As some of you might remember I am working on my PhD and asked for people to fill out a survey last year to help do some research. I used several different ways to analyze this data and will try to break down some of it below. If you want to know more let me know. I am making sweeping conclusions based on the collective results and it may not represent everyone here; however, I would like to hear your thoughts on what you think about it. I apologize if the below is phrased overly "scientific" its largely copied from a draft I am working on.
In short, I used network correlation/regression models and exponential random graph models to analyze the patterns in depressive symptoms, social support, and speaking to others online about important life matters. 37 members filled out the entire survey. Members who responded to the survey were on average 24.76 years old. Members reported they spent, on average, 12.57 hours (SD=8.60) on the site per week.
To measure depression a standardized scale was used called the PHQ. The average PHQ score was 6.92 (SD=5.86). By using the suggested score of 10 as a cut-point, 13.5% (5 members) of the sample would be listed as depressed or at-risk. Members reported significantly more support from their “in-real-life” network, when compared to the support they reported from their online network.
When asked to nominate other members with whom they spoke to about important life matters, members reported an average of 6.11 other members. Using this network I was able to analyze network effects.
Depressive symptoms, online support, and IRL support were not spatially autocorrelated. This means that these features were not clumped in certain areas of the network.
Members were significantly more likely to reach out and speak to other members about important life matters if they reported more site hours, more depressive symptoms, and less IRL support.
When a model was developed to predict depressive symptoms, IRL support decreased the severity of depressive symptoms;however, members who reported more online connections also reported more depressive symptoms on average. The model also showed a significant network effect. A significant network effect means the depressive symptoms present within a someone’s network were significantly associated with the depressive symptoms of that person.
Too long; Didnt read summary:
Thank you to everyone that did this. I am eternally grateful and this is really cool stuff.
I think this site is a very unique place where people can reach out and create connections regardless of IRL.
Depression is a very real thing and people shouldn't be hesitant to reach out to formal and informal ways of helping.