Question

Mary wants to find out whether eating sugary snacks before an exam leads to better performance on the exam. Which of the following strategies would answer her question most conclusively?

a) Identify a large number of students who perform exceptionally low and exceptionally high in exams, ask them whether they eat sugary snacks before exams and see whether high performers eat more sugary snacks before exams than low performers.
b) Wait for exam time in a big class, ask everyone whether they ate sugary snacks before the exam, and see whether those who ate sugary snacks do better compared to those who didn't.
c) Wait for exam time in a big class, give a random half of students M&Ms before the exam and see whether the students who ate M&Ms perform better.
d) Pick a big class, give all students sugary snacks before one exam and salty snacks before the next exam, then see whether students score lower on the second exam.

Answer 1

The most conclusive method to determine the impact of sugary snacks on exam performance is to conduct a randomized controlled trial, giving half the students M&Ms before the exam and comparing their performance with those who did not consume them.

Step-by-step explanation:

Designing an Experiment to Test the Effects of Sugary Snacks on Exam Performance

The most conclusive strategy to determine if eating sugary snacks before an exam improves performance would be option (c): Wait for exam time in a big class, give a random half of the students M&Ms before the exam, and see whether the students who ate M&Ms perform better. This strategy represents a randomized controlled trial, which is the gold standard in experimental design. Randomization ensures that the two groups (those who eat the M&Ms and those who do not) are likely to be similar in all respects except for the intervention being tested, reducing the impact of confounding variables or biases.

Option (a) involves observational data which can be influenced by confounding variables, such as study habits. Option (b) also involves observation but is limited as it does not control for students' natural propensity to eat sugary snacks. Option (d) provides a comparison between sugary snacks and salty snacks, which could present separate effects that are not strictly comparable in terms of influencing exam performance.

An experiment with a control and random assignment lays out a clearer path towards establishing a cause-and-effect relationship, as opposed to correlational studies. Moreover, ensuring the sample size is large enough helps provide more reliable results that could be generalized to a wider population.