4 ms·

> A lower score offers better protection. I don't understand why anyone would present the data like this, sure, 'lower is better' makes sense in benchmarks whe

by NikkiA 7y ago

> A lower score offers better protection. I don't understand why anyone would present the data like this, sure, 'lower is better' makes sense in benchmarks where you're presenting a raw timing or such, but with an arbitary unit like 'score' just invert the data so bigger -> better and let people's human nature take over.

blacksmith_tb 7y ago
Especially when more stars is better! Overall their testing methodology looks good, so it's a little disappointing that the presentation is so counterintuitive.
- psetq 7y ago
  Their primary presentation is the 0-5 star rating which is perfectly intuitive. If I'm reading the methodology correctly, the actual STAR score represents the incidence of concussion, in which lower is certainly better. The only real confusing thing I see is the "star" naming of the "STAR evaluation score" vs. using a "star rating" (0-5) and how they are actually separate things.
  - blacksmith_tb 7y ago
    I agree that the stars are clear, I wonder if there's any point in displaying the score at all, given that.
jolmg 7y ago
I think it depends no how this score is determined. If their rating methodology allows them to determine an upper bound on protection better than a lower bound, then it's better to set that bound to score zero, to not risk later needing scores with negative numbers. The formula seems to be on the following pdf on the third page. https://vtechworks.lib.vt.edu/bitstream/handle/10919/83760/Bicycle%20STAR%20Methodology.pdf?sequence=1&isAllowed=y https://vtechworks.lib.vt.edu/bitstream/handle/10919/83760/B... That's a funny query parameter, by the way. I tried setting isAllowed to n, but it still gave me the pdf, so who knows what it's for.
wolfgke 7y ago
> I don't understand why anyone would present the data like this, sure, 'lower is better' makes sense in benchmarks where you're presenting a raw timing or such, but with an arbitary unit like 'score' just invert the data so bigger -> better and let people's human nature take over. Also for bicycle helmets it makes sense, i.e. force to the head -> lower is better.

by NikkiA 7y ago

blacksmith_tb 7y ago

Especially when more stars is better! Overall their testing methodology looks good, so it's a little disappointing that the presentation is so counterintuitive.

psetq 7y ago
Their primary presentation is the 0-5 star rating which is perfectly intuitive. If I'm reading the methodology correctly, the actual STAR score represents the incidence of concussion, in which lower is certainly better. The only real confusing thing I see is the "star" naming of the "STAR evaluation score" vs. using a "star rating" (0-5) and how they are actually separate things.
- blacksmith_tb 7y ago
  I agree that the stars are clear, I wonder if there's any point in displaying the score at all, given that.

jolmg 7y ago

I think it depends no how this score is determined. If their rating methodology allows them to determine an upper bound on protection better than a lower bound, then it's better to set that bound to score zero, to not risk later needing scores with negative numbers. The formula seems to be on the following pdf on the third page. https://vtechworks.lib.vt.edu/bitstream/handle/10919/83760/Bicycle%20STAR%20Methodology.pdf?sequence=1&isAllowed=y https://vtechworks.lib.vt.edu/bitstream/handle/10919/83760/B... That's a funny query parameter, by the way. I tried setting isAllowed to n, but it still gave me the pdf, so who knows what it's for.

wolfgke 7y ago

> I don't understand why anyone would present the data like this, sure, 'lower is better' makes sense in benchmarks where you're presenting a raw timing or such, but with an arbitary unit like 'score' just invert the data so bigger -> better and let people's human nature take over. Also for bicycle helmets it makes sense, i.e. force to the head -> lower is better.