Online Program

Comparing Daily and Weekly Diary Data Analyses of Relationship Conflict Reports in a Nationally Representative Young Adult Sample (309992)

Keywords: Inter-personal violence, digital diary data, nationally-representative sample, young adult population, Cox mixed effects model, generalized linear mixed effects model, GLM, respondent burden

The Interpersonal Conflict and Resolution (iCOR) study examined daily reports of conflicts across relationships in a nationally-representative sample of 589 young adults, aged 18 to 32, over a six-week period. For each daily survey, respondent attitudes and experiences of conflict were collected by sending participants questionnaires about the past day. These digital diary data were used to investigate several research questions regarding the nature, incidence, and coincidence of different forms. We examined the loss of information in moving from a granular model, using the daily survey information, to a less costly and more coarsely categorized version (weekly or biweekly data collection). To the extent that the conclusions derived from the aggregated data appear to represent experiences of interest reliably, the results would suggest the value and opportunity of reducing the respondent burden by surveying them less regularly. We concluded that while the use of coarsely measured behavior data does not appear to impact effect direction, we lose some accuracy in drawing statistical inferences with more coarse data, compared to daily measures.