Labs for Data Analysis

Periodically we will devote a class session to applied data analysis. The purpose of the lab is two fold: (1) to assist with planning your analysis while pointing out some common problems that might occur in your own research, and (2) to provide some practical experience with using statistical software and interpreting thre results. You can prepare for the labs by considering how you would describe and analyze the following datasets, and then carrying out your analysis plan in whatever software program you would like. During the class, we will discuss your ideas for how to analyze the data, conduct the analysis using R, and interpret the results. Datasets are posted in excel format and text (comma separated value or csv).

In the future, let us know if you have a particular dataset that you would like to see analyzed from your own research. We will incorporate it into the course and/or a future lab as appropriate.

Lab 1

  • Dataset 1: The investigators measured a pre (before intervention) value in 20 mice, assigned each mouse to treatment or placebo groups, and remeasured the same outcome after receiving the treatment/placebo. Is there a treatment effect?
  • Dataset 2: The following contains the results from two studies in one file to simplify the process of importing data into your software package. Both trials contain 40 subjects with the same values for age and gender. However, treatment/control status (trt1 or trt2) and the outcome (y1 or y2) differ for study 1 and study 2, so separate analyses should be performed. For both studies, the research question is the same: Is there an association between the treatment and the outcome?
  • Note that there is something unusal occurring in each of these datasets so that just results from statisical tests may not appropriately answer the research questions.

Lab 2

  • cPLA2 dataset description
    • Wild type mice usually deliver around day 19.5 of pregnancy, but it is believed that cPLA2 knockout mice have a delay in the timing of delivery.
    • While the knockout mice deliver later, they also have smaller liter sizes; it has been argued that the later delivery is due to smaller litters.
    • The investigators believe that there is more to it than just smaller liter sizes, so they collected data on the timing of delivery and litter.
    • They also attempted to rescue the delay in delivery by treating with PGE+PGI.
  • Using graphics and statistical tests, examine the association of genetics and PGE+PGI treatment with
    • length of gestation
    • litter size
    • length of gestation adjusted for liter size
  • CPLA2.teaching.csv
  • CPLA2.teaching.xls
Topic attachments
I Attachment Action Size Date Who Comment
CPLA2.teaching.csvcsv CPLA2.teaching.csv manage 1.3 K 20 Jan 2009 - 11:57 ChrisSlaughter  
CPLA2.teaching.xlsxls CPLA2.teaching.xls manage 118.0 K 20 Jan 2009 - 11:58 ChrisSlaughter  
prepost.csvcsv prepost.csv manage 0.4 K 26 Oct 2008 - 11:19 ChrisSlaughter Pre and post measures
prepost.xlsxls prepost.xls manage 75.5 K 26 Oct 2008 - 11:19 ChrisSlaughter Pre and post measures
trials.csvcsv trials.csv manage 1.1 K 26 Oct 2008 - 11:19 ChrisSlaughter Two clinical trails
trials.xlsxls trials.xls manage 74.0 K 26 Oct 2008 - 11:18 ChrisSlaughter Two clinical trails
Topic revision: r5 - 20 Jan 2009, ChrisSlaughter
 

This site is powered by FoswikiCopyright © 2013-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback