Vocal Satisfaction in Transfeminine Speakers
Restricted (Penn State Only)
- Author:
- Romo, Nina
- Graduate Program:
- Communication Sciences and Disorders
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 05, 2022
- Committee Members:
- Nicole Etter (she/her), Professor in Charge/Director of Graduate Studies
Samantha Tornello, Outside Unit & Field Member
Navin Viswanathan, Major Field Member
Michael Robb, Chair & Dissertation Advisor
Aarthi Madhavan (she/her), Major Field Member - Keywords:
- Transgender
Voice
Satisfaction
Acoustics
Context
Identity - Abstract:
- Purpose: Vocal satisfaction is a commonly used outcome measure in gender-affirming voice and communication training. This measure may have the strongest relationship with a transgender individual’s voice-related quality of life. Vocal satisfaction may be influenced by an individual’s own personal feelings about their voice, or the social experience of using one’s voice with others. However, past research has not established conclusive relationships between the acoustics of the voice, and vocal satisfaction, in transfeminine speakers. It is possible that sophisticated acoustic analysis techniques, such as LTAS and analysis of formant spacing, may be well-suited to characterize the voice and form a basis for any existing relationships. Furthermore, past research has shown that some transfeminine speakers feel decreased vocal satisfaction when speaking with individuals they do not know well. It is possible that this may be demonstrated experimentally. Transfeminine speakers may show different levels of vocal satisfaction in some interactive contexts than others, and their vocal satisfaction may be influenced by different factors depending on context. Finally, the relationship between one’s sense of overall gender identity and voice acoustics has been explored in cisgender speakers, but not with transgender speakers. It is predicted that it may be possible to establish relationships between transfeminine speakers’ overall sense of gender identity, voice acoustics, and, ultimately, vocal satisfaction. Methods: Twenty-one transfeminine speakers, and sixteen student speaking volunteers, were recruited through the University of Utah Speech and Language Pathology Clinic. Transfeminine speaking participants were asked to provide two speech samples in a counterbalanced order. During one of the speech samples, participants were by themselves in a recording room. During the other condition, participants were in the recording room with a speaking volunteer who they had never met before. In each condition, participants provided a sample of free speech, read aloud a reading passage, and uttered four carrier sentences containing corner vowels. Immediately following the provision of each speech sample, participants were asked to rate their vocal satisfaction on a VAS. Participants also answered questions regarding their transition and vocal health, their baseline mental well-being, and their overall sense of gender identity. Following this, LTAS analyses were performed on recordings of participants’ free speech and reading passages. The following measures were obtained: mean F0, mean spectral energy, and spectral tilt. Measurements of the first two formants were taken from the corner vowels in recordings of the carrier sentences. On the basis of spacing between these formants, values representing two features – related to tongue advancement and tongue elevation – were calculated for each vowel in each condition. Results: The condition in which participants spoke – whether with an unknown individual, or in isolation – was not, on its own, predictive of vocal satisfaction. Furthermore, participants did not produce significantly different values for any acoustic variable when speaking in different conditions. However, different relationships between voice acoustics and vocal satisfaction were established for each condition. The acoustic features of mean F0 and tongue advancement of /i/ were able to significantly predict vocal satisfaction when participants spoke to an unknown individual, but not when they speak in isolation. Furthermore, among those who have had voice training, number of weeks in training were able to significantly predict vocal satisfaction when speaking to an unknown individual only. Finally, transfeminine speakers’ sense of overall gender identity, as well as its interactions with acoustic features, were not able to significantly predict vocal satisfaction. Conclusions: The results imply that transfeminine speakers may produce voices that suit their personal needs best irrespective of interactive context. However, they also imply that they judge their own vocal satisfaction using different acoustic criteria according to interactive context. These acoustic criteria broadly align with common therapeutic goals for transfeminine speakers, and may suggest that some transfeminine speakers seek out voice training in anticipation of future interactions with unknown individuals. Future research may more closely examine the role of interactive context for vocal satisfaction in transfeminine speakers, possibly by exploring different speaking environments and relationships between speakers. It may also look at the underlying source of voice acoustics by examining vocal physiology in transfeminine speakers, and determining its relationship to voice acoustics and vocal satisfaction.