SoylentNews is people
SoylentNews
from the would-you-recommend-this-car-to-a-friend? dept.
Lexus, Toyota and Buick are the most reliable brands in Consumer Reports' latest survey, a reward for their conservative approach to new technology.
It's the fourth straight year that Lexus came in first and Toyota came in second. Two of their hybrids—the Toyota Prius and the Lexus CT 200H—were named the most reliable vehicles. But Buick—General Motors Co.'s near-luxury marque—is the first domestic brand to crack the top three since the magazine began tracking vehicle reliability in the early 1980s.
Audi and Kia rounded out the top five brands.
Dodge, Chrysler, Fiat and Ram—all owned by Fiat Chrysler Automobiles—were the worst performers. Electric car maker Tesla Motors also fared poorly.
The magazine released its annual reliability survey Monday. It's closely watched by the industry, since many buyers look to the magazine for recommendations.
http://phys.org/news/2016-10-lexus-toyota-buick-auto-reliability.html
Top Vehicles in Consumer Reports' Reliability Survey
[Details]: Consumer Reports News
(Score: 2) by RamiK on Thursday October 27 2016, @04:20AM
The ignorance of the consumers is a given. Even an auto-mechanics has significant observation bias after all. So, you can go around expecting people who maybe owned 5 automobiles from 2-4 makes to rate a market filled with different price tiers, part prices, service costs, models, engine sizes and etc...
Saying that, the format can still produce good data. Say, by following something like this format:
Rate the 5 most recent cars you've owned in descending order by date of purchase:
1. Reliability: |1| |2| |3| |4| |5| Make:______ Model:______ Year:_____
2. Reliability: |1| |2| |3| |4| |5| Make:______ Model:______ Year:_____
3. Reliability: |1| |2| |3| |4| |5| Make:______ Model:______ Year:_____
4. Reliability: |1| |2| |3| |4| |5| Make:______ Model:______ Year:_____
5. Reliability: |1| |2| |3| |4| |5| Make:______ Model:______ Year:_____
Ideally, you'd use a web form and drop-down boxes so the numbers would crunch automatically.
Following this format, you'd eliminate multiple biases and have a good built-in criteria to evaluate the quality of the feedback (diversity of past ownership and years of ownership). You'd be able to say things like "People who owned a GM Truck for 3-7 years rate their reliability this or that". Or, "Toyota is often rate most reliable by people owning models past X year and that owned the car for Y years.".
Of course, making a determination of what's "Most reliable" will be a bit tricky. But still, I'd rather debate the interpretation of the data then the method of it's collection and it's very quality.
compiling...