Text Analysis in Consumer Research

Abstract

This chapter provides an overview of key methods used to quantify language and researchers ability to offer insights into consumer behavior. It presents an overview of three major approaches to text analysis: Manual coding, top–down dictionary-based approaches, and bottom–up data-driven approaches. The chapter explains the basic logic behind each approach and its subdivisions, provide an introductory “how-to” for each approach, and conclude with a discussion of each approach’s pros and cons. An imputation approach proposes that language, and the individual words that it consists of, can provide a deeper and more nuanced understanding of the level of a construct. Although word counts can be conducted by human coders, the process is sufficiently easy to automate and can be conducted with spreadsheet programs or text analysis programs such as Linguistic Inquiry and Word Count. In essence, topic modeling techniques are similar to a factor analysis of one’s texts as researcher distills a large set of text data into more concise categories.