3 ms·
It shouldn't be a problem if there's a bias towards certain car models. It just means that some models will get a larger sample size and other models will get a
by rcthompson 10y ago
It shouldn't be a problem if there's a bias towards certain car models. It just means that some models will get a larger sample size and other models will get a smaller sample size. They could easily account for this by sticking error bars on all their graphs, and ideally, reporting their sample sizes. The real problem is the other aspect you mentioned, which is that for any given model, owners of that model might be more likely to consult a 3rd-party mechanic for some issues than others.
- Nitramp 10y agoYou could imagine that owners of a more expensive car are more likely to use the original dealer for most repairs, and only seek a third party shop for very expensive repairs. Which, depending on methodology, might skew Your Mechanic's numbers. But that's only speculation given they don't reveal their numbers, methods, or sample sizes.
- wrsh07 10y agoBut it is an issue if there's bias towards certain (more expensive) repairs for those models.
- rcthompson 10y agoThat's what I said.
- wrsh07 10y agoOops, agreed
- hueving 10y agoA specific model being more expensive to repair is something I would want to know about.
- rcthompson 10y agoNo, that's differences between models. The problem comes when people go to the dealer for cheap repairs and only go to a 3rd party for more expensive stuff. Then taking the average of repair costs from a 3rd party shop would overestimate the true average.