Does Network analysis relate to co-occurrence?

Apr 21, 2010 at 7:12 PM

I apologize for being a bit of a luddite. I'm trying to figure out if NodeXL, and its basis in Network analysis, would be more useful than the co-occurrence modeling I am currently doing. I certainly don't expect a full lesson on network analysis but maybe there's a quick way to tell me if I'm headed in the right direction.

I conduct qualitative research with sample sizes of about 10-50. Each participant goes through an activity in which they choose stimuli and then group those stimuli. Imagine selecting possible features (e.g., ringer mute) for a cell phone and then grouping those features according to use (e.g., answering the phone). I use a frequency and co-occurrence application to tell me which stimuli are most often picked with which others (e.g., ringer mute and ringer volume were often chosen together, more so than with display dimmer).The work can get a little more complicated in that sometimes we pre-supply the participants with the categories in which to place things (e.g., controls). Then we're curious not only if certain stimuli co-occurred but if they did so in the same group (that'd be like the voting example, but if the votes were broken down into Environmental, Financial, Procedural, etc.

That seems to be very much in line with the use of NodeXL, especially as I think about the senate voting example.

FYI, Tony previously replied to a question of mine, but I'm trying to think a bit broader at this point - probably this is the question I should have asked first.

Generally, I'm curious if I've found a more powerful tool with which I can do stronger analysis or if I'm going to end up mis-using a jackhammer when I'm just trying to hit a nail? Any thoughts would be appreciated.

Apr 22, 2010 at 12:34 AM

I am not sure the NodeXL is the right tool for your problem, yet. 

This may require a conversation to work through.

Please visit my blog at to give me a ring (click the "Call me" link) or send me an email (marc {at} to arrange a time to chat.