1.5 Measuring the Gender Binary: A Novel Approach
1.5.1 Implicit Measures
1.5.1.1 The Implicit Relational Assessment Procedure
Given the number of implicit measures that have been developed in recent years, it was important to ensure one was selected which best suited the needs of the current thesis. Following a review of the literature, the Implicit Relational
Assessment Procedure (IRAP: Barnes-Holmes, Barnes-Holmes, Stewart & Boles, 2010) was considered the most appropriate for an assessment of binary biases. To give a brief overview of this paradigm, the IRAP shares many procedural properties with other mainstream implicit measures. As in the IAT and other measures, a typical IRAP examines a participant’s ability to relate two different stimulus categories (e.g., men and women) with two other stimulus categories (e.g., stereotypically masculine or feminine traits). It has a two-block structure, with participants required to relate the categories in different ways in different blocks, usually under one of two response “rules”. To continue with the example of the above categories, participants in an IRAP would relate stimuli in one block according to the rule “Respond as if men have stereotypically masculine traits and women have stereotypically feminine traits”, and the other to the rule “Respond as if women have stereotypically masculine traits and men have stereotypically feminine traits”. Blocks are made up of multiple trials, in which participants relate a specific pair of stimuli from each category. Stimuli may be presented on-screen as a pair (e.g., “Men” and “Dominant”) or in the form of a statement (e.g., “Men are
dominant”). Participants then respond using a set of relational terms (corresponding to different response keys) such as “Similar/Different” or “True/False”.
It is this trial structure and format which makes the IRAP particularly well- suited to an assessment of the binary. Most implicit measures require participants to categorise or pair stimuli together using a common key press. In the IAT, for
example, the evaluation or attribute categories are at the top right and left of the screen and participants categorise a particular stimulus (e.g., “Men”) as one or the other (e.g., masculine or feminine). As such, the measure can only produce a
relativistic, either/or assessment of bias (e.g., men are more masculine than feminine / women are more feminine than masculine: Greenwald et al., 1998; Nosek & Banaji 2001). In the IRAP, however, the target and evaluation/attribute stimuli are presented on-screen at the same time and thus can be scored to produce four different effects for each trial type (i.e., men-masculine, men-feminine, women-masculine, women- feminine). As the IRAP is scored in a similar way to other measures (i.e., the effect size of the response latency differential across task types), the individual trial-type scores would be as follows: men are/are not masculine, men are/are not feminine, women are/are not masculine, and women are/are not feminine.
The IRAP thus allows for an assessment of several theoretically interesting features of gender binarism. First, it provides an index of how both sex and gender categories are related to one another; that is, whether they constitute distinct or oppositional categories. This would be demonstrated by the magnitude and significance of the IRAP effect, as there would be no notable difference across block/task types if the categories were not meaningful opposites (see Rothermund & Wentura, 2004 for a review of the evidence of IATs using non-distinct social
categories). Second, it provides a unique way to examine the structure and make up of these sex/gender relations. Because the IRAP can be scored to produce individual trial-type effects, it is able to index the strength of both role-congruity (i.e., men- masculine and women-feminine) and role-incongruity effects (i.e., men-feminine and women-masculine). This means it can assess whether the strength and magnitude of biases are comparable across gender categories (i.e., if the significance of the
men-masculine effects are symmetrical to the women-feminine effects), and also if there is a comparable amount of resistance to forming role-incongruent relations (i.e., if the significance of the men-not-feminine effects are symmetrical with the women-not-masculine effects). Put simply, the IRAP allows for a novel, quantitative and individual-level metric of an important feature of the gender binary: that men are masculine and not feminine, and women are feminine and not masculine.
In addition to its procedural advantages, the IRAP was considered an
appropriate measure given its substantial evidence base. To date, the IRAP has been used to examine biases towards racial and ethnic minorities (Barnes-Holmes et al., 2010; Barnes-Holmes, Murphy, Barnes-Holmes & Stewart, 2010; Drake et al., 2010), people who are overweight (Nolan, Murphy & Barnes-Holmes, 2013; Jurascio et al., 2010), and gay or bisexual individuals (Cullen & Barnes-Holmes, 2008). It has also been used to examine sexual attraction (Timmins, Barnes-Holmes & Cullen, 2016) and has employed extensively in clinical settings to explore self- concept (Timko et al., 2010; Vahey et al., 2009), perspective-taking abilities (Barbero-Rubio; López-López; Luciano Eisenbeck, 2016; Kavanagh, Barnes- Holmes, Barnes-Holmes, McEnteggart & Finn, 2018), and behavioural avoidance and disgust (Nicholson et al., 2013, 2014). While it has not been applied as
frequently in the context of gender, a small number of studies have used the IRAP to assess sexist or anti-women bias (e.g., Farrell & McHugh, 2017; Scanlon,
McEnteggart, Barnes-Holmes & Barnes-Holmes, 2014). To date, however, no IRAP has examined core gender binary biases (i.e., the extent to which participants can fluently relate or juxtapose feminine and masculine attributes to both women and men).