Viraj and Paul,
If memory serves the poll being discussed in #60 and the 2 quotes above, is this one,
Poll: does NL glare depend on eyeglasses?.
I have a wee bit of experience with both statistical process control in manufacturing and market research polling from marketing, (albeit since I'm an old fart), both from a career now well in the past. I have a fair understanding of these tools. I am not claiming I should be teaching them to anyone.
I fear the term statistical significance has become popular and like a lot of things that go that way, the actual meaning tends to wander. You can read about it here,
A Refresher on Statistical Significance.
It is doubtful the results of this poll should be thought of as statistically significant, for several reasons. First worldwide population is estimated at 7.7+ billion. How many millions own binoculars? Can 32 people represent what this rather larger cohort experiences with glare from any binocular?
Sample size can be small but other conditions need to be met, (see article above), when that is the case. What one is measuring - parts/dimensions or people's experience is relevant. What one is using to measure - questions or tools is critical. It’s hard to accurately measure the diameter of a hole with a 12” straight ruler. Maljunolo, is exactly correct. The sample here of 32 respondents might work if the questions were designed differently. As well though, it’d be better if the outcomes were more distinctly different, then these 4 sort of close responses - 28.1, 31.3, 25.0, 15.6. One doesn't need statistics if you grab a handful of 5 parts from a bin of say 25 and all measure exactly the same. The odds are good the rest are good. If those same 5 had a range in measurement, (depending), the confidence in what's in the rest of the bin goes down.
Respectfully though, as I enjoy hanging here at least between migrations, the real problem is something else. Birdforum's Binocular subgroup is probably the last place one should direct questions such as this if we really needed a reliable outcome.
Why?
The subject of glare has been controversially discussed in this quite closed community for years. Sides have been drawn. A few are ardent supporters, and cite personal experiments to prove its existence. Others just as enthusiastically say something like “I can't see it, therefore how can you say it exists?" It's not clear, reading the back and forth, that there is agreement as to what glare is. Is it a little crescent of light in the lower portion of the FOV? Is it some sort of veiling thing making the whole view cloudy? Some combination of those? Or is it environmental conditions, being blamed as a product of a binocular thing? Mostly, we don't know each other personally. We haven't stood in the same place, looking at the same conditions, through the same binoculars and shared what we each saw to be able to definitively say we all agree what it is. What if someone is not trained, comes here and reads a bit, then goes out and looks out over water on a sun shiny day and sees glare and thinks "Aha, these binos are bad?”
Im taking bets.
Based on the small sample, question construction, close responses, and biased pool, this is what can safely be said based of this survey.
28.1% of 32 Birdforum respondents use the NL without glasses, do not see glare
31.3% of 32 Birdforum respondents use the NL without glasses, do see glare
25.0% of 32 Birdfourm respondents use the NL with glasses, do not see glare
15.6% of 32 Birdforum respondents use the NL with glasses do see glare.
Thats it.
One BFer shortly after this survey occurred, made the statement in a different thread, 46% of “birders” see glare in Swarovski NLs, citing the survey as proof. Sorry, no. This limited questionnaire cannot be interpreted to produce that conclusion. It seems even a stretch to predict things like 46.9% of BF respondents when viewing through an NL binocular will see glare or 53% will not, based on this.