‘Describe sharks in three words’: investigating the usefulness of automated statistical packages, and manual approaches, to analyse short descriptions of sharks

Context Social science research, particularly qualitative data collection, is vital in wildlife conservation science. Aims While large representative samples are valuable, scaling qualitative data collection is often challenging due to its cumbersome nature. Therefore, it is necessary to explore the effectiveness of less time-intensive approaches. Methods This study explores the use of word association (three words to describe sharks) as qualitative data, analysed using both automated statistical tools in R Studio and manual content analysis. We employed word clouds, sentiment analysis, latent Dirichlet allocation, log-odds ratio, and content analysis to analyse 1006 words from 371 participants. Key results Results indicated that word clouds, sentiment analysis, and log-odds ratio were effective, while manual content analysis was better for categorising words than latent Dirichlet allocation. Conclusion Given the low participant burden, similar items should be considered in future surveys exploring human perceptions of wildlife contexts.