Explain why sampling error matters and how to reduce it.

• A. It arises from using a sample; reduce with larger samples, proper sampling design, weighting.
• B. It is the difference between populations and samples due to nonresponse only.
• C. It cannot be reduced; it's inherent in sampling.
• D. It is the same as bias.

Answer: (A)

Sampling error is the random difference between a sample estimate and the true population value, because a sample is only a subset of the whole group. This matters because it sets the precision of your estimate; even with a perfectly planned random sample, repeated samples would produce slightly different estimates, and you quantify that uncertainty with standard error and margins of error. You can reduce sampling error by collecting more data—the larger the sample, the closer the estimate tends to be to the population parameter. A carefully designed sampling plan also cuts variability: use random selection so every unit has a known chance of selection, employ stratification or other designs to ensure key subgroups are represented, and consider techniques like clustering or multistage designs that minimize variance. Weighting helps align the sample with the population and can reduce the variance of estimates when probabilities of selection or response are unequal. It’s helpful to keep in mind that nonresponse and measurement error are separate sources of inaccuracy; sampling error specifically concerns the random variation from drawing a sample, not systematic bias or missing data.