Wednesday, July 19, 2006

A Survey

Last month, in a fit of hubris, I went bought a new car. Brand new, 2006 Chevy Cobalt. Wonderful car, love it. Runs well, gets decent gas milage, has heated leather seats, power tinted windows-- in short a bit loaded. I'm not bragging, really, I'm not... I got my first monthly bill for it the other day and suddenly a paid off 2000 two door, White Cavalier suddenly looks a wonderful, but the piper must be paid and the bed that hath been made must be laid in. So, I'll reconoiter my finances (again) and commence to pay the sum of approximately 324 dollars a month for the next 70 months of my life.

Now, a few weeks ago I got a survey from Chevy asking me to rate the dealership. And I have been sitting on the survey because I just wasn't very happy with the treatment I got. I almost walked out three times. By the time I was done I was exhausted. I'm not going to get into it, but lets just say it was 6 hours of dickering....

I have the survey filled out, but I'm not sure I'm going to send it. Its not a good report card for them. The survey basically has these choices: completely satisfied, somewhat satisfied, satisfied, not at all satisfied... talk about splitting hairs. I would of rather a 1-10 scale or something. My answers range somewhere between Satisfied and Somewhat Satisfied. The reason I haven't sent the survey back is this: I'm not happy with the choices they give me. Maybe it would be better if I wrote them a letter and told them what it was I didn't like, but I'm too lazy for that and I don't want them bugging me trying to make me like them.

So, the survey sits on my desk, in an envelope waiting to be mailed.


Kletos Sumboulos said...

If you put those options in the order in which they appear on the card then they have committed a psychometric sin. A survey like this makes sense to consolidate a lot of data and make sense of buying behavior - older people are more likely to be satisfied than younger people for example. It's also useful to evaluate employees individually (Steve has lower satisfaction than any of the other salesmen - back to training with him.)
However, the anchors (words that denote values) have to go in logical order. Completely satisfied, somewhat satisfied, satisfied, not at all satisfied - is out of order. The survey makes sense if it reads - "Completely satisfied, satisfied, somewhat satisfied and dissatisfied." Believe it or not, there is a lot of research on the use of scales - whether to use an even number of response options or odd (which gives people a 'neutral' choice, which can work against you, because some people use it as a lazy choice and others actually use it as a middle ground. There's also debate about the perfect length. 10 is too long. I use 6 or 7 depending on whether allowing a middle point makes sense. 4 can feel constraining to the respondant (like you experienced). comment is boring.

SO HOW ABOUT THAT KINKY FRIEDMAN? I'm trying to convince my family (Houston) to vote for him.

thirdworstpoetinthegalaxy said...

Mail it. Please. Mail it for those of us who are tired of being yanked around by these guys.

Buying a car shouldn't be such a miserable experience. But it almost always is.