Thread by @lpsatchell

New paper!
Are you missing spending time with your friends? Perhaps during this time of self-isolation you need to be with other people?
You could be an offline friend addict! After all 69% of people are addicted to their friends... or are they?

Thread:
psyarxiv.com/7x85m

There are now a lot of social media addiction scales cropping up in the literature. These scales are used in a number of papers to report a % of people addicted to social media (up to 34% of some samples, see Andreasson, 2015).
This number is surprising as a pathology prevalence.

We looked into how the social media addiction scales are developed.
They usually adapt the wording of an existing scale like how the Social Media Craving Scale was adapted from the Penn Alcohol Craving Scale (Savci & Griffiths, 2019). They just change 'alcohol' for 'social media'

Then they validate this by showing correlations with existing trait measures, risk measures and amount of time spend on social media.
In these studies (and we encourage people to read their method sections in detail) addicts are classified using a 'polytetic scoring' system

This is the frequently cited guidance provided in Andreassen et al. (2012, p. 512):“scoring 3 or above [out of 5] on at least four of the six items”
That is, if people 'neither agree nor disagree' on just over half of the items, they are classified as an addict. (Not norm-based)

The second issue is that the items in these measures often are framed in a way that sound like something people would 'neither agree nor disagree' about with social behaviour in general ("Thought about how you could free more time to spend on Facebook?" / with friends)

So maybe, if these measures are tapping into social addiction, we could also discover social addiction in offline contexts? Afterall everyday social interaction involves seeking information from others... maybe people are Offline Friend Addicts too?

So with a brilliant team of people (see thread end) we developed the O-FAQ (pronounced 'Oh-fahq') in a preregistered study that followed the exact scale development used in social addiction scales. We turned 'facebook'/social media to 'spend time with friends' (table 1)

We collected N=807 participants' responses to this and validation measures (in the style of previous papers) on the Big Five, risk-taking and amount of time people spend with friends.
We collected a 28-day later test-retest in N=313.

EFA suggested O-FAQ three-factors (rather than our preferred one-factor solution), which we named Social Rumination, Life Disruption and Affective Reactions.
These were 'validated' against Big Five and Risk-taking domains

We also got really strong test-retest reliability (all r=> .72).
And this correlated with self-reported number of hours spent with friends (much like how social media addiction scales report self-reported screentime...)

Finally, using the polythetic criteria of scoring 'neither agree nor disagree' or higher to 19/37 of our questions, we could classify 69% of people as addicts.

But come on... does this seem right? Should it be this easy to consider social information seeking a pathology?

Our discussion reflects on how we, like social media addiction research, have no psychiatric validation criteria for our pathology. Our cut-off is way too liberal and we don't consider how normal and health social information is for humans. Something to think about...

This fun (and sardonic) project was coathoured with my amazing second authors @dfido1 @CraigHarper19 @H_Shawberry and a great team of collaborators @BritDavidson @davidaellis @Dr_Narcissism @RJalil @DrAliceJones @LindaKKaye Gary Lancaster and @MelissaPavetich
:D go REDTEAM!

PS if you are concerned that you are addicted to spending time with your friends, you can take the O-FAQ here, where there are *norms* to compare your response too...
(PS and our data and R code...)
osf.io/9y2rh/

#Schmeasurement #Measurement #ValidityCrisis