See who is already coming to #socialrecruiting summit in November!
Who's hiring, who's not, and what else is hot
I'll hand it to Careerbuilder for the day's most unusual press release.
It's the "world's only validated color-based personality test." Validated for use in hiring? I'm not so sure.
I checked out the Dewey site , which says I "seek to know about spiritual values to understand better the meaning of life and create a sense of purpose."
True, I guess. But true with many people, if not most.
I wonder what Wendell Williams thinks about all this.
| 2 comments
|
flag
Log in or register to post a reply.
I just finished exploring the Dewey System. As junk systems go I found it fun, easy and not likely to be predictive when it comes to career counseling. I think their Personality Profile was more accurate than their Career Counselor. I tried it a couple of times and intentionally manipulated the system so I got different results. In addition, our feelings about the colors may change over time too, and let's just hope our screen's colors are calibrated properly! In the long run, Dewey may be better for EHarmony than CareerBuilders. Ron
My amazement at the junk tests people sell is only exceeded by the junk tests they buy. I have never read a single research paper that showed color preferences predicted a person's job skills...and I'm pretty sure my professors would have failed any serious student who proposed one. But let's give this vendor the benefit of the doubt and do some probing:
Color Preferences Colors have a long history of being associated with emotions: red with "arousal", blue with "calm" and so forth. So let's grant them that. Using Neural Nets For Analysis Neural nets are wondrous mathematical tools. They work by iteration: Fitting, testing, fitting, and so forth until the Net finds a mathematical solution that minimizes predictive error. But that is only the beginning. The hard proof of a Net's strength occurs AFTER the Net is trained. That is, will a trained Net do as well with data it has NEVER SEEN BEFORE? I looked for it, but it seemed like the only data the author included in the report was training data...not testing data. I don't know if this was an oversight or intentional...but the impressive correlations cited, unless tested and replicated, may be misleading junk. Validated? Validation means test scores accurately predict job performance. I saw no evidence of this anywhere. The fact that people agreed with their profile is nice-to-know, but usually has nothing to do with job ability or job performance. A "validated" color test, for example, might show that people with green preferences would score high in problem solving or similar. However, in case of the Dewey Color System, green preferences seem to be associated primarily with making money at someone elses' expense.
Sorry folks...without hard data and a professionally conducted study to support these claims, I have to default to the DOL's Guidelines. Specifically, "Under no circumstances will the general reputation of a test or other selection procedures, its author or its publisher, or casual reports of it's validity be accepted in lieu of evidence of validity."
Buyer beware!
