implicit bias seemed to, at face value, serve disciplinary objectives.

Likewise reading “Recitatif,” Morrison’s only published short story, offers the potential for readers to learn about hidden bias. The story tells the fictional tale of two girls who are roommates at an orphanage. The story is told retrospectively from the position of one of the girls, Twyla, and recounts their impromptu meetings over the course of their lives, from girlhood to middle-age. The story is useful for thinking about bias from a number of perspectives: gender, ability, age, class, and religion. However, race becomes the central issue because the two girls’ races are never revealed, aside from a reference to them being like “salt and pepper,” one black and one white. Instead, Morrison offers the reader a series of coded and ever-changing racial stereotypes that intersect with other stereotypes about gender, ability, age, religion, etc. Morrison’s overarching objective is clear by the narrative’s ending because the girls’ races are never revealed. Instead, the story works to make the reader complicit in the activity of decoding racial stereotypes—e.g., frequency of hair washing, entitlement, Christian devotion/faith, types of food, access to education, wealth, mental illness/wellbeing, music, etc.—to attempt to solve the mystery of the charters’ races. The overlap between the story’s potential outcome for readers and the textbook’s presentation of the purpose of taking IAT seemed to provide a well-rounded approach to the overarching goal of the writing assignment: the exploration of hidden bias and analysis of the potential effects of this phenomenon.

In exploring the IAT, I found it to be lacking in a number of areas: in its potential to introduce test-takers to biased conceptions about groups of people without attending to individuality, its assumptions about the identity of the test-taker (e.g., “I would be willing to have a sexual relationship with a Jewish person”), and its broad and transparent question format that both failed to attend to intersectional identity and seemed to point to explicit, rather than implicit, bias. The most alarming observation is evidenced in the following from Valaree Ford’s comments in the Fall of 2020:

I do know what it’s like to be a black person doing the black bias test (Figure 4.1). By which I mean, I couldn’t even finish it, even to get to the regular questions, because the button test was like an emotional wall to me. I couldn’t even feel angry about it, just tired. Seeing the blank stares from those black people reminded me of the collective tiredness we feel towards how unfair society treats us. Being made to compare their face to the negative words (hurtful, selfish, angry, nasty, dirty, tragic, and scorn) reminded me of a history of black faces seen as intimidating; nothing more than angry apes that have to immediately be bussed to prisons. Going from that to  using empty positive words doesn’t make up for having to experience that. Using hateful words that are steeped in our tragic history, and then turning around and using such trite compliments on the same people (my people) that I called dirty and angry, just makes those positive words (friend celebrate, terrific, cheerful, fabulous, spectacular, glorious, and glad) disingenuous. Frankly, it’s down-right traumatizing for a minority to have to this. The test doesn’t consider the societal consequences of how cold it handles the humanity of the racial groups it’s asking about.

Figure 4.1. Sample question from IAT test for race

Here, rather than experiencing the IAT as an activity that offered the potential for implicit bias exposure and self-reflection, the test-taker is demoralized and dehumanized.

These negative observations about the IAT’s content were at times supported by students during subsequent semesters. Further, the IAT’s question phrasing and the timed, no-review, iterative design and scale of the online test left some examinees at sea and frustrated by both the test’s inflexibility and the resulting perceived lack of accuracy. In some like cases taking the test counterintuitively led to student perception that the IAT promoted racial and gender stereotypes, while not offering the meaningful self-reflection that participating in the activity seemed to promise. In this vein, some student comments highlighted the binary presentation of the IAT as divisive.

However, many other students reported that they found components of the test revelatory and effective. In redeveloping curriculum for the online classroom, I considered ways to ethically include the IAT, per the recommendations of research assistant Valaree Ford in a crafting assignments for students in subsequent semesters (post-Fall 2020). It was clear that a change was necessary. Using a literary work that might fill the gaps in the test and offer alternatives for learners who are more progressive or sensitive to the potential perceived flaws of the IAT seemed the solution. In other words, a simple workaround could be to assign the IAT and the short story reading as alternatives to each other, where students would either choose to take the IAT or read the short story.

When taking students’ experiences of the IAT into account, this solution seemed too facile, however. How could students anticipate, prior to completing the IAT, that they might experience something they would later feel was harmful? Further, wouldn’t the digitally accessible format of the IAT and its inclusion in their assigned textbook compel students to more often choose to take the IAT rather than read a literary work (a seemingly much less straightforward and scientific approach)?

Therefore, I developed a method to assess student feelings about the IAT to both attempt to get a more well-rounded picture of the IAT’s perceived effectiveness and a closer look at specific comments to decide whether and how the IAT should be used in future assignments and what could serve as an effective replacement if it was discovered that student comments reflected the second author’s sentiments. Only in categorizing, quantifying, describing, and analyzing student comments over the three-semester period might I find the basis for making ethical amendments to the assignment for future courses.  

Methods

The writing assignment outlined above garnered 75 essays from 75 general education students, 25 from the Fall of 2020, 26 from the Spring of 2021, and 24 from the Fall of 2021. I sought to compile relevant comments from the submitted essays about the experience of taking the IAT into four categories: 1. negative comments about the test and results, 2. negative comments about the logistics of taking the test (how the test works), 3. positive comments about the test and results, and 4. positive comments about the logistics of taking the test.  

Category Creation

Division of student comments into positive and negative categories based on direct evaluative statements about the IAT and any disclosed results as well as process of taking the IAT (logistics) seemed the most logical way to proceed. These categories are crafted to be intentionally broad at the outset to capture as many relevant comments as possible. Categorizing comments should be straightforward: students report whether they felt taking the IAT was effective in revealing results that were personally revelatory and reflect on whether these revelations might inspire growth. Comments asserting the IAT’s ineffectiveness in any area would be categorized into as negative and further delineated into two sections as cited above: the test itself as well as any volunteered reaction to results (1) and the process of taking the test (2). Comments asserting the IAT’s effectiveness in any area would be categorized as positive and further delineated into two sections as cited above: the test itself as well as any volunteered reaction to results (3) and the process of taking the test (4). The latter category should be accounted for not only to delineate these types of comments from others about the content of test, but also because taking the test relies on the implementation of a new set of rules/skills (pressing specific keys on a keyboard to indicate examinee answers). Specific student comments about their own IAT results (e.g., “moderate preference for black people over white people”) and the logistics of taking the test (e.g., “was the IAT easy to complete?”) were not prompted by the assignment itself or in discussions leading up to and following the assignment.

I had to plan for the possibility that some comments might either positively or negatively “bundle” assessments of the taking IAT with evaluation of the experience of reading the short story. This bundling could make student reflections on the IAT specifically difficult to discern. Therefore, after reviewing all of the essays I created a fifth (5) and sixth (6) category of statements that bundled the IAT with the story in a way that made discrete assessments too difficult to discern. These types of comments were compiled into positive (5) and negative (6) categories to potentially disambiguate comments about the IAT and short story and make final totals as accurate as possible (see Table 4.1). Working in this way would allow us to get a broad picture of student perception of the IAT’s effectiveness in terms of exposure to the concept of implicit bias and any potential personal revelations/reaction to results that the students chose to disclose.

Table 4.1. Categories of student comments on taking the IAT

Analysis of Compiled Comments

Once all relevant comments are compiled and categorized from the 75 submitted essays (with multiple comments potentially pulled from each essay, as are relevant), the number of individual comments from Category 5 (positive statements that “bundled” the IAT with “Recitatif”) should be deducted from the number of individual comments in Category 3. The same process should be performed for the bundled negative comments (deducting the number of comments that fall into Category 6 from the number of comments in Category 1).

Categorization should offer the potential to get significant results while setting the stage for qualitative analysis, a more granular look at where issues with assigning the IAT might have cropped up as well as possible remedies suggested by the results. For instance, did students think that taking the IAT was effective in revealing implicit bias? Did they feel that the activities (taking the IAT and reading the short story) served as complements to each other in uncovering bias? This categorization would also allow us to interrogate any of our initial misperceptions; in other words, if the IAT did what it set out to do, the positive comments should bear this out. This approach would also shed light on perceived positive takeaways in terms of changed mindsets and, perhaps most importantly, potential future actions.

If the goal of both the reading story and taking the IAT is to expose students to implicit bias and provoke discussion, critical thinking, and personal reflection, then any hindrance to that process could be identified through our granular examination of specific comments within the compilations, particularly when they are in-spirit repetitions of each other. Potential hindrances in the context of the IAT could include, for example, results’ rejection (i.e., “my results were incorrect”), mis-keying/typos, frustration with the test’s design, intentional falsification of results (e.g., because of perceived judgement), inconsistent results between multiple attempts on the same section of the IAT, perception that taking IAT normalizes or naturalizes bias (making it an unsolvable problem in the mind of the test-taker), a triggering effect also relevant to the short story (whereby the student is first exposed to harmful stereotypes by the test), etc.

Of course, test-taker discomfort has been cited as having the potential for a positive outcome in the long term (Staats et al., 2017). Certainly, being confronted with one’s own potential bias is uncomfortable and the first step in understanding bias and its negative implications in practice in the world. However, when test-takers cannot move beyond frustration or discomfort or when revelations about bias are generalized, normalized, or naturalized, the hidden bias activities have failed in so much as the goal of prompting analytical thinking about this important topic cannot be met. It should be noted that the short story also has the potential to create discomfort and frustration in the reader in its unresolved ending (in terms of the characters’ races).

Roadblocks to implicit bias education, reflection, and analysis could be accounted for in our closer examination of the compiled comments, as time allows and as relevant. Indeed, the data cannot fully address the value of taking the IAT in the educational setting for a number of reasons. The most obvious biasing factor in assessing negative comments about the IAT is the fact that it was part of an assigned project. Many students would feel compelled to offer minimal critique of the activity and their results, as it was presented as a valuable resource prepared by an elite group from the upper-echelons of higher education through its very inclusion in the course. This, of course, could also apply to the short story. However, its lack of inclusion in the textbook and the fact that it is literary fiction minimizes this type of biasing effect.

Looking in detail at illustrative and relevant negative and positive statements could reveal more specific information or allow for a greater range of conclusions to be drawn from the feelings of the examinees post-test. Further, this qualitative approach can supplement the findings by identifying student suggestions for possible solutions to potential issues that arose for them while taking the IAT, thereby working toward meeting the overarching assignment objective of presenting and examining implicit bias.

Results

Results of Comments about the IAT

Of the total 102 relevant comments compiled, slightly more than half, 55 (or roughly 53%), were positive. Despite prior unforeseen ethical concerns about the way the IAT was administered (as part of a graded assignment), it seems the majority of comments about the IAT were generally positive, with 53 discrete positive comments about the IAT’s content and results and 2 discrete positive comments about the logistics of taking the test. However, almost half, 47 (or roughly 46%), of the comments were negative, with 33 discrete negative comments about the IAT’s content and results and 14 discrete negative comments about the logistics of taking the test (see Table 4.2).

Table 4.2. IAT categorized breakdown by number of comments

With a gap of only about seven percentage points in a small sample size of 102 relevant comments from 75 essays, it is difficult to unequivocally say that the majority of comments, never mind students, expressed that completing the IAT was a distinctly positive experience.

To get a clearer picture of student perceptions about the value of taking the IAT, it is necessary to examine particular statements of critique of the IAT, both in Categories 1 and 2 (negative assessments) and in Categories 3 and 4 (positive assessments).

Negative Assessments of the IAT’s Content and Results (Cat. 1)

Potential hindrances to the provocation of discussion, critical thinking, and personal reflection on the issue of implicit bias in the context of the IAT cited above are represented in some of the negative assessments of the IAT’s content and results. One such example falls into the area of results’ rejection: “My results can only be looked at in the scope of this research survey test and do not determine the extent of my bias nor does it capture my personal beliefs about race and my actions with other individuals” (Fall 2020). While, as previously discussed, discomfort can be seen as the first step in education, other students who felt negatively about the content of the IAT and their results explicitly ruled this possibility out. An essay from the Spring of 2021 asserts, “Discomfort can often derail us from becoming self-aware.” In other words, far from perceiving discomfort as first step toward an eventual positive change, some students understood this feeling as a baffle to change.

Other students who felt negatively about the IAT cited their anxiety about their potential results while taking the test as a factor in their negative assessment:

The test required me to put people into boxes, and it felt wrong, especially people who are not like me…I hated the process of reading a sentence, such as these people ‘do not command respect’ (Young 45) and having an immediate group of people come to mind. I was raised by parents who taught me to celebrate differences, to not be color blind and recognize my privilege…When I was taking the hidden bias test, I was so concerned with being wrong. I thought that putting the wrong people in the wrong category meant that I was a bad person. (Student, Spring 2021)

This learner describes a moral and ethical bifurcation of their identity between their upbringing/morality (i.e., “being a good person”) and their potential results. They recount that this process unfolded as they took the test. This effect of participant fear in the face of being presented with their own potential immorality, though perhaps unintended by the IAT’s developers, was a common assertion from those who viewed the IAT negatively. One student essay from the Spring of 2021 recounted:

            the anxiety and self-monitoring I experienced while taking the IAT…This fear I had of     getting an unwanted test result caused me to have anxiety and monitor my answers. I    worried about answering the questions the correct way even though there are no    correct answers. Even though I tried to answer the questions honestly, my emotions          and fears took over causing the IAT to have the possibility of human error. I did not       like the test…however, I am not rejecting the results.

Here, the student makes an important distinction: they are not rejecting the results as some others who felt negatively did. Yet, their comment makes clear their perception that their anxiety had the potential to negatively influence the accuracy of their results.

Some negative comments, rather than rejecting results or qualifying result (in)accuracy, instead indicated that the IAT’s disclosure of hidden biases was ultimately meaningless. An essay from the Fall of 2021 states, “Bias tests may awaken people in discovering an unconscious bias, but the test will never provoke enough change to eliminate racism or sexism. I am sure most people that have taken the bias test have not changed their ways and still believe in the same things as they did before they took the test.” Here, the learner questions the overall effectiveness of the test as a tool for positive change in the world. Sometimes, these types of statements took the form of generalizations about unidentified unenlightened “others.” One such assertion, duplicated in spirit by at least one other student, appears in an essay from the Spring of 2021: “I think for…many [people] they don’t believe it and don’t want to listen on how to educate themselves to change; they’d rather stay ignorant and happy because they’re most likely white and bias doesn’t affect them.”

While generalizations like the previous comments are less compelling in terms of determining student feelings about the IAT than more specific and arguable claims, some students’ observations cut to the heart of the WGSS discipline. Another comment from the Spring of 2021 points out that “The categorizations we make are organized in a very binary way being strictly female or male, which leaves a lot of individuals from the LGBTQ+ community out of the picture.” This comment clearly critiques the binary foundations of the content of the IAT for non-heterosexual people. Problematic on a number of fronts, the IAT’s potential “binaryness” has not only a negative effect on the individual student observing this, but undercuts the WGSS course’s core aim of educating students about diverse sexualities and genders in an effort to breakdown binaries.

Negative Assessments of the Logistics of Taking the IAT (Cat. 2)

In many of the negative assessments of the logistics of taking the IAT potential hindrances to course aims arose. Some such hindrances, cited above, are mis-keying/typos and frustration with the test’s design. One comment from the Fall of 2020 reports, “During the test I messed up a couple times, and I felt anxious that it would alter my results drastically…I would forget what I was doing for a split second and place a picture or word in the wrong category.” Here, the test-taker asserts that they felt their anxiety distracted them and resulted in incorrect or unintended responses.

While the previous student’s cited anxiety and distraction are slightly different than statements about mis-keying, the result is similar: a feeling that the results might not be accurate because of these “errors.” A student from the Fall of 2020 says, “Throughout the test, I felt like I answered every question to the best of my ability. My only moments of error were mis-keying due to trying to complete the test quickly as suggested…I do not necessarily agree with the results.” Here, the test-taker feels they answered honestly, yet seems to associate mis-keying with their circumspect statements about their IAT results. Others who felt they had also mis-keyed directly attribute this to the IAT’s design:  

            When taking the race IAT, I found it easy until they displaced my already established       rhythm. I only noticed this displacement because I felt like I was sure about what            category the picture or word belonged to by title but not by the position it was now           moved to. In other words, I felt like the change in the new association of the words      was not the problem, instead, the problem arose because they changed the keys I was   supposed to press for the category. (Student, Fall 2020)

This student’s account of mis-keying points to the alteration of response keys and their correlating answers between different bias tests as the mitigating factor.

In the vein of IAT design negatively influencing the accuracy of results, some comments observed the IAT’s game-like qualities. One student from the Fall of 2020 writes, “It all seemed less like a test and more like some sort of game. I thought that perhaps the format of the test would affect my results because it seemed to me that it was a somewhat difficult task that could interpret mistakes in the test as sign of unconscious bias.” Another student from the fall of the following year (2021) echoes the previous comment more vehemently in their assessment of the IAT’s game-like quality: “These tests do not show whether you are biased or not…simply saying click ‘E’ if light skin and ‘I’ if dark skin, which becomes a mind game to the eyes and brain, which does not show if one is biased or not.”

Other students who remarked on the game-like quality of the IAT as a source of frustration articulated that this might influence results. One student from the Fall of 2020 recounts, “I found myself at times identifying a pattern and falling victim to the ‘practice effect,’ when you change your behavior even slightly once you know how the game is played…These hidden bias tests do not thereby invoke change.”^[1] This criticism when paired with the comment from above about answer-key switching reveals that the IAT’s design attends to the practice effect. However, test-takers are not privy to the specifics of the IAT’s design, and so both these students perceive a flaw in the design, one because the rules of the game seem to change between tests and one because they feel they do not change enough and can be manipulated. There has been some previously cited research that invalidates the concern that examinees can effectively misrepresent their answers to alter results. It is clear, though, that some students felt differently because of they observed the game-like qualities of the IAT and, hence, in criticizing the design of the IAT, applied their understanding of the premise of a game (i.e., consistent rules) or knowledge of psychology (i.e., practice effect).

Positive Assessments of the IAT’s Content and Results (Cat. 3)

Many comments reflected positive sentiments about the IAT. The most repeated in-spirit comment overall comes from a Spring 2021 student: “The ‘Test for Hidden Bias’ truly shows that bias can fly way past over heads without even noticing it. Although biases can be profoundly evident in any situation, understanding and recognizing bias along with being able to change those behaviors is extremely beneficial to growth.” Another student from the Fall of 2020 offered the following comment: “In general, I believe that people should take this test to help them become self-aware of stereotypes, prejudice, and bias they may have, to change societal norms and end negative stigmas around specific groups.” It is worth noting that this comment and the like emphasized an IAT role in promoting self-recognition but qualified the significance of these personal revelations by referring to potential resulting action on the part of the individual or society writ large.

This qualification did not lead some students to negate the significance of the IAT, however. A learner from the Fall of 2020 writes, “Even though these IAT tests made me uncomfortable for certain aspects, I felt like it made a positive impact on me because being uncomfortable may be negative at the moment, but beneficial to me in the long journey of my life.” Here, previously cited observations about discomfort leading to positive change in perspective and action seem to be indicated. Most revelatory I think are observations like the following, which cite the overall scientific format as impactful:

            I think the tests are effective in their purpose due to the method of presentation…a            straightforward process, easy for an individual to grasp…an input-output basis: you      take the test, you get immediate results with understandable percentages. The tests           offer a measurable scientific result which can serve to soften the blow when dealing    with introspective topics. I believe the tests to be a highly effective tool for helping            people come to realize their biases.   

This student from the Fall of 2021 and one other focused on the appealing straightforwardness of the IAT. On some level, this speaks to the logistics of completing the IAT. Yet rather than taking a specific look at the way the IAT is completed, these positive assessments more generally cite the simplicity of taking an online test.

Positive Assessments of the Logistics of Taking the IAT (Cat. 4)

Only two comments total, both from the Fall of 2021, went into depth about positive aspects of the logistics of taking the IAT. The first comment of this kind focused on the speed at which the examinee is instructed to complete the tests:

            The first thing I noted was that the instructions for the test start with the phrase ‘Go as      fast as you can’ when sorting words or images (Project Implicit). It keeps the test     subject from thinking too carefully about sorting, forcing the subject to make quick           judgments, which aids in test-taker complicity. Through the quick sorting, the test      pushed me to associate the words/race and gender/career/names with the categories,       for each respective test, even when switched. The switch was meant to test me on what     biases I had attached to each category. This was to discern if I naturally associated           certain “good” or “bad” connoted words with different races along with female and           male names with career and/or gender. Once a snap judgment was made while sorting the terms, there was no taking it back, which reflects how biases are natural and      impulsive in people’s everyday judgments.

Here, the IAT’s stated goal in designing a test that attends to the hidden or implicit is explicitly treated in the student’s contemplations about the effect of completing the IAT quickly. In other words, it is the student’s perception that if you answer the IAT questions quickly, your results will reveal accurate hidden bias.

The second comment also attends to IAT results as accurate based on the test’s design. Although in analyzing those results, the comment indicates that historically negative depictions of African Americans are to blame for personal bias assessed by the IAT, as follows:

            While reflecting, I thought about how much easier it was for me when the African            group was paired with the negatively connoted words and wonder why that was.       Flashes of news clips, history lessons popped into my head, all of them shining     African Americans in a negative context: slavery, riots, robbery, poor, dirty. It then made sense to me that since I’ve repeatedly seen and heard African Americans in a             negative light, many more times than in a positive, that it was easier for my brain to          link those together. (Student, Fall 2021)

While it is perhaps unsurprising that undesired results are linked by the test-taker to historical fact and societal influence outside of the examinee’s control, one is left to wonder what the takeaway from the activity is for such a test-taker. One logical outcome is that, similar to students who rejected their IAT results, this student might dismiss results by dismissing personal ownership. Similar to the equivalent of an intellectual shrug from students who thought the test reinforced their understanding that everyone has hidden bias (hence, potentially normalizing and dismissing bias), this student’s assertions that their bias is probably a mirror image of bias writ large seems unactionable.  

Discussion and Conclusion

The analysis of the compiled comments seem to support the use of the IAT, in general. However, some important caveats and considerations are raised in our closer look at some representative comments that form the basis for our concerns with the ethical implications of using the IAT in the WGSS classroom setting. Some of these considerations have to do with (1) student perception that the IAT makes assumptions about the identity of the test-taker in presenting unqualified statements about preferences and life experiences (e.g., sexual activity, identity, and gender) that is sometimes reflected in the IAT’s binary question format; (2) that, whether negatively or positively assessing the IAT, students often qualified and sometimes discounted their own IAT results or the implications of taking the IAT writ large. Although the number of qualifications far exceeded outright rejection both in comments about the IAT’s content and the logistics of taking the test, (3) the potentially biasing factor of the participation in the IAT presented as mandatory in the initial iteration of the assignment must be acknowledged as significant. Also, as previously mentioned, the alteration in subsequent semesters of presenting the IAT as one option of two that also included reading the short story is problematic because of the IAT’s inclusion in the textbook and potential to be perceived as more straightforward at the outset (4).

Assessments of the short story were generally positive. A common theme I noticed is that students who felt positively about both taking the IAT and reading the short story observed that taking the IAT helped them be more aware of their biases while reading the story. This application of the IAT to the short story taken has some implications. First, it seems that many students appreciated the dual approach to contemplating implicit bias—having two different sources from two different mediums. Second, the compiled comments fail to demonstrate (at least in so far as the research completed to date indicates) that reading the short story alone is perceived by students as adequate in addressing the topic of implicit bias. Therefore, a complementary exercise that is not the IAT seems the best compromise. Along with further analysis of student perception of the short story, such exercises need to be investigated.

One possible alternative to the IAT includes videos, such as the 1997 Prodigy music video banned from MTV. This film depicts a very wild night from the point of view of a person you cannot see. The twist ending finalizes the narrative (i.e., the viewer eventually fully sees the character), allowing for the viewer to draw conclusions about their potential misinterpretation of the character’s identity and reflect on those assumptions. Such videos could be accompanied by questions for reflection (e.g., Did you anticipate the ending? Why would this be a surprising ending?). In this way, concerns about student bias against the short story’s literariness (e.g., time it takes to read, its “unscientific” nature) and student rejection of the never finalized ambiguous racial identities of the characters in the short story could be mitigated by a more accessible medium. Other resources, like short documentary films on bias from the New York Times or podcasts like Hidden Brain’s episode on the psychology of surprise endings could make good alternatives/complements as well, especially considering the graphic nature of the aforementioned music video.

Whatever the proposed alternative to assigning the IAT, the aforementioned caveats and considerations (1 – 4) alone are not enough to unequivocally make a prohibitive case against the IAT’s use. However, comments that echoed Valaree Ford‘s original assessment of the IAT when combined with the above considerations make a compelling case for seeking out a replacement activity in the service of feminist pedagogy. 

[1] This assertion is repeated in spirit at least once in a subsequent semester by another student.