I tested prospective differences from the web site, geographical part, and you may ethnicity having fun with t-testing and data of difference (ANOVA) towards LIWC class percent. To the several other sites, six of your own several t-assessment was extreme regarding the following the kinds: first-people singular [t(3998) = ?5.61, p Supplementary Dining table dos to possess setting, fundamental deviations, and contrasts ranging from cultural organizations). Contrasts revealed extreme differences when considering White and all most other ethnic teams in the five of your six significant ANOVAs. Thus, we included ethnicity because a great dummy-coded covariate for the analyses (0 = White, step one = Various other cultural groups).
Of the several ANOVA screening regarding geographic area, simply two was indeed extreme (members of the family and confident feeling). While the variations weren’t officially important, we don’t think geographical area for the then analyses.
Show
Regularity from term use goes without saying inside the descriptive statistics (find Dining table 1) and you may via keyword-clouds. The expression-cloud method depicts the most widely used terms and conditions along the entire take to as well as in each of the age range. The word-cloud system instantly excludes certain terminology, as well as stuff (a beneficial, and you will, the) and you can prepositions (in order to, with, on). The rest blogs terms are scaled in dimensions according to their volume, starting an user-friendly portrait of the very most common content conditions all over the latest test ( Wordle, 2014).
Contour step one shows the brand new 20 most typical content terms and conditions utilized in the entire take to. As can rise above the crowd, the quintessential frequently used words was love (looking inside the 67% regarding profiles), like (searching for the 62% away from profiles), appearing (lookin within the 55% from profiles), and you can anyone (appearing during the fifty% away from users). Thus, the most popular words have been similar all over age groups.
Contour dos reveals the next 30 most common content terminology into the the newest youngest and you will oldest a long time. By eliminating the first 20 popular stuff terms over the decide to try, we train heterogeneity from the relationships pages. In the next 30 conditions for the youngest age group, raised percentage terms and conditions provided get (36% of profiles throughout the youngest generation), go (33% from users regarding the youngest age group), and you will works (28% out of users regarding youngest age group). On the other hand, the fresh oldest generation had large percent regarding terms such as travel (31% out-of profiles regarding oldest age group), higher (24% regarding pages in the eldest age bracket), and you can dating (19% of profiles regarding eldest generation).
2nd 31 common conditions about youngest and you can oldest decades groups (just after deducting the newest 20 most frequent terminology regarding Contour step one).
Theory Comparison of age Variations in Code inside Relationships Users
To evaluate hypotheses, the portion of terms from the matchmaking reputation that fit each LIWC classification served since established parameters during the regressions. I checked out years and you will gender due to the fact separate variables in addition to adjusting for site and you may ethnicity.
Theory 1: Older decades is in the a high portion of terms about after the kinds: first-individual plural pronouns, family, family members, wellness, and you can positive feeling.
Conclusions mostly offered Hypothesis 1 (get a hold of Table dos). Five of one’s five regressions found a life threatening chief impression to possess decades, such that as the age the latest reputation publisher improved, the brand new percentage of words throughout the class enhanced in the after the categories: first-individual plural, nearest and dearest, health, and you can confident feeling. I located no extreme age perception into proportion of terminology from the household members class.
a Gender: 0 (female) and you can 1 (male). b Webpages: The 2 websites had been dictomously coded as the step 1 and you will 0. c Ethnicity: 0 (White) and you will step one (Ethnic or racial minority).
an excellent Gender: 0 (female) and you can step 1 (male). b Webpages: The two other sites was basically dictomously coded because the step one and you may 0. c Ethnicity: 0 (White) and you will 1 (Cultural or racial minority).