The Meaning Extraction Method: An Approach to Evaluate Content Patterns from Large-Scale Language Data

Abstract

Qualitative content analyses often rely on a top-down approach to understand themes in a collection of texts. A codebook prescribes how humans should judge if a text fits a theme based on rules and judgment criteria. Qualitative approaches are challenging because they require many resources (e.g., coders, training, rounds of coding), can be affected by researcher or coder bias, may miss meaningful patterns that deviate from the codebook, and often use a subsample of the data. A complementary, bottom-up approach - the Meaning Extraction Method - has been popular in social psychology but rarely applied to communication research. This paper outlines the value of the Meaning Extraction Method, concluding with a guide to conduct analyses of content and themes from massive and complete datasets, quantitatively. The Meaning Extraction Method is performed on a public and published archive of pet adoption profiles to demonstrate the approach. Considerations for communication research are offered.