WBR0365

Author	[[PageAuthor::Serge Korjian, M.D., Gonzalo A. Romero, M.D. [1] (Reviewed by Serge Korjian)]]
Exam Type	ExamType::USMLE Step 1
Main Category	MainCategory::Biostatistics/ Epidemiology
Sub Category	SubCategory::General Principles
Prompt	[[Prompt::A research group is trying to study the effects of a short course of concentrated coffee bean extract on weight loss in healthy volunteers. They design a randomized double-blinded controlled trial and recruit 154 subjects. Subjects are randomized in into 2 groups one receiving 3 daily doses of coffee bean extract and the other receiving 3 daily doses of a placebo. At the end of the 6 month period, the biostatistician wants to compare the mean change in body weight from baseline in the 2 groups. Which of the following tests should the biostatistician use in order to obtain a valid comparison?]]
Answer A	[[AnswerA::χ2 (Chi-square) test]]
Answer A Explanation	[[AnswerAExp::Chi-square is a test which allows to determine a difference between 2 or more proportions or percentages of categorical outcomes.]]
Answer B	AnswerB::ANOVA
Answer B Explanation	[[AnswerBExp::ANOVA or analysis of variance allows to determine if a difference exists between the means of 3 or more groups. It is used in the same context as the t-test but when comparing more than 2 groups.]]
Answer C	AnswerC::T-test
Answer C Explanation	[[AnswerCExp::The T-test allows the comparison of 2 means for statistical significance. It can only be used with 2 means whereas ANOVA evaluates 3 or more mean values.]]
Answer D	AnswerD::Pearson's correlation coefficient
Answer D Explanation	[[AnswerDExp::The Pearson correlation coefficient is a measure that demonstrates the correlation between 2 variables. It is not used to determine if statistical significance is achieved and cannot be used to compare means. This correlation varies from +1 to -1; with negative values showing inverse relationship. The closer the absolute value is to 1, the stronger the correlation whether positive or negative.]]
Answer E	AnswerE::Median
Answer E Explanation	AnswerEExp::The median is the numerical value separating the higher half of a data sample from the lower half. It may be used to determine a skew in the distribution of the study population.
Right Answer	RightAnswer::C
Explanation	[[Explanation::The student's t-test is used to evaluate a significant difference between 2 means of continuous variables. In this study, the researchers are looking to compare the mean Δbody-weight between the 2 groups making the t-test the most appropriate choice. The t-test operates under the assumption that the population has a normal distribution. The t-test can be either independent (unpaired), used when two independent sets of similarly distributed samples one from each of the two groups being compared, or paired used in cases where the study population consists of matched pairs. Educational Objective: The T-test is used to detect a statistically significant difference between the means of 2 groups. References: Gerstman BB. Basic Biostatistics, Statistics for Public Health Practice/ Formula and Tables. Jones & Bartlett Learning; 2007.]]
Approved	Approved::Yes
Keyword	WBRKeyword::t-test, WBRKeyword::ANOVA, WBRKeyword::Chi-square, WBRKeyword::Median, WBRKeyword::Pearson's correlation coefficient, WBRKeyword::correlation coefficient, WBRKeyword::clinical trial
Linked Question	Linked::
Order in Linked Questions	LinkedOrder::