Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Exact Regression for Small and Wide Data: Analyzing Transgender Participation in a Game-Based Intervention

Hecmanczuk, Violet (2023) Exact Regression for Small and Wide Data: Analyzing Transgender Participation in a Game-Based Intervention. Master's Thesis, University of Pittsburgh. (Unpublished)

Download (930kB) | Preview


Background: The public health significance of this work is to push forward new strategies for measuring transgender populations, especially study populations that face small sample size challenges. Transgender individuals face unique health disparities. Intervention methods to mitigate these disparities are still evolving and it is of interest to see what factors within transgender populations are associated with intervention fidelity.
Methods: Data was pulled from the intervention arm of a randomized controlled trial (n = 120). Participants had been instructed to download and play a computer game aimed at LGBT adolescents. Transgender and cisgender participation were compared across three outcomes: download (binary), hours played, and completion (binary). Among transgender participants (n = 62), participation in those three outcomes was modeled on social covariates. A purposeful selection process that combined field knowledge and statistical testing was used to select social variables. Linear and logistic regression models were estimated. Additionally, exact logistic regression was implemented in final models measuring completion and download within the transgender subgroup.
Results: Compared to their cisgender LGBT peers, transgender adolescents were equally likely to play the game, averaged about the same amount of hours, and were more likely to complete the game (OR: 2.95). Within the transgender population, family social support, the presence of a gender-sexuality alliance, and how a participant felt their gender mannerisms were perceived affected participation. Higher self-reported support from family was associated with playing more hours (CI: 0.10, 0.81). Students who knew their school had a gay straight alliance also tended to play more hours (CI: 0.2, 2.5) and were more likely to download (Exact CI: 0.78, 173.99).
Conclusion: Exact inference aided in reasonably estimating social covariates. Future studies facing sample size challenges should consider using this method.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Hecmanczuk, Violetviolet321@pitt.eduseh201
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairCarlson, Jjnc35@pitt.edujnc35
Committee MemberYouk, Aayouk@pitt.eduayouk
Committee MemberCoulter,
Committee MemberSidani, J.jaime.sidani@pitt.edujaime.sidani
Date: 11 May 2023
Date Type: Publication
Defense Date: 24 April 2023
Approval Date: 11 May 2023
Submission Date: 28 April 2023
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 150
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: exact inference, transgender, logistic regression, intervention fidelity, behavioral health
Date Deposited: 11 May 2023 15:33
Last Modified: 11 May 2023 15:33


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item