Pediatric Hospital Medicine 2012

Biostatistics for the Non-Statistician: Understanding Frequently used Methods in Medical Research
Matt Hall
At the conclusion of this session, participants will be able to:

  1. Understand basic bivariate tests and learn when they are appropriate.
  2. Understand the process for selecting an appropriate multivariable model
  3. Interpret the results of common statistical tests

Understanding statistical concepts is essential when critically reading manuscripts, allowing the reader to assess the validity of the conclusions and implications for practice. It is also important when designing your own studies and communicating with your biostatistician. This session will introduce many of the statistical concepts that are frequently used in medical research, and will focus on appropriately selecting tests and interpreting results. We will briefly explore different types of data. Next, we will discuss hypotheses and many of the common tests (e.g. t-test, Wilcoxon rank sum, chi-square) and models (e.g. logistic, linear, mixed, GEE), and describe why they are selected from the statistician’s tool box. Along the way, we will emphasize the appropriate interpretation of statistical findings (e.g. p-values, odds ratios, model coefficients) and common pitfalls to avoid. Participants will evaluate the appropriateness of the statistical methods used in several recently published pediatric research papers.


 

Scholars

Renew Membership

Join APA

Donate

APA Journal

MailingList

ResidentFellow