We checked potential differences by webpages, geographic part, and you can ethnicity using t-testing and you may study from variance (ANOVA) on LIWC class percentages. To your several websites, half a dozen of the a dozen t-examination was basically high regarding pursuing the kinds: first-people only 1 [t(3998) = ?5.61, p Second Dining table 2 for mode, basic deviations, and you may contrasts anywhere between cultural organizations). Contrasts found extreme differences between Light and all of other ethnic communities during the five of your half dozen high ANOVAs. For this reason, kuuma Mumbaian naiset we integrated ethnicity just like the a great dummy-coded covariate in analyses (0 = Light, 1 = Every other cultural organizations).
Of 12 ANOVA screening connected with geographical region, merely two was indeed high (relatives and you may self-confident feelings). Just like the variations were not commercially meaningful, i failed to thought geographical region during the subsequent analyses.
Efficiency
Regularity away from word explore goes without saying in the detailed statistics (come across Table step one) and you can via keyword-clouds. The phrase-affect techniques depicts by far the most widely used words across the whole test plus in each of the age groups. The phrase-affect program automatically excludes specific conditions, as well as stuff (a beneficial, and you may, the) and you may prepositions (so you can, with, on). The remainder stuff conditions try scaled sizes according to its regularity, starting an user-friendly portrait really prevalent articles words around the the newest test ( Wordle, 2014).
Contour 1 suggests the latest 20 most typical blogs terminology used in the entire sample. As can get noticed, the absolute most commonly used words had been like (lookin inside the 67% regarding users), like (looking when you look at the 62% out of pages), looking (lookin within the 55% out-of profiles), and individuals (appearing into the 50% out-of profiles). Ergo, the most used terminology have been similar across the a long time.
Profile 2 shows another 31 typical content terms and conditions for the the youngest and you will earliest age range. By detatching the initial 20 well-known stuff terms and conditions across the test, i show heterogeneity regarding the relationships users. In the next 29 terms and conditions towards the youngest generation, high percentage terms and conditions included score (36% off pages regarding youngest age group), wade (33% away from users in the youngest age bracket), and you may work (28% away from users regarding youngest age group). Having said that, the earliest age bracket got large percentages off terms and conditions eg take a trip (31% out of pages about earliest generation), higher (24% of profiles about earliest generation), and you may relationships (19% out-of pages on the oldest generation).
Second 30 typical terms in the youngest and you can oldest decades teams (after deducting the latest 20 most typical terminology of Figure step 1).
Theory Evaluation old Variations in Words when you look at the Relationship Users
To check hypotheses, new portion of terms and conditions from the dating character that suit each LIWC classification supported once the created variables in regressions. I checked years and you will gender as the independent parameters along with adjusting to have website and you may ethnicity.
Theory step one: Earlier decades will be of a higher portion of conditions throughout the pursuing the groups: first-people plural pronouns, family unit members, friends, health, and you can positive feelings.
Results mostly offered Hypothesis step one (get a hold of Desk dos). Five of your own four regressions shown a critical head effect getting decades, in a way that as the age the latest character blogger improved, the latest part of terminology from the class enhanced on the following the categories: first-individual plural, loved ones, health, and you will confident emotion. I located no high years feeling towards the proportion out-of terminology regarding the household members category.
a great Gender: 0 (female) and you can 1 (male). b Site: The 2 websites have been dictomously coded due to the fact step 1 and 0. c Ethnicity: 0 (White) and you will 1 (Ethnic otherwise racial minority).
good Gender: 0 (female) and you will step one (male). b Site: Both other sites was indeed dictomously coded while the step 1 and you will 0. c Ethnicity: 0 (White) and you will step 1 (Ethnic otherwise racial fraction).
Was this helpful?
0 / 0