3 ms·
Take what you can get reliably, use it as a sample and try and extrapolate out making adjustments if the sample population differs from the overall population.
by dsfyu404ed 4y ago
Take what you can get reliably, use it as a sample and try and extrapolate out making adjustments if the sample population differs from the overall population.
This is SOP in god knows how many fields.
- MrJohz 4y ago> making adjustments if the sample population differs from the overall population. This feels like a step you can't just handwave away as step three of three. To quote the meme, it's a bit of a "draw the rest of the fucking owl" situation. Extrapolation is hard, and there are very good reasons not to do it willy nilly. Where it is standardised in other fields, it's usually because researchers have already figured out good ways to extrapolate the data and proven the the extrapolation appears to be valid (and even then, usually only within limited bounds).