We checked out possible distinctions by the webpages, geographic area, and you may ethnicity playing with t-screening and studies from difference (ANOVA) on LIWC group percent. On the several websites, half dozen of one’s several t-tests had been high from the pursuing the kinds: first-individual only 1 [t(3998) = ?5.61, p Second Desk dos getting mode, practical deviations, and you may contrasts anywhere between cultural teams). Contrasts found high differences between White and all of almost every other ethnic communities during the five of your own six significant ANOVAs. Hence, we provided ethnicity since the a good dummy-coded covariate from inside the analyses (0 = White, step 1 = Virtually any cultural teams).

Of one’s a dozen ANOVA examination connected with geographical region, simply several had been extreme (nearest and dearest and you can self-confident feeling). Because differences weren’t theoretically significant, i didn’t believe geographic area within the then analyses.

Results

Regularity of keyword explore is obvious inside the descriptive statistics (pick Desk step one) and you can through term-clouds. The phrase-cloud strategy depicts one particular widely used terms and conditions over the entire take to plus each of the age groups. The definition of-cloud program automatically excludes particular words, together with posts (a, and you will, the) and you will prepositions (in order to, having, on). The remainder content terms and conditions are scaled in proportions relative to its volume, carrying out an user-friendly portrait really prevalent content conditions across brand new take to ( Wordle, 2014).

Contour step 1 reveals new 20 most commonly known blogs terminology utilized in the complete take to. As well as be seen, by far the most frequently used terms was like (appearing inside the 67% regarding pages), such (lookin inside the 62% off profiles), looking (looking inside 55% out of pages), and someone (appearing inside fifty% out-of profiles). Therefore, the preferred words were equivalent round the age range.

Profile dos reveals another 29 most commonly known stuff words in the the new youngest and earliest age groups. By removing the first 20 well-known content terms along side try, we show heterogeneity on relationships profiles. Within the internet next 29 words into the youngest age group, raised percentage words incorporated get (36% away from pages regarding the youngest age group), go (33% from pages regarding the youngest age bracket), and you may functions (28% off users on the youngest age group). However, new oldest age bracket had higher proportions of terms like travel (31% of users about earliest age group), higher (24% out of users on eldest age bracket), and relationship (19% off profiles regarding oldest age bracket).

Second 29 most commonly known words about youngest and you can eldest decades groups (immediately following subtracting the fresh new 20 common conditions from Figure 1).

Theory Testing old Differences in Code in Matchmaking Users

To check on hypotheses, the newest percentage of terms and conditions regarding relationship profile that suit for each and every LIWC group supported while the founded variables into the regressions. We checked-out decades and you can gender once the independent variables plus modifying to own webpages and you may ethnicity.

Theory 1: Earlier years might be of this increased part of terms and conditions about following groups: first-individual plural pronouns, family relations, nearest and dearest, health, and you may confident feeling.

Results largely supported Hypothesis step 1 (see Desk 2). Five of your four regressions shown a significant head perception having age, in a manner that as chronilogical age of the new reputation copywriter enhanced, this new part of terms on group increased in the following the categories: first-individual plural, members of the family, health, and you will confident feeling. We discovered no significant ages effect on the proportion from terminology on family group.

an effective Gender: 0 (female) and 1 (male). b Web site: Both other sites was indeed dictomously coded since the step one and you will 0. c Ethnicity: 0 (White) and 1 (Cultural otherwise racial fraction).

an effective Gender: 0 (female) and you can step one (male). b Web site: The 2 other sites was basically dictomously coded once the 1 and you may 0. c Ethnicity: 0 (White) and you will step one (Cultural or racial minority).