Survey says… (Too often, nothing)
“Hope Rising” by Computer World editor Don Tennant (Computer World Magazine, November 2008). Don writes a great column that’s always challenging and often venturing outside the scope of traditional IT publication. This was a good topic but unfortunately he missed the mark in some respects. Here’s what AU had to share with Don:
Hello again Don,
With regards to the Hope Rising column, and the issue of race… Do I think there’s racism in America and specifically corporate America? Without out a doubt I do. Do I believe your column helped clear things up? Nope. First thing, you used the S word – survey. In short, unless there’s some sort of verification I don’t see how such result can be trusted to make any sort of assessment of reality. At the very least any assumptions (and we all know about assume) should be prefaced with, “Of those who responded to this unscientific survey. The input to this survey is subjective and is not audited for survey taker accuracy, etc.”
There are just tons of possibilities:
– For all we know whites tend to exaggerate more than Afro-Americans about their pay.
– If Afro-Americans really are more dissatisfied then I’m willing to bet that those who are unhappy are going to be pretty well aware of the figure that’s making them unhappy. Note: I’d trust someone’s self assessment of satisfied vs. dissatisfied (which is subjective) far more than someone’s recall of their salary (objective).
– Is average really the best way to look at this?
– Pay is also related to geographic location and that certainly could be related to racial concentration. For example, Atlanta verse the more expensive and thus higher pay of Dan Diego.
Well, I think you get the point. I often enjoy your thoughts because they provoke more thoughts. However in this case I think the assumptions and generalizations have been counter productive. If we’re to solve this problem – or any problem – then hard facts are going to be more effective then over-generalizations of surveys.
Keep’em comin’, please!