If that 10k is representative, that should give a very small uncertainty interval, less than 1%. You can get 95% confidence interval with only a few hundred samples depending on the standard deviation, so 10k is actually massive. It’s pretty standard statistics, here’s more info on how it’s calculated.
Comment on Precarious finances: 38% of Europeans no longer eat three meals a day
Mojojojo1993@lemmy.world 11 months ago
Where do they get these stats from ?
JohnDClay@sh.itjust.works 11 months ago
shalafi@lemmy.world 11 months ago
I know jack shit about statics, or this study, but 10K participants seems more than solid if it’s proper science.
AlexandroffExtension@lemmy.dbzer0.com 11 months ago
Dug around a little. Seems the 10k were split between 10 different states. Here is an infographic from the source:
The numbers do seem inflated and don’t add up very well comparatively between the different countries either.
Mojojojo1993@lemmy.world 11 months ago
Ah didn’t see the breakdown. Incredibly leading questions and highly charges but seems legit
match@pawb.social 11 months ago
you’re right! can you calculate the sampling size that would’ve given 95% confidence
CazzoBuco@lemmy.world 11 months ago
10k is a great number for this kind of extrapolation. The only concern is how varied that population was
Mojojojo1993@lemmy.world 11 months ago
For this kinda thing ? Really ? 10k to extrapolate into the population of what 300 mil ?
Think that’s pretty wild. For maybe a single question with a few variations but something as complicated and complex as poverty and spending habits.
Iyf like to see the sources to back up that claim.
quicksand@lemm.ee 11 months ago
Yup, check this parent comment for the source you’re asking for