a study of 110 men ? Thats only slightly better than the mmr bullsh*t isnt it ?
Andrew Wakefields paper was based on 12 children, but that was the least of its problem. No ethics and false data were far greater problems...!
I've had a look at the paper and it does raise some concerns. The statistics are quite lacking.
The blogger incorrectly states that 'this created what is known as the P-value in statistics and this means there is clinically significant evidence to prove that vasectomy links men to a higher risk of PPA (a form of dementia)'. It doesn't. It just indicates whether the effect could be explained by chance alone. It doesn't actually state what the effect is or the size of the effect.
Thankfully with a little bit of maths you can calculate it. 19/47 PPA patients had vasectomy whilst 9/57 controls had vasectomy. (19*48)/(28*9) = 3.6. So having a vasectomy increases your risk of PPA by a factor of 3.6. (Quite how the paper got accepted without even reporting the Odds Ratio is quite worrying, but that's psychology journals for you....). However you want a measure of how accurate that estimate is, what we call the 95% confidence intervals. These are 1.44 -9.08. So the true estimate could range from only a 40% increase, to a 9 fold increase. It's so large because it's the discordant pairs that are important. In this study, you have on 9 control patients with vasectomy.
Furthermore, you want to control for things like age and medical history and all sorts to make sure the groups aren't different in other ways that could explain the increase in vasectomies in the PPA patients. The study does control for age, but not much else other than education. It also doesn't give the adjusted rate (3.6 is the unadjusted rate), and their 95% confidence intervals. The effect could get even less once you control for these factors. Again, lacking in any detail.
Basically IMO, for the blogger to conclude "the link to vasectomy is present and clear" is rather perplexing. Certainly could be a link, and does provide theoretical grounds for the biological mechanism, but the evidence is very weak and ambiguous. Need a much larger scale study controlling for more confounding effects to conclude anything either way....
Hope that psychology graduate improves her understanding of statistics before going to a PhD....!