A 13-year study of 1328 adults randomly selected from a population carefully monitored the personal habits and health conditions of participants. Personal habits included tobacco use and coffee consumption. Health conditions included incidence of stroke. Which of the following questions about this population CANNOT be answered using data from this study?

(A) Are coffee drinkers more likely to smoke than adults who do not drink coffee?
(B) Does coffee consumption cause a reduction in the incidence of stroke?
(C) Do coffee drinkers have fewer strokes than adults who do not drink coffee?
(D) What percentage of the population are coffee drinkers?



S-IC.3: Recognize the purposes of and differences among sample surveys, experiments, and observational studies; explain how randomization relates to each.

S-IC.6: Evaluate reports based on data.

Correct answer and commentary

At the advanced level of the GAISE framework students ask their own statistics questions and seek generalizations as part of the Formulate Questions component of the statistical problem-solving process. While multiple-choice questions do not allow students the freedom to respond with their own statistics questions of interest, this advanced-level question assessing the Formulate Questions component asks students to identify which statistical questions can be answered using the information that is provided about the study. At the advanced level, students are expected to have a deeper understanding about what types of statistics questions can be answered using statistical analyses and the data available.

In this question, the questions in options (A), (C), and (D) can all be answered using the data that were collected. Because this is an observational study – personal habits and health conditions were monitored but not changed by the researchers – the question in option (B) cannot be answered. To answer a question involving causation as indicated in option (B), participants must be randomly assigned to experimental conditions. Without random assignment, the researchers can only make claims about correlation, not causation. As students become more sophisticated in their understanding of statistical questions, they should be able to recognize the types of questions that can be answered by data collected from surveys, experiments, and observational studies and to evaluate claims arising from such studies.

Student performance