Multivariable data analysis part 2 Flashcards
what command is used to regress a continuous competing variable
regress exposure outcome variable
what command is used to regress a categorical competing variable
xi: regress exposure outcome i.catergorical variable.
what does the coefficient information generated in STATA show
y=mx +b
gives you the m + b
what does the 95% confidence interval show
confidence interval of the coefficients.
what do the p values show
significance of each variable in the model
what does R^2 show
Proportion of VARIATION explained by the linear model
how well it fits the model
What is the range of R^2
0-1
does a higher or lower R^2 show a better ft
higher
what is another way to describe R^2
correlation
what effect does adding more variables to your model have on the r^2
improves the fit (including adjusting factors)
what does adjusted R^2 show
measure of fit that accounts for the number of variables
what 2 methods are used to compare models
adjusted r^2
p values for each additional variable.
what 2 variates are included in statistically analysis to improve fit
confounders
competing exposures.
when is logistic regression used
when the outcome is binary
how must the outcome be coded in logistic regression.
0 and 1
