Weighting Survey Data: Not Necessarily a Brilliant Idea

Aug 312009
 

Should one weight their survey data? Is it worth the effort? The short answer must be ‘maybe’ or ‘it depends’. A slightly longer and much more useful answer was given by Leslie Kish in his enormously helpful paper ‘Weighting: Why, when and how’. Today (well, actually I submitted the final manuscript 2.5 years ago – that’s scientific progress for you!), I have added my own two cent with a short chapter that looks at the effects and non-effects of common weighting procedures (in German). The bottom line is that if you employ the usual weighting variables (age, gender, education and maybe class or region) as controls in your regression, weighting will make next to no difference but might mess with your standard errors.

Weighting Survey Data: Not Necessarily a Brilliant Idea

Political Science Peer-Review Survey: 836 respondents and counting

Feb 122009
 

With about 100 new respondents, yet another brilliant week for the Political Science Peer-Review Survey draws to a close. While the snowball is still rolling, and while we cannot know for certain because the survey is anonymcountries Political Science Peer Review Survey: 836 respondents and countingous after all, we might soon reach a point of saturation: I have received a number of very friendly replies from people who tell me that they have already heard about the survey once (or twice) from someone else. The Netherlands in particular seem to be a hotspot of peer-review survey related activities. You could guess that much from the distribution of our respondents. While the US dominate the field (as they should), Switzerland and the Netherlands come an amazing 5th and 6th, accurately reflecting the standing of these countries as Social Science strongholds.

Technorati-Tags: , , , , , ,

Political Science Peer-Review Survey: 500+ respondents

Jan 272009
 

The title says it all: yesterday, respondents 500-506 took the Political Science Peer-Review Survey, which is obviously great. A neat detail is that so far, more than 60 current or previous editors of political science journals have taken part in the survey. Tomorrow, we will resume or email campaign (aimed at those who have published in SSCI journals over the last eight years or so) to get even more people on board.
Technorati-Tags: , , , , , , ,

Update on the Political Science Peer Review Survey

Jan 152009
 

On Monday, we started a new initiative to boost response to the Political Science Peer Review Survey. Thanks to some very industrious research students, we were able to identify about 21,000 individual authors who have published in Social Science Citation Index-covered Political Science Journals between 2000 and 2008. For about 8,000 of these, the SSCI lists their email addresses (that’s the EM field in the SSCI records), and so we started contacting them and asked them to participate in the survey. Obviously, some addresses are not longer valid because people have moved on to different places or have left academia altogether. Nonetheless, I was slightly surprised by the rather poor quality of the address data supplied by Thomson. In some cases, letters were missing whereas in other cases similar looking letters (e.g. ‘v’ and ‘y’) had been confused. This looks like either a weak OCR routine or an non-native and underpaid data typing slave has been used. Overall, we have contacted 962 people so far. About 200 of our messages have bounced, and we have 61 new responses to the survey (assuming that without the mailout, no one would have responded during these four days), which brings us to a new total of 238 responses

 Older Entries  Newer Entries