On Thursday, February 21, 2019 at 2:12:14 PM UTC-8, Andre Jute wrote:
On Thursday, February 21, 2019 at 5:13:35 PM UTC, Frank Krygowski wrote:

Thompson & Rivara did not take kindly to her rebuttals. They never again
allowed her access to their data.

If it is true -- give us an independent reference to their refusal, Franki-boy -- that Thompson & Rivara deny other researcher their raw data, plus a comprehensible description of any manipulation of the data to arrive at their published set, then any paper on which they do so falls. because it is logical to assume that there is something wrong with their work, and that they know it. It is to avoid this suspicion that journals of scientific record demand to have the raw data on file before they publish the article analysing the data, and also that the data be made available.

There has been a case recently that is notorious around the world of not just a researcher who refused to provide the raw data, but a whole subclass of researchers, who then brought an entire branch of science dependent on statistical analysis into deep, deep disrepute when their "adjustments" were exposed as self-serving lies, and the reason they obstructed legitimate freedom of information requests. The researcher was Michael Mann, the subbranch of a scientific field was dendrochronology, and the field that was damaged by Mann's now disgraced "hockey stick" was climate studies -- more specifically, worthless climate alarmism.

Andre Jute
There are codes of practice for a very good reason. They are for the self-protection of decent professionals.

Ms. Robinson mentioned that she could not get later datasets if memory serves.

But that doesn't change the fact that you cannot correct for confounding factors such as those most likely to be involved in accidents are NOT experienced sports riders but those that ride on the wrong side of the road with cruiser-type bikes in poor mechanical conditions. ER's do NOT make any such judgements and can only report "with and without" helmets. None of the research I've ever seen make the slightest sense with carefully analyzed. They are ALL heavily biased one way or the other. I seem to have made the only comparison that was totally neutral.

I saw that motorcycle safety helmets were effective ONLY on tracks and assumed that this safety would show up in far more effectiveness of bicycle helmets because of the reduced speed. My analysis was quite a surprise to me.

I recently tried extending the ratio of bicycle deaths vs. pedestrian deaths and it appears to hold the same ratios. So helmets as we know have not changed and pedestrians haven't improved their judgements.