Read more
Many research projects involve analyzing sets of texts from the social web or elsewhere to get insights into issues, opinions, interests, news discussions, or communication styles. For example, many studies have investigated reactions to Covid-19 social distancing restrictions, conspiracy theories, and anti-vaccine sentiment on social media. This book describes word association thematic analysis, a mixed methods strategy to identify themes within a collection of social web or other texts. It identifies these themes in the differences between subsets of the texts, including female vs. male vs. nonbinary, older vs. newer, country A vs. country B, positive vs. negative sentiment, high scoring vs. low scoring, or subtopic A vs. subtopic B. It can also be used to identify the differences between a topic-focused collection of texts and a reference collection. The method starts by automatically finding words that are statistically significantly more common in one subset than another, thenidentifies the context of these words and groups them into themes. It is supported by the free Windows-based software Mozdeh for data collection or importing and for the quantitative analysis stages. This book explains the word association thematic analysis method, with examples, and gives practical advice for using it. It is primarily intended for social media researchers and students, although the method is applicable to any collection of short texts.
List of contents
Acknowledgments.- Introduction.- Data Collection with Mozdeh.- Word Association Detection: Term Identification.- Word Association Contextualization: Term Meaning and Context.- Word Association Thematic Analysis: Theme Detection.- Word Association Thematic Analysis Examples.- Comparison Between WATA and Other Methods.- Ethics.- Project Planning.- Summary.- References.- Author Biography.
About the author
Mike Thelwall, Professor of Data Science, leads the Statistical Cybermetrics Research Group at the University of Wolverhampton, UK. He has developed free software and methods for social media sentiment analysis and for systematically gathering and analyzing web and social web data. He is particularly interested in social science applications of web data. His program SentiStrength is sold commercially, given free to academics, and used in several academic social media analysis toolkits. He has co-authored hundreds of refereed journal articles and has written three books. He is an associate editor of the Journal of the Association for Information Science and Technology and sits on four other editorial boards.
