Comments on: Groundswell: Application to Gauge Social Networks Attitudes http://davidcrow.ca/article/6698/groundswell-application-to-gauge-social-networks-attitudes Helping companies with marketing, product strategy, user experience design, and technology development. Thu, 22 Jul 2010 15:01:08 +0000 hourly 1 http://wordpress.org/?v=3.0 By: Thomas Purves http://davidcrow.ca/article/6698/groundswell-application-to-gauge-social-networks-attitudes/comment-page-1#comment-3036 Thomas Purves Wed, 02 Apr 2008 14:35:48 +0000 http://davidcrow.ca/article/6698/groundswell-application-to-gauge-social-networks-attitudes#comment-3036 When I first read that headline I thought it said Application to *Gouge* Social Networks file under the social graft/graph/gaffe etc. ;) I like this ladder model but I think there's one key factor that's missing and that's the tone of contributions. What are the quality and not just quantity of participation. How much snark/trolling/bashing is there relative to constructive contribution. It's hard to measure that empirically. Or maybe not? Maybe best you could do is some automated Godwin metric? community godwin-index = avg thread length / first accusation of somebody being a nazi etc. Or perhaps some clever bayesian algorithm to index exactly how closely any given contribution resembles the average YouTube comment. What percentage of of outbound links do not actually point only to rick-astley-related video content? (twitter would score low...) When I first read that headline I thought it said Application to *Gouge* Social Networks

file under the social graft/graph/gaffe etc. ;)

I like this ladder model but I think there’s one key factor that’s missing and that’s the tone of contributions. What are the quality and not just quantity of participation. How much snark/trolling/bashing is there relative to constructive contribution. It’s hard to measure that empirically. Or maybe not?

Maybe best you could do is some automated Godwin metric? community godwin-index = avg thread length / first accusation of somebody being a nazi etc.

Or perhaps some clever bayesian algorithm to index exactly how closely any given contribution resembles the average YouTube comment.

What percentage of of outbound links do not actually point only to rick-astley-related video content? (twitter would score low…)

]]>