Saturday, August 14, 2010

Clement's version of robust and reliable data

On the CBC Radio program The House, Tony Clement stated today that there would be no further census concessions. The reason it seems is that Clement believes that he "[is] meeting the objections of those who are concerned about robust and reliable data." I guess he thinks that a 50% response rate is robust and reliable.

The sad fact though is that it isn't. It is fair to say that experts at Statscan, who have made it their life's work to deal with statistics, would know more about what constitutes robust and reliable data than he does, and that therefore we should follow their advice of maintaining the status quo. That is if you're not talking to the experts Clement alleges exist at Statscan that believe that a voluntary long form is a perfectly acceptable substitute for the mandatory long form
Recommend this Post at Progressive Bloggers If you liked this post, please vote for my blog at Canadian Blogosphere Canadian Blogosphere

2 comments:

  1. The short form is not and would not be 'voluntary' as you have written. Everyone keeps repeating this error. The short form is in fact mandatory.

    ReplyDelete
  2. I have not written that the short form is voluntary is voluntary.

    ReplyDelete

Progressive bloggers