Search Articles

View query in Help articles search

Search Results (1 to 2 of 2 Results)

Download search results: CSV END BibTex RIS


Seeing the “Big” Picture: Big Data Methods for Exploring Relationships Between Usage, Language, and Outcome in Internet Intervention Data

Seeing the “Big” Picture: Big Data Methods for Exploring Relationships Between Usage, Language, and Outcome in Internet Intervention Data

To account for topic correlation, we ran exploratory factor analysis on topic scores using varimax rotation. Based on the scree plot, we decided to use a 50-factor solution that exhibited a close fit (root mean square error of approximation of 0.01807) explaining 32.5% of overall variation. The resulting factors may be viewed as patterns of user behavior. In summary, the LDA topics clustered together aspects of messages that likely had some similarity.

Jordan May Carpenter, Patrick Crutchley, Ran D Zilca, H Andrew Schwartz, Laura K Smith, Angela M Cobb, Acacia C Parks

J Med Internet Res 2016;18(8):e241