Which sampling method is most likely to be biased because selection is not random and may reflect researchers' preferences?

Which sampling method is most likely to be biased because selection is not random and may reflect researchers' preferences?

• A. Purposive Sampling
• B. Systematic Random Sample
• C. Stratified Sampling
• D. Snowball Sampling

Answer: (A)

The main idea here is how the way participants are chosen can introduce bias when randomness isn’t used. Purposive sampling is when the researcher selects individuals based on specific characteristics or a judgment about who will provide the most informative data. Because this choice is driven by the researcher’s aims and expectations rather than chance, personal preferences or theoretical leanings can shape who gets included. That makes the sample more likely to reflect the investigator’s views and expectations rather than a representative slice of the population, which in turn can bias the results and limit how well they generalize. In contrast, methods that incorporate randomness—like selecting participants at random from a larger group, or stratifying the population and then randomly sampling within each stratum—help reduce this kind of bias. Snowball sampling, while non-random and potentially biased by social networks, is driven more by participants recruiting others than by the researcher’s explicit preferences, though it can still introduce biases due to who is connected.

The main idea here is how the way participants are chosen can introduce bias when randomness isn’t used. Purposive sampling is when the researcher selects individuals based on specific characteristics or a judgment about who will provide the most informative data. Because this choice is driven by the researcher’s aims and expectations rather than chance, personal preferences or theoretical leanings can shape who gets included. That makes the sample more likely to reflect the investigator’s views and expectations rather than a representative slice of the population, which in turn can bias the results and limit how well they generalize.

In contrast, methods that incorporate randomness—like selecting participants at random from a larger group, or stratifying the population and then randomly sampling within each stratum—help reduce this kind of bias. Snowball sampling, while non-random and potentially biased by social networks, is driven more by participants recruiting others than by the researcher’s explicit preferences, though it can still introduce biases due to who is connected.