,
8 tweets,
3 min read
Read on Twitter
Let's try an exercise that we would do in grad school.
Can any science fans spot how this study's methods/results don't match its main takeaways? nature.com/articles/s4159…
Can any science fans spot how this study's methods/results don't match its main takeaways? nature.com/articles/s4159…
Some clues from the paper...
P.S. I only dunk on a questionable study when more than 2 people bring it to my attention, in case anyone is wondering.
WE HAVE A WINNER!
Their whole premise is based on a logistics model that connects 😈 with:
- sex
- forward head protraction
- age.
But the FHP is high in older adults (as one might expect; Fig.3) as well as the prevalence of 😈 (fig. 4).
- sex
- forward head protraction
- age.
But the FHP is high in older adults (as one might expect; Fig.3) as well as the prevalence of 😈 (fig. 4).
In other words, older adults contribute to their trends, which they admit...
They try to shore up their case by using Chi-square to compare the age groups...but my stats-lovers out there might recognize the issue with using Chi-square for such an analysis.
They try to shore up their case by using Chi-square to compare the age groups...but my stats-lovers out there might recognize the issue with using Chi-square for such an analysis.
Those statistical blips -- plus the fact that they didn't actually measure cell phone usage 👇-- make the takeaways pretty suspect.
Best part of this thread: I'm on deadline for another story, so I thought my bosses might consider this thread to be light procrastination.
But now my bosses want me to turn this thread into a story.
But now my bosses want me to turn this thread into a story.
