Descriptive Statistics and you may Artwork out of Popular Terms and conditions

We checked prospective differences of the website, geographical area, and you may ethnicity playing with t-evaluation and you will studies of variance (ANOVA) towards LIWC classification percentages. Toward a couple of other sites, six of several t-evaluating was basically high about pursuing the kinds: first-people just one [t(3998) = ?5.61, p Supplementary Dining table dos having mode, practical deviations, and you will contrasts ranging from ethnic groups). Contrasts shown extreme differences when considering White and all sorts of most other cultural groups in the five of one’s half a dozen significant ANOVAs. Ergo, i included ethnicity while the an excellent dummy-coded covariate for the analyses (0 = Light, step 1 = Virtually any ethnic organizations).

Of your a dozen ANOVA testing connected with geographic area, simply several have been significant (loved ones and confident feelings). Since the distinctions weren’t theoretically important, we did not imagine geographic part inside the next analyses.

Abilities

Volume from phrase play with goes without saying in descriptive analytics (come across Table step 1) and via phrase-clouds. The word-affect technique illustrates many commonly used conditions along the entire shot plus all the age groups. The term-affect program instantly excludes particular words, as well as stuff (good, and you can, the) and you can prepositions (in order to, which have, on). The remaining blogs terms are scaled sizes according to their regularity, starting an user-friendly portrait of the very prevalent blogs terms around the brand new sample ( Wordle, 2014).

Figure step one reveals the brand new 20 most frequent content conditions found in the complete sample. As can be seen, one particular commonly used terms and conditions was in fact love (searching from inside the 67% out-of profiles), eg (appearing within the 62% out of profiles), lookin (searching in 55% off profiles), and some body (looking during the 50% out of profiles). Therefore, typically the most popular conditions had been equivalent across the age groups.

Contour dos shows the following 30 most frequent blogs words when you look at the the fresh youngest and you may eldest age groups. By removing the original 20 preferred posts terms and conditions across the sample, i instruct heterogeneity regarding the dating users. Next 30 words with the youngest age group, high percentage terms incorporated get (36% away from pages regarding the youngest age bracket), go (33% from users regarding youngest generation), and performs (28% of users on the youngest age bracket). In contrast, the fresh new oldest age group got high percentages regarding words eg traveling (31% from users on earliest age group), higher (24% of pages on oldest age bracket), and relationship (19% out of profiles on eldest age group).

Second 31 most frequent terms in the youngest and you can oldest age teams (shortly after deducting the newest 20 most common terms from Profile step 1).

Hypothesis Comparison of age Differences in Language in the Relationships Users

To evaluate hypotheses, the fresh new part of terms and conditions regarding the relationships character that fit for every LIWC category served once the based parameters in regressions. I tested years and gender because independent parameters including adjusting to have webpages and ethnicity.

Hypothesis step 1: Old ages could well be on the a high part of conditions throughout the following the categories: first-people plural pronouns, family relations, nearest and dearest, fitness, and you can confident feelings.

Conclusions largely supported Theory 1 (come across Table 2). Four of one’s four regressions found a significant fundamental impact to have years, in a fashion that because chronilogical age of the newest reputation writer enhanced, this new portion of terminology regarding the classification improved in the pursuing the categories: first-people plural, family, health, and you may positive emotion. I receive no extreme many years feeling toward proportion off words from the nearest and dearest classification.

a sex: 0 (female) and you may 1 (male). b Website: The 2 other sites was indeed dictomously coded while the 1 and you may 0. c Ethnicity: 0 (White) and step one (Cultural or racial minority).

a sex: 0 (female) and you will step one (male). b Webpages: Both websites was dictomously coded just like the 1 and you will 0. c Ethnicity: 0 (White) and you may 1 (Ethnic otherwise racial minority).