Importance of Culturally Competent Research & Practice Tools for LGBTQ Populations

Background and Purpose:

LGBTQ individuals are often underrepresented in research, including the empirical validation of tools and assessments for both research and practice. The underlying assumptions that guide the development of such tools are often rooted in heteronormative ideologies, which may result in not only inaccurate results, but even increased marginalization for LGBTQ people. The purpose of this presentation is to provide an example of how measures that lack cultural sensitivity can lead to further stigmatization of already stigmatized groups.

Methods:

Data were collected using an empirically validated Healthy Relationship Belief Scale. OLS regression was used to determine that having an LGBTQ identity was a significant predictor of a lower score, or less healthy beliefs about romantic relationships (β = -1.52, t = -2.01, p = .046). Following this finding, independent samples t-tests were conducted for each item of the scale individually, to examine which questions were answered differently by LGBTQ and non-LGBTQ participants, and adjustments were made to the composite score based on these findings. Following this, the OLS regression model was run again with the newly adjusted Healthy Relationship Belief Scale.

Results:

Independent samples t-tests showed that LGBTQ participants scored lower than their non-LGBTQ peers on two items of the Healthy Relationship Belief Scale: 1) give each other gifts (t(260 ) = 2.28, p < .05) and 2) enjoy spending time with each other’s family (t(261) = 3.81, p < .001). Based on the well-documented history of LGBTQ individuals being ostracized or disowned by their families after coming out, this item was removed from the scale. While the authors propose theoretical explanations for why LGBTQ individuals may place less importance on giving each other gifts, we did not feel there was sufficient support in the literature at this time to remove this item, and it remained in the scale. Once the “enjoy spending time with one another’s family” item was removed, an OLS regression for the adjusted composite score showed that LGBTQ identity was no longer predictive of less healthy relationship beliefs (β = 0.73, t = 1.14, p = 0.255).

Conclusions and Implications:

This example illustrates how using instruments with populations for whom they are not designed can be stigmatizing and harmful. It highlights the need for research and practice tools that are developed and validated with people from diverse groups in mind.