Step 1: Understand snowball sampling.
Snowball sampling is a nonprobability, network-based technique used for hard-to-reach/rare populations. Initial respondents recruit others. There is typically no explicit frame; inclusion depends on social links.
Step 2: Implication for generalization.
Because selection probabilities are unknown and the underlying population often isn’t explicitly enumerated, estimates cannot be directly generalized to a defined population with known sampling error.
Step 3: Rule out distractors.
- (b) Judgmental and (c) Convenience also limit generalization, but the stem highlights population not defined explicitly, a signature problem in snowball designs (hidden populations).
- (d) Simple random sampling is probability-based and supports generalization.
\[
\boxed{(a)}
\]