I checked out possible variations from the web site, geographical region, and you can ethnicity having fun with t-evaluation and research off variance (ANOVA) to your LIWC classification proportions. Toward several other sites, half a dozen of twelve t-tests were high from the following the classes: first-individual only 1 [t(3998) = ?5.61, p Secondary Desk 2 to possess form, important deviations, and you will contrasts anywhere between ethnic communities). Contrasts found high differences between White and all almost every other ethnic groups into the four of your own six extreme ANOVAs. Therefore, i provided ethnicity as the an excellent dummy-coded covariate inside the analyses (0 = Light, step 1 = Every other ethnic organizations).
Of your own twelve ANOVA testing related to geographical area, merely a couple of have been significant (family members and you may confident feeling). As the differences weren’t commercially important, we don’t envision geographical region within the next analyses.
Results
Volume out-of keyword explore goes without saying within the descriptive analytics (see Dining table step 1) and you may thru phrase-clouds. The term-affect strategy depicts one particular popular terminology across the whole sample plus each one of the age groups. The word-cloud program immediately excludes certain conditions, and content (an effective, and you may, the) and you may prepositions (to help you, which have, on). The remainder posts conditions try scaled in size in accordance with the frequency, creating an user-friendly portrait of the very commonplace content conditions around the this new take to ( Wordle, 2014).
Profile step 1 reveals new 20 popular articles conditions included in the entire try. As well as rise above the crowd, more frequently used terms and conditions was basically love (searching for the 67% regarding users), like (searching into the 62% out of users), searching (appearing from inside the 55% away from pages), and you can some body (lookin inside the fifty% off users). Therefore, widely known terms was basically equivalent around the age range.
Figure dos shows the following 29 most frequent posts conditions when you look at the the fresh new youngest and you can eldest age range. By eliminating the original 20 prominent stuff terms along side attempt, we illustrate heterogeneity regarding dating users. Next 29 words into youngest generation, raised percentage terminology incorporated rating (36% of profiles about youngest age bracket), wade (33% out of pages on youngest generation), and you will work (28% away from users about youngest age bracket). On the other hand, the new earliest generation got large rates regarding words such as for instance traveling (31% regarding pages throughout the oldest generation), higher (24% regarding users in the eldest generation), and you may relationships (19% regarding users throughout the eldest age bracket).
Next 30 most frequent terminology in the youngest and you may eldest years teams (just after deducting brand new 20 typical terminology of Shape step 1).
Theory Review old Variations in Code for the Dating Pages
To check hypotheses, the newest portion of terms on the relationship profile that suit per LIWC class offered as the based details within the regressions. I tested ages and you can gender because independent parameters as well as adjusting to own website and you may ethnicity.
Theory step 1: Elderly years could well be for the increased portion of words from the following the classes: first-person plural pronouns, family, members of the family, health, and you will confident feelings.
Findings largely supported Hypothesis step 1 (find Desk 2). Four of five regressions shown a critical head impact getting decades, such that since the age the reputation journalist enhanced, the Guams seksikkäitä naisia brand new part of terms on the category improved in the adopting the categories: first-person plural, friends, wellness, and you may confident feelings. We located no significant decades feeling on the ratio of terms regarding the members of the family classification.
a great Gender: 0 (female) and step 1 (male). b Site: The two other sites was indeed dictomously coded while the step one and 0. c Ethnicity: 0 (White) and you can step one (Cultural otherwise racial minority).
a beneficial Gender: 0 (female) and step 1 (male). b Site: The 2 websites have been dictomously coded because step one and you may 0. c Ethnicity: 0 (White) and step one (Cultural otherwise racial minority).
Leave a Reply