orgtheory.net

org survey response rates

Bleg: What tricks do we have for increasing response rates for people working in organizations? The older literature suggests that org surveys have widely varying response rates. For example, this 1999 review in Human Relations finds that top  management journals publish studies with an *average* response rate of 36%. This 2008 Human Relations article finds a response rate of 35%.  So, how can we pop up the response rate? We have the Dillman method (letters), payment and multiple contacts. How else can we reach orgs?

50+ chapters of grad skool advice goodness: From Black Power/Grad Skool Rulz 

Written by fabiorojas

March 28, 2014 at 12:01 am

Posted in fabio, mere empirics

3 Responses

Subscribe to comments with RSS.

  1. Go back in time to the 1980s before people started getting survey fatigue and you could routinely get response rates >50%.

    That’s the approach we take in Rossman, Brown, and McFly (forthcoming).

    Like

    gabrielrossman

    March 28, 2014 at 12:07 am

  2. Offer them something they want. I have experienced great success (>45%) with increasing response rates using what I call “survey apps” that a) mine social media data produced by orgs; b) survey orgs to provide context for such data; and c) return the results of the analysis back to the user as incentive to participate. In my case, I offer orgs a free audit of their social media outreach based upon a comparison to a large group of their peers- but I think the same technique could be applied using other incentives. I have a paper on how to implement the approach- unfortunately it still requires some familiarity with programming and internet savvy (Python, Google App Engine, Facebook API), etc. More details are here: http://www.findyourpeople.org (though the project is currently down while I improve the app for a second round).

    Like

    Chris Bail

    March 28, 2014 at 12:42 am

  3. By going to them.

    Like

    Hector

    March 28, 2014 at 11:24 am


Comments are closed.

%d bloggers like this: