Data is Beautiful

852 readers

3 users here now

Be respectful

founded 6 months ago

MODERATORS

[email protected]

'Who Makes the Most Reliable New Cars?' (2024-12) (lemmy.ca)

submitted 4 days ago* (last edited 4 days ago) by [email protected] to c/[email protected]

39 comments fedilink hide all child comments

From this article: https://www.consumerreports.org/cars/car-reliability-owner-satisfaction/who-makes-the-most-reliable-cars-a7824554938/

Related to this post: https://lemmy.ca/post/34496572

you are viewing a single comment's thread
view the rest of the comments

[–] [email protected] 11 points 3 days ago (1 children)

Not all surveys are opinion surveys like political ones are. It’s just a method of data collection.

We calculate predicted reliability ratings for almost every new car, truck, and SUV on the market using data from Consumer Reports’ annual reliability surveys, which ask members about problems they’ve had with their vehicles

This year we calculated brand-level score by first examining the weighted overall problem rate for all models within a brand for each model year. Then the brand reliability score was calculated by averaging models from 2022 to 2024, and some early 2025 data for each brand, where there was sufficient sample size. We had insufficient data to create brand rankings for Alfa Romeo, Chrysler, Dodge, Fiat, Infiniti, Jaguar, Land Rover, Lincoln, Lucid, Maserati, Mercedes-Benz, Mini, Mitsubishi, Polestar, Porsche, and Ram.

[–] ori 13 points 3 days ago* (last edited 3 days ago) (1 children)

Good point but still questionable methodology.

This basically defines most reliable as "least issues in first 3 years". Also why would you collect them as a survey of your members and not try to get a broader statistic.

[–] [email protected] 0 points 3 days ago (1 children)

What broader statistic? The next step up would be cold calls and that comes with all kinds of issues of it's own.

[–] ori 1 points 3 days ago* (last edited 3 days ago) (1 children)

There should be statistics how many cars are sold and then you could collect data from repairshops instead of users from a website. Even if a user has 3 cars, asking 1 repairshop will give you data for 100's of cars. Its just odd to do a survey on the users of your website because thats probably a small and skewed sample size.

Edit: "We had insufficient data to create brand rankings for Alfa Romeo, Chrysler, Dodge, Fiat, Infiniti, Jaguar, Land Rover, Lincoln, Lucid, Maserati, Mercedes-Benz, Mini, Mitsubishi, Polestar, Porsche, and Ram." Gives you a hint on how sufficient their data is in total.

[–] [email protected] 0 points 3 days ago (1 children)

I suppose, but they're not necessarily going to hand over data just because you asked. Maybe it's not the best possible choice, but it doesn't seem odd to me at all.

[–] ori 1 points 3 days ago

Well maybe Odd is the wrong word, too small and biased?