I tested prospective variations by webpages, geographic part, and ethnicity playing with t-assessment and analysis regarding variance (ANOVA) with the LIWC classification proportions. Towards a couple of other sites, half dozen of your own several t-examination was significant on following the classes: first-people just one [t(3998) = ?5.61, p Second Table dos getting mode, basic deviations, and you may contrasts between ethnic organizations). Contrasts shown extreme differences between Light and all sorts of most other cultural communities into the four of your half a dozen tall ANOVAs. Thus, we integrated ethnicity since the an excellent dummy-coded covariate in analyses (0 = White, step one = Some other ethnic groups).
Of one’s 12 ANOVA tests related to geographical part, just a couple were extreme (household members and self-confident feeling). Just like the differences just weren’t technically meaningful, i did not believe geographic area in the next analyses.
Abilities
Regularity regarding keyword use is obvious inside descriptive analytics (find Desk 1) and you will via term-clouds. The term-cloud techniques portrays the absolute most commonly used terminology across the whole sample as well as in each of the age groups. The phrase-cloud program instantly excludes certain conditions, and additionally blogs (a great, and you will, the) and prepositions (in order to, with, on). The rest articles terms and conditions are scaled in size according to the regularity, performing an intuitive portrait of the most extremely commonplace posts terms and conditions all over the test ( Wordle, 2014).
Profile 1 suggests the 20 common posts terms included in the complete decide to try. As can rise above the crowd, the absolute most frequently employed terms had been love (searching for the 67% off pages), for example (appearing inside the 62% out of users), lookin (searching during the 55% regarding pages), and some one (lookin when you look at the fifty% off pages). Therefore, the most popular terminology was in fact similar all over age groups.
Profile dos shows next 30 popular content conditions inside the this new youngest and you can oldest a long time. By eliminating the first 20 common stuff conditions along side try, i teach heterogeneity in the matchmaking pages. Within the next 29 terms into the youngest generation, high level percentage terminology provided rating (36% regarding profiles regarding youngest age group), wade (33% out of profiles regarding youngest age group), and you can really works (28% out-of pages about youngest age bracket). Conversely, brand new oldest age group got higher percent out of conditions such as for instance take a trip (31% from users about earliest generation), higher (24% off profiles on the eldest age group), and you can dating (19% out-of pages on earliest age bracket).
Second 31 typical terminology throughout the youngest and you may eldest ages groups (once deducting the fresh new 20 most typical terms out-of Contour 1).
Theory Testing of age Differences in Vocabulary inside the Relationships Pages
To test hypotheses, the newest percentage of terms regarding dating character that suit for each LIWC classification supported once the built variables during the regressions. I checked out years and you can gender once the separate variables plus adjusting for site and you can ethnicity.
Hypothesis step one: More mature decades could well be of the a top portion of words in the pursuing the classes: first-individual plural pronouns, family unit members, household members, health, and you will confident feelings.
Findings largely served Theory step one (see Table 2). Five of your four regressions revealed a life threatening head perception to possess many years, in a manner that as the ages of the profile blogger improved, the newest part of conditions regarding the class enhanced from the following the categories: first-person plural, relatives, fitness, and you may confident feeling. I found no significant decades impression into proportion out-of terms regarding family class.
an effective Gender: 0 (female) and you will step 1 (male). b Webpages: Both websites was in fact dictomously bravodate legit coded since step 1 and 0. c Ethnicity: 0 (White) and step one (Ethnic otherwise racial minority).
a Gender: 0 (female) and you will step one (male). b Web site: Both websites had been dictomously coded just like the step one and you may 0. c Ethnicity: 0 (White) and you will step 1 (Cultural or racial minority).