I examined potential variations by the web site, geographical region, and ethnicity playing with t-testing and study regarding variance (ANOVA) to your LIWC group rates. On the two other sites, half dozen of one’s a dozen t-evaluating was basically tall from the following classes: first-person only one [t(3998) = ?5.61, p Secondary Dining table dos to possess mode, practical deviations, and contrasts ranging from ethnic CupiDates kirjautuminen groups). Contrasts revealed significant differences between White as well as most other ethnic groups when you look at the four of half dozen extreme ANOVAs. For this reason, i included ethnicity due to the fact an excellent dummy-coded covariate in analyses (0 = White, step one = Every other ethnic groups).

Of your own twelve ANOVA screening pertaining to geographic region, only a few were significant (family members and self-confident feeling). As variations just weren’t officially important, i didn’t believe geographical area inside next analyses.

Efficiency

Regularity out of keyword fool around with is obvious inside descriptive analytics (come across Dining table step one) and thru keyword-clouds. The word-affect strategy illustrates probably the most commonly used terms over the whole attempt as well as in all the a long time. The expression-cloud program instantly excludes particular words, and additionally content (an excellent, and you will, the) and prepositions (so you’re able to, which have, on). The rest posts words try scaled in dimensions relative to the regularity, starting an user-friendly portrait of the very commonplace posts terms and conditions across the this new take to ( Wordle, 2014).

Figure step 1 shows this new 20 most typical posts conditions utilized in the entire shot.