Posted by OddipusRex on January 22, 2003, at 14:06:56
In reply to Re: How do people decide which groups to join?, posted by M. Lee on January 22, 2003, at 6:41:51
Here's a link to another study of online groups. Efficacy was measured with Depression scales and interviews at intervals. I didn't see any mention of a control group. Also wouldn't a self selected group of volunteers be more likely to be biased in favor of believing internet groups were useful? One of the groups studies was Walkers. I did not see the others named. The abstract and an article describing the study are online. The study itself is pay for view :(
http://www.mcmanweb.com/article-222.htm
http://ajp.psychiatryonline.org/cgi/content/abstract/159/12/2062
As for how people choose PB it would seem like the emphasis on meds would attract a group of people who were already diagnosed and probably on medication. There would probably be fewer posts from younger people who are wondering what is wrong with them. There would probably be fewer people who were believers in "nurture" vs "nature" Google of course would multiply this effect by sending more and more people here from seaches for specific drugs. People looking for just general information or support have a lot more sites to choose from so probably would be fewer in number here.
I wonder if like doesn't attract like too. Although there might be some change over time. You could do your reading level comparisons over time. Do the numbers become more similiar? Are people whose reading level is the same as the majority more likely to stay? You could simply count the initial subject of the first post. Was it specific request for information or request for support, etc. Are information seekers less likely to stay than support seekers? You might also look into whether the number of initial interactions or replies influences people to stay. Do people who get more responses tend to stay longer than people who don't? Do people who reply frequently to others stay longer than people who only post about themselves? You could estimate level of personal disclosure. Do people who tell their life story feel more invested in the place and tend to stay longer? Are people who are warned to be civil more likely to leave permanently or are people who are more emotionally invested here more likely to make emotionally charged statements that cause them to be banned? Well those are things you could count based on what you've got.
I think to really measure whether it was helping people to feel better or function better you would need interviews and rating scales like in the other study. And they would need to be carried out over time. But I guess you could measure support by number of responses and information by whether the specific question was addressed.
I think the idea of people being attracted by controversy is interesting but it would be hard to compare with other groups because most of them don't encourage or allow this kind of ongoing controversy. Also most groups don't keep these masssive archives.
Bob do you get permission from the other groups you study or do you think because it's posted on the internet you don't need permission? In that case why do PBers need to sign consent forms?
I really think the administrative style is less unique than the emphasis on drugs. I don't know of any large group that has ongoing involvement in the actual discussion with the administrator. Just the opposite in fact.
One thing that interested me from your article was the idea of "super posters". You could try to define the term and follow the career of these super posters. One advantage of looking at them would be that they would probably be willing to do in depth interviews with you.
> > But let me ask you, how do you think people decide which groups to join?
> >
>
>
> I suspect much of the choice is guided by unconscious motivations. If a person grew up in an environment rich in conflict, a group where conflict exists side by side with caring conversation might be very familiar.
>
> If so, then one might see a correlation between the spread of conflict (administrative involvement could be a measure of conflict) and increase in numbers. The more difficult question would be whether increased population led to increased conflict, or whether the conflict attracted and reinforced the commitment of group members.
>
> This analysis isn't intended to demean anyone's motivations. I think we return to these environments hoping to recapitulate some troublesome experience. Maybe we are hoping something that led to conflict in our memory will work out better if we try it again elsewhere.
>
> > > > > The groups are then systematically compared in terms of activity level (overall, by topic, and by user), types of communication (questions vs. information vs. opinions vs. suggestions)...
> >
>
> For me, efficacy is a more interesting question than why people join. I wonder how readability correlates with efficacy. I also wonder how readability correlates with conflict and with administrative policies.
>
> It might also be worth an investigators time to review tools used for literary analysis. The taxonomy you suggest (questions vs. information vs. opinions vs. suggestions) might be all one could get their pencil around in one effort. But eventually, tools used in language and literature departments might contribute to a deeper understanding of how people use electronic publishing tools. Useful measures might include a count of first, second or third person voice, passive or active voice, and perhaps some regional correlation - some regions have different language patterns that might correspond with readability.
>
poster:OddipusRex
thread:8891
URL: http://www.dr-bob.org/babble/admin/20021128/msgs/8956.html