(WITHDRAWN) Negative Social Exchange Among Adults: An Applied Thematic Analysis Using Egocentric Social Network Data

Background. Social relationships are powerful determinants of health and mental health (Umberson & Montez, 2010). Social exchange theory argues that people invest in relationships that benefit their well-being (Rook, 2014). However, research has overlooked the deleterious effects of negative social exchange (Offer, 2018). Negative social exchange is defined as violations of relationship norms, including failed or begrudging social support, rejection or neglect, misdirected control, unsound advice, and undermining actions (Rook & Charles, 2017). Research has focused on older adults' experiences and has not captured the relationships of early adulthood. The purpose of the current study is to evaluate whether major sources of stress can be categorized using existing forms of negative social exchange for a younger sample of adults.

Methods

An egocentric network survey design was used. Respondents between 18 and 65 years old (N = 301) were recruited using a crowdsource data collection company. To measure negative social exchange, a name generator asked respondents to list individuals who were sources of stress in their life. An open-response name interpreter provided additional information about the major source of conflict associated with each listed person. An applied thematic analysis was conducted using participants' responses. The categories of negative social exchange identified by Rook (2015) were used to develop codes. An additional " other " category was added for responses not matching existing categories. Three coders reviewed responses for N = 864 relationships reported by N =301 respondents. Interrater agreement was calculated using simple agreement and Fleiss' Kappa. A Fleiss' Kappa between .40 - .55 indicates fair interrater agreement (Stoyan et al., 2016). Responses with two or more ratings of "other" were examined. Codes were developed inductively, and responses were recoded. Coding was assessed for interrater agreement.

Results

The average age of respondents was M = 31.75, SD = 10.58. The sample was largely female (52%), reported a bachelor's degree (30%), and was single (47%). Overall, 85% of responses were categorized as some form of negative social exchange. Experiencing misdeeds and violation of relationship norms was the most frequently applied (42%), followed by misdirected control/unsound advice/undermining (30%), failed social support (21%), and "other" (22%). Rejection/neglect (4%) was the least used code. Simple agreement indicated raters completely agreed 37% of the time, while Fleiss' kappa (.43) indicated fair levels of agreement. The responses (n = 130) labeled by raters as "other" were reviewed. Seven codes were developed. General relationship problems (29%) were the most frequently applied code, followed by work-related (27%), none/other (25%), third-party source (10%), school-related (8%), children (6%), and health-related (2%). The use of inductively developed codes resulted in higher simple agreement (73%) and produced a higher Fleiss' Kappa (.71)

Discussion

The current study examined whether existing negative social exchange categories matched the relational experiences of young and middle adults. Findings indicate that many young and middle adults' responses can be categorized as existing forms of negative social exchange, but only fair agreement can be reached about the type of negative social exchange experienced.