WTL Chocolate and Cycling Prompt

You have been hired by the U.S. bicycle team to help them train for the Tour de France. The head trainer recently read an article, which claims that consumption of dark chocolate results in increased oxygen consumption during cycling. The experimental setup consisted of a randomized crossover design where the oxygen consumption of n = 9 male participants was measured in two trials after participants consumed either dark chocolate or white chocolate. A crossover design is a repeated measurements design such that each subject receives the two different treatments (dark chocolate versus white chocolate) during the different time periods, i.e., the patients cross over from one treatment to another during the course of the experiment. The order of which treatment was received in the first time period was randomized. Prior to receiving the first treatment, each participant underwent baseline measurements. Data was gathered and analyzed as depicted in the table below. Maximal Oxygen Consumption* (Note: n = 9 for each condition) Baseline White Chocolate Dark Chocolate Mean (ml/kg min) 41.89 41.84 44.52‡ Std dev 5.4 5.6 6.43 p-value† - 0.071 0.037 †p-value is for statistical comparison with respect to baseline ‡ Dark Chocolate: 95% Confidence Interval for the population average change in maximum oxygen consumption (over baseline) is 0.21 ml/kg min to 5.05 ml/kg min. The trainer knows you have some statistics background and wants your opinion about whether or not dark chocolate should be added to the athletes’ diets. Based on the results from the article, write a memo to the trainer explaining what the statistics show and make an argument for or against inclusion of dark chocolate in the athletes’ diet. Your memo should include a little discussion about how a crossover design affects the data analysis. Describe what the p-values indicate about the results, and the meaning of statistical significance. Finally, comment about the provided confidence interval, including both an interpretation of the confidence interval itself and the meaning of the confidence level in context.