Sampling variability vs. sampling bias
SAT glossary · Course Foundations
Definition
Margin of error accounts for random sampling variation only and says nothing about a bad sampling frame, leading questions, or non-response. A biased survey with a huge sample has a tiny margin of error and a worthless estimate — options claiming a bigger sample cures a flawed method test exactly this.
Sampling variability vs. sampling bias in plain English
These are two totally different reasons a survey estimate might be 'off,' and mixing them up is a serious trap. Sampling VARIABILITY is just the normal, expected wobble that comes from randomly picking a sample instead of surveying literally everyone — margin of error accounts for exactly this, and only this. Sampling BIAS is a completely different problem: it happens when your method for choosing who gets surveyed is flawed from the start (like a bad sampling frame, leading/unfair questions, or lots of people simply not responding). Here's the trap: a badly biased survey can still have a tiny, impressive-looking margin of error if the sample size is huge — but a tiny margin of error says NOTHING about whether the underlying method was any good. So an answer choice claiming 'a bigger sample fixes a flawed survey method' is wrong — bigger samples fix variability, not bias.
Sign up free: all 157 SAT terms as flashcards, plus practice questions with full explanations →