Are Dr. McIntire's comments ad hominem attacks also or just mine? Just curious.
What you are missing is this is not falsifying data or a false reading of the data. The data are what they are. That's your misunderstanding of my comments. Anyone working with data will tell you that it can be presented in various ways. The researchers chose one way. Everyone usually chooses one way.
Dr. McIntire, working it this field stated: (emphasis added).. the first comment appears to address the per review in part...
Again, what you (hdugger) are missing is this is not falsifying data or a false reading of the data. The data are what they are. That's your misunderstanding of my comments. Anyone working with data will tell you that it can be presented in various ways. The researchers chose one way. Everyone usually chooses one way.
Here is my best explanation for not including critical data in the publication:
Here is Dr. McIntire again which I never disagreed with... (emphasis added)
And I'll point out that publishing the average end points would add ONE LINE to the table but would require at least one paragraph to explain. So maybe there is a space limitation but I would think they would drop something else out before that. Not my field, their call.
Finally, you CONTINUE to make this about me using tenuous, if not completely ridiculous examples. Your breathless accusations to avoid responding to my points are not getting you anywhere. I could go back and make a list of points you have avoided. I will continue to point this out as long as you continue to do this.
What you are missing is this is not falsifying data or a false reading of the data. The data are what they are. That's your misunderstanding of my comments. Anyone working with data will tell you that it can be presented in various ways. The researchers chose one way. Everyone usually chooses one way.
Dr. McIntire, working it this field stated: (emphasis added).. the first comment appears to address the per review in part...
They don't publish the curve characteristics at the end of treatment.... and that makes it in the NEJM.... I can not believe this....... They don't calculate any type of "curve-delta". If it's ok to do that because the success/failure classification is >50°, then why would it be necessary to include the curve sizes at baseline? Why publish any data at all? Why not just say, "Curves were <40° at the beginning and 75% were <50° at the end. Trust us."
I side with you (Pooka) on this one. When you don't report data that is easily calculated and meaningful, there is a reason. ANY data from the endpoint concerning curve size is absent in the article and supplementary material.
The baseline data has plenty of detail. There is a maturity scale as well as different curve types and whatnot. Either in the paper or in the supplementary info. MY critique is primarily with the end point data. The average curve at baseline was ~35° in all groups (IIRC). The end point curve size would be good to know, although I don't think average is the best measure for curve size. Median is probably better unless the data are REALLY normally distributed.
Here is my best explanation for not including critical data in the publication:
It occurred to me that Weinstein/Dolan/et al. might be planning more publications. They almost certainly are in my opinion. I am guessing that is almost certainly the case as the first publication is really just an extended abstract because they don't show the deltas.
As far as any kind of nefarious motives, I tend to side with the scientists and give them the benefit of the doubt. It's not that I doubt their data as much as I question how meaningful it is if "success" is determined as <50°. If the pre-treatment group is 33° and the "success" group is 48°, well then I'm not sure I'd consider that to be a great outcome. The brace group had fewer surgeries than the observation group, and I think that is real. I like the creativity they used to randomize or to self select the treatment. I like the stratification of brace wearers to time in brace. I think they did an amazing amount of work to do a serious study and to control for a BUNCH of things are have been traditionally difficult, if not impossible, to do. They might have had a very strict limit on the amount of tables and figures they could have, which is why they probably had the supplemental info. So I WANT to say they left out the end point data as somewhat of an oversight because they had a bunch of other stuff to report. But, having published, the scientist knows their data better than anyone. To not mention anything about the end point other than success/failure percentage.... makes it tough to believe it was an oversight and probably means it brings the results into question/doubt. It would be great if I were wrong.
Finally, you CONTINUE to make this about me using tenuous, if not completely ridiculous examples. Your breathless accusations to avoid responding to my points are not getting you anywhere. I could go back and make a list of points you have avoided. I will continue to point this out as long as you continue to do this.
Comment