Detailed Analytics and you may Design regarding Widely used Terminology
I checked out prospective distinctions by website, geographic part, and you may ethnicity having fun with t-screening and you can investigation away from variance (ANOVA) with the LIWC class percent. Toward two websites, six of your own several t-evaluating was in fact tall about after the groups: first-individual one [t(3998) = ?5.61, p Additional Table 2 for function, fundamental deviations, and you will contrasts between ethnic teams). Contrasts revealed extreme differences when considering White and all of most other cultural organizations inside the five of your six high ANOVAs. For this reason, i included ethnicity because the a beneficial dummy-coded covariate from inside the analyses (0 = White, step 1 = Every other ethnic groups).
Of your a dozen ANOVA tests connected with geographic region, just a few was in fact extreme (nearest and dearest and self-confident feeling). While the variations weren’t commercially meaningful, i didn’t imagine geographical region when you look at the next analyses.
Performance
Regularity of phrase fool around with is obvious from inside the detailed analytics (come across Dining table step 1) and you can via keyword-clouds. The phrase-affect method portrays many commonly used terminology over the whole try plus each one of the a long time.