Project Working Title: Relational Facebook Use Moderates The Link Between Ostracism And Aggression
Authors: David Koch, Judith Knausenberger, Gerald Echterhoff
Affiliation: Department of Psychology, University of Münster, Germany
A. Hypotheses
1. There is a statistical significant correlation between ostracism and aggression
- This is a replication of Poon & Teng (2017), who reported a correlation of r = .38, p < .001, so I expect a moderate correlation
2. Relational Facebook use moderates this correlation; the higher relational facebook use, the smaller the correlation between ostracism and aggression
- The moderation is expected to be linear
These predictions arise from two lines of research within the ostracism literature:
1. Ostracism leads to aggressive responses (Twenge, Baumeister, Tice, & Stucke, 2001)
2. Subtle reminders of facebook can mitigate the negative consequences of ostracism (Knausenberger, Hellmann, & Echterhoff, 2015)
B. Research questions
1. Does horizontal individualism or horizontal collectivism moderate the pattern proposed in Hypothesis 2 so that it emerges only for those high in horizontal collectivism, not for those low in horizontal collectivism, or only for those low in horizontal individualism, not for those high in horizontal individualism?
2. Does daily time spent on facebook moderate the pattern proposed in Hypothesis 2 so that it emerges only for those who spend a lot of time on facebook on a typical day but not for people who spend little time on facebook on a typical day?
C. Methods
Design: Within-subject design; all participants filled out all measures and there were no conditions
1. Independent variables
- Ostracism measured with the 10-item Workplace Ostracism Scale (WOS; Ferris, Brown, Berry, & Lian, 2008)
- Depending on employment status, we present the scale either a) in original form (for employed participants), b) modified for university context (for student participants) or c) in original form but with respect to the last work place of the participant (for unemployed, self-employed and pensioned participants)
2. Dependend variables
- Aggression measured with the Brief Aggression Questionnaire (BAQ; Webster et al., 2014)
- The order of presentation of the Workplace Ostracism Scale and the Brief Aggression Questionnaire was randomized to exploratory test for effects of order
3. Moderators
- Relational Facebook use measured with a new self-constructed 20-item scale (Relational Facebook Use Scale; RFU)
- 6 additional items of non-relational facebook use are measured but are not part of the RFU and are not used for testing hypotheses and research questions
- Horizontal individualism (3 items) and horizontal collectivism (4 items) measured with a reduced version (Sivadas, Bruvold, & Nelson, 2008) of the Horizontal And Vertical Individualism and Collectivism Scale (INDCOL; Singelis, Triandis, Bhawuk, & Gelfand, 1995)
- Time spent on Facebook on a typical day (measured in 4 levels between “10-30 minutes” to “more than 3 hours”
4. Additional items
- Age, gender, employment status, number of Facebook friends for exploratory analysis
- Items measuring seriousness of participation and distraction
- 1 attention-check item
- Facebook account yes/no for exclusion of participants without facebook accounts
Planned Sample
- We will recruit participants via the platform Prolific Academic and we will pre-select people whose first language is English to guarantee that they comprehend the material
- We couldn’t find any studies with a comparable design, so we decided to recruit 230 participants as a rule of thumb
Thereby we want to make sure to have at least 200 participants after excluding those who don’t have a facebook account
- Data collection will terminate after 230 people completed the study
Exclusion criteria
- People who don’t have a facebook account will be excluded from data analysis
- People who fail the attention-check item will be excluded
- We will exploratory test the hypotheses with several exclusion criteria (demographics, outliers, distraction); however, this will be explicitly exploratory and will be reported as such
Procedure
- Order of presentation: consent form, employment status, (WOS, BAQ: randomized), possession of facebook account, RFU and items for non-relational facebook use (not presented for participants without facebook account), horizontal individualism and collectivism, demographics, seriousness and distraction, debriefing
D. Analysis plan for Hypothesis 1 and Hypothesis 2
- The arithmetic mean for every person is computed for WOS, BAQ and RFU
- Hypothesis 1: Regression analysis with BAQ as DV and WOS as IV
- WOS as IV Hypothesis 2: Regression analysis with BAQ as DV and WOS, RFU and WOS*RFU as IV
- Only items with an item-rest correlation of r > .30 will be included in the RFU