[Home Page] [Previous Page]
[Old Roundtable Archives] [New Roundtable]
Posted by Ian on April 13, 2000 at 02:31:47:
In Reply to: benchmarking posted by Jenn on April 13, 2000 at 02:30:55:
You may be making this more complex than it has to be.
As I've ever seen the term used, a "benchmark" is the first wave of a study intended to be redone periodically.
One should think about how to make the sample similar in the later waves, or what deviations will be acceptable. I'm currently dealing with a "benchmark" study done so that it's unacceptably expensive to duplicate the sample as specified, and I consider this a sad lack of foresight.
Also, think about the sample sizes which will be necessary to detect changes of the size you deem likely. For instance, if you are benchmarking awareness, will the next wave show as significant, a result which is in the likely range for advertising which you can afford - will it give you confidence in saying whether your ad budget has achieved the goal?
A "benchmark" study may be longer and more complex than the follow-on waves, because it may enable you to delete questions which are not heavily weighted or redundant: For instance, after running the benchmark, regression modeling may reveal that some questions are not worth asking again, at least for some years until you re-benchmark to take account of long term changes in the industry.
Subject: