How team sports participation affects mental health amongst university students (language: EN, study location: United Kingdom)
Project links
Skills
About this project
In this study, I collaborated closely with a research partner to explore how perceived authorship affects the appreciation of algorithmically generated poetry, specifically in sonnet and haiku forms. To begin, I prepared the dataset, ensuring it was properly organized and formatted for hypothesis testing. This involved initial descriptive statistical analysis to provide an overview of participant demographics and establish foundational data insights.
For the core analysis, I employed a Two-Way Mixed ANOVA to examine appreciation levels across three conditions of perceived authorship: Poet, Machine Learning Algorithm, and No Authorship. The analysis assessed participant appreciation for sonnets and haikus, both within each authorship condition and across groups. This methodological approach allowed for a nuanced understanding of how, or if, authorship perception influences literary appreciation in algorithmically generated contexts.
Ultimately, the results showed that perceived authorship had no statistically significant effect on poem appreciation. Poems attributed to human authors were not rated higher than those attributed to algorithms or those with no specified authorship, nor did poem type (sonnet or haiku) impact appreciation levels. I reported and interpreted the findings in APA style, providing detailed statistical tables and figures to support the results.