Browsing by Subject "puheen tuottotapa"

Goals: The aim of this study was to examine a possible correlation between speaker’s voice quality and his/her persuasiveness in speech performance. The aim was also to detect what speaker characteristics would correlate with different voice qualities in Finnish speech culture. There has not been a lot of research done about Finnish speech culture from voice quality perspective, so any research that can be linked to speech culture offers more information for future studies. Voice quality has been studied for a long time and it is known that it has impact to the speech and can determine how it will be received by listeners. This is a good feature to examine and it helps to measure speaker’s convincingness. Phonetic methods give a new view of inspecting the phenomenon how voice quality can affect convincing speaker’s detection. Methods and material: This qualitative study consists of spontaneous speech samples by 12 different Finnish speakers and of 30 different auditory detection reviews. The data is examined using auditory and acoustic methods. The auditory analysis is based on the results from an auditory detection reviews made by 30 Finnish speaking listeners. This data is collected with e-form. The listeners were detecting and assessing voice quality traits, educational background, speakers’ characteristics and convincingness. Convincingness is measured using Likert scale. The voice quality in those spontaneous speeches is also measured using acoustical methods. Examination is done with glottal inverse filtering program called Aalto Aparat. The Voice quality parameter was normalized amplitude quotient (NAQ) and it was compared to the Likert scale evaluation results of convincing speaker. The correlation between these two parameters are calculated with Kruskal-Wallis test. All the results have been presented statistically. Results and conclusions: Voice qualities modal, creaky, hoarse and breathiness are more easily detected in Finnish speakers. The most detected speaker characteristics were genuine, monotonous, enthusiastic, trustworthy, analytical, credible and familiar. These characteristics can be found as good values in Finnish speech culture. The results show that tense voice production was associated with modal voice and with convincing speakers. Speakers who were evaluated to be convincing were also detected to have creak in their voice. Although the acoustical parameter NAQ did not show any correlation results to Likert scale assessments of convincing speaker based on this data, the acoustical results gave confirmation that voice quality is an independent voice characteristic. The auditory analysis gave important information on the interaction of speaker’s traits and voice quality. Based on this study and data, I could not find defining correlation between voice quality and a convincing speaker, but the results give example about how different voice qualities are detected in a speech that aims to be convincing in Finnish speech culture.