Which statement correctly distinguishes sampling error from non-sampling error?

• A. Sampling error arises from using a subset; non-sampling error arises from bias, measurement, nonresponse, etc.
• B. Sampling error arises from bias, measurement error.
• C. Non-sampling error occurs only in full census.
• D. Non-sampling error is resolved by increasing sample size.

Answer: (A)

The main idea here is how errors arise in data collection. Sampling error is the random difference that comes from studying only a subset of the population. When you draw a sample, you’re likely to get a slightly different estimate (like a mean or proportion) than you would from the whole population, simply due to chance. This type of error shrinks as your sample gets larger, but it never disappears completely.

Non-sampling error covers issues not tied to the act of sampling itself—systematic problems like bias in how the study is designed, mistakes in measurement, nonresponse bias (certain groups not responding), data entry or processing errors, and other flaws. These can occur regardless of how large or small your sample is, and they can occur even in a complete census.

So the correct statement captures that sampling error arises from using a subset, while non-sampling error stems from bias, measurement error, nonresponse, and similar problems. The other choices mix these concepts or claim things that aren’t true—non-sampling error does not only occur in a census, and it isn’t fixed just by increasing sample size.