Measuring and Exploring LGBTQ+ Stigma Reduction from a Contextual Behavioural Science Perspective

Abstract

This thesis is comprised of five empirical studies which were designed to measure and explore lesbian, gay, bisexual, transgender, and queer/questioning+ (LGBTQ+) stigma reduction, through a Contextual Behavioural Science (CBS) lens. This thesis offered unique contributions to the field via its introduction of a novel stigma measure, its empirical testing of the euphemism treadmill effect, and its introduction of a novel form of perspective-taking. The thesis begins with a literature review chapter, followed by the first phase of experimentation. This involved the creation of a new psychometric scale to measure others’ stigmatising attitudes. Data were collected from a total of 429 participants, measuring attitudes toward two different populations. Exploratory factor analysis and item-reduction was undertaken using data from one sub-set of participants resulting in a unidimensional scale. Parallel forms were created using an odd-even split. The resultant factor was compared against the second sub-set of participants using a confirmatory factor analysis. The one factor, 24-item structure was confirmed and retained. The second phase of experimentation involved exploring a previously employed reduction technique (i.e., language and the euphemism treadmill effect) using both a methodologically robust approach and ecologically valid approach, across two studies. The first study used a technique familiar to CBS, match-to-sample, while the second used a more ecologically valid approach, a vignette. Both contained the same three hypotheses. Results from the first study indicated some significant correlation between pre-intervention and post-intervention scores, but no significant effect within the gay condition specifically. Results also showed that word valence was a significant moderator between pre- and post-intervention scores. Results from the second study indicated no significant change in scores from pre- to post-intervention labels, with pre-intervention scores and post-intervention scores showing a strong positive correlation. Word valence was not a significant moderator between pre- and post- scores. The third phase of experimentation involved exploring both a previously utilised perspective-taking intervention, and the creation of a novel form of perspective-taking, across two studies. For the first study, 280 participants were randomly assigned to one of six different conditions varying in requirements expected and type of perspective-taking. Each ix condition also varied in participant burden, across three levels ranging from least burdensome to most burdensome. The second study utilised 235 participants who were randomly assigned to one of three different perspective-taking conditions. Attitudes toward gay people were measured pre- and post-intervention as well as after a two-week follow-up period and compared both within and between subjects. Results from the first study showed no significant effect of change scores nor of type of intervention, as well as level of participant burden, on two of the three measures utilised. Results for the second study found no significant effect of condition on change scores. However, attitudes were shown to have significantly changed from pre-intervention to post-intervention on all three measures, and this significant change remained between pre-intervention to follow-up on two of the three measures utilised. The general discussion chapter gives an overview of the key findings that emerged from this thesis as well as a discussion of implications, limitations, and future directions as a whole and complete work.