Outline 1) Descriptive Statistics 2) Define “association”. 3) Practice reading Table 1 for evidence of confounding, effect modification. 4) Practice reading scatterplots for evidence of confounding effect modification. 5) Review comparing adjusted and unadjusted analysis for confounding effect modification. 6) Read STATA output: saturated model predicted group summary is actual group summary
10
Embed
Outline 1) Descriptive Statistics 2) Define “association”. 3) Practice reading Table 1 for evidence of confounding, effect modification. 4) Practice reading.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Outline1) Descriptive Statistics 2) Define “association”.3) Practice reading Table 1 for evidence of
confounding, effect modification.
4) Practice reading scatterplots for evidence of confounding effect modification.
5) Review comparing adjusted and unadjusted analysis for confounding effect modification.
6) Read STATA output:saturated model
predicted group summary is actual group summarycorrelationevidence of association from overall hypothesis test
Descriptive Statistics Goals:1) Identify measurement of data entry errors2) Characterize the methods and materials3) Assess the validity of the scientific and statistical assumptions4) Get a straight forward estimate of the association you are interested in5) Explore data to generate hypothesis for future studies
Reporting descriptive analysis:Give the reader the ability to judge the scientific evidence and the importance of your work.
Methods: Describe what you did so that the reader could reproduce your work. Results: Report what you actually realized in this repetition of the research.
Present a preliminary estimate of the association. Indicate whether this sample supported the scientific and statistical
assumptions used to do the statistical analysis.
Association : The distribution of two variables are not independentThe conditional distribution of an outcome variable changes depending on the value of the predictor.
Rather than look at the entire distribution we use a summary measure for the distribution of the outcome and compare the value of that summary measure at different values of the predictor.