Read on Twitter
1/ **new paper** Sharing the load: Contagion and tolerance of mood in adolescent social networks
https://psycnet.apa.org/record/2020-99621-001
with @block_per
https://psycnet.apa.org/record/2020-99621-001
with @block_per
2/ is mood contagious from one day to the next in real world social groups? will hanging out with a happy friend make me happy? is there an emotional cost to supporting a friend in need?
3/ we gathered daily social interaction (n=1775) and mood (n=4724) data from N=79 adolescent classical musicians during 2 residential performance tours. Each day, ppts reported the top 2-6 peers they hung out with that day and rated how much they experienced each of 12 mood words
4/ we tested whether individuals who spent time together experienced similar mood type/valence, and how this arose over time, i.e. was it due to social influence (mood contagion at the network level) or social selection (e.g. mood homophily)?
5/ descriptive analysis showed that connected (vs unconnected) adolescents were indeed more similar in mood. Clusters of connected adolescents changed congruently in their mood over time. Next, we used stochastic actor oriented models (SAOM) to find out how this similarity arose
6/ inferential analysis using RSiena showed a significant MOOD CONTAGION effect: Connected individuals became reciprocally more similar in mood over time. i.e. FIRST participants formed a tie, THEN their mood becomes more similar.
7/ So, to answer the questions from the start: will hanging out with a happ(ier) friend make me (more) happy? YES (in this dataset). Is there an emotional cost to supporting a friend in need? YES (in this dataset) - spending time with a sad friend increases my sad mood
8/ The mood contagion effect was large (a ppt was 1.26x more likely to experience a particular mood state for each interaction partner that reported this mood) and greater for negative than positive mood.
9/ the finding that neg mood was MORE contagious than pos mood (in this dataset) is interesting and contra to prior findings that neg mood leads to social withdrawal. Possibly the particular sample (age range, group) contributed; warrants further work (any support?)
10/ Interestingly, despite good statistical power, we found no evidence for social selection based on mood. Ppts did not choose their interaction partners based on similar or desirable mood (mood homophily/popularity) and did not become more/less sociable based on mood
11/ We call this 'mood tolerance'. Adolescents did not abandon a friend after the friend had a bad mood day. Interesting to see if this finding generalises to other social groups and ages - would adults, or less integrated groups, withdraw to protect their own mood?
12/ finally we computed 'relative importance' parameters to compare model parameters to each other in terms of their impact on mood. For both pos and neg mood the biggest (50%) impact was from other similar mood states and mood pos feedback. However pos and neg mood differed in
13/ the relative importance of fixed effects (i.e. daily events) vs. social influence: For pos mood, daily events were as influential as contagion, whereas for neg mood, contagion was much more influential. [Emotion researchers: Get your head round Fig5, you're in for a treat]
![13/ the relative importance of fixed effects (i.e. daily events) vs. social influence: For pos mood, daily events were as influential as contagion, whereas for neg mood, contagion was much more influential. [Emotion researchers: Get your head round Fig5, you're in for a treat]](https://pbs.twimg.com/media/Eq5rYFbXUAArjNx.png "13/ the relative importance of fixed effects (i.e. daily events) vs. social influence: For pos mood, daily events were as influential as contagion, whereas for neg mood, contagion was much more influential. [Emotion researchers: Get your head round Fig5, you're in for a treat]")
14/ TLDR summary (much needed I know; sorry): Evidence for mood contagion in adolescent social networks interactive face to face. Negative mood more contagious than positive mood.
15/ So many outstanding questions, esp in COVID times. Would effects replicate in other adolescent groups (e.g. school)? what happens when interaction is not face-to-face, i.e. how does mood contagion differ when people are restricted to mediated interactions?
15/ another pertinent and timely question: If everyone is struggling, is it 'risky' (in terms of mood) to connect, or are there still benefits? Under what conditions are there benefits, e.g. group size, cohesiveness, individual vs. collective mood reserves?
16/ main limitation = sample representativeness (age, demogs, setting). however, for reasons we outline in the paper, we expect this to limit generalisability in terms of magnitude of effects but not presence vs. absence of mood contagion
17/ what we lose in representativeness we gain in control - sampling each day from bounded, real world social networks undergoing highly similar daily experiences robustly justifies causal conclusions re. interrelationship between mood and social connections
