Sprint 5.2: Sampling And Bias

Let's practice spotting research flaws by evaluating two reasoning scenarios. In Study A, researchers claim that drinking five cups of coffee daily causes lung cancer, but they failed to check if the participants also smoked cigarettes. In this case, smoking acts as a because it is an unmeasured third factor that influences the outcome, making the cause-and-effect claim invalid. In Study B, a student surveys just ten close friends in a school of two thousand students, concluding that 90% of the school wants online classes. Relying on such a tiny number of participants results in a high , meaning the results cannot reliably represent the entire student body. Furthermore, selecting only close friends instead of a random group introduces a severe , as friends often share similar views and do not reflect the diversity of the school. By systematically identifying these specific flaws, you can accurately evaluate the validity of any scientific or social research claim.

To ensure all socioeconomic groups were proportionately represented in the sociology project, the students used a sampling method. In the subsequent biology experiment, the trials for the new plant nutrient were completely to prevent any selection bias. Before administering the experimental treatment, the researchers carefully measured the health metrics of the to establish an accurate point of comparison. Later, when analyzing the survey data for their economics paper, the team reported a of plus or minus three percent to account for statistical uncertainty. Finally, during the analysis of the physics lab data, the software applied advanced to remove background interference and isolate the true signal.