Regression

The analysis continues...

5. Complete a logistic regression to examine the association of each test with diagnosis, controlled for confounders. Consider how to decide what confounders to include. Consider the pros and cons of adding in adidtional predictors. Consider how to present our data to help clinicians make decisions in the real world

Remember to closely inspect the model output summary, which should look something like this:

> > m_testB <- glm(diagnosis ~ + testB ++ depression + + age, + data = df,+ family = 'binomial')> summary(m_testB)
Call:glm(formula = diagnosis ~ testB + depression + age, family = "binomial",     data = df)
Coefficients:            Estimate Std. Error z value Pr(>|z|)    (Intercept) 20.9387 3.9966 5.239 1.61e-07 ***testB -0.1435 0.0422 -3.400 0.000674 ***depression 0.7566 0.1553 4.871 1.11e-06 ***age -0.2857 0.0540 -5.291 1.22e-07 ***---Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
    Null deviance: 275.87 on 198 degrees of freedomResidual deviance: 161.63 on 195 degrees of freedom  (1 observation deleted due to missingness)AIC: 169.63
Number of Fisher Scoring iterations: 6

Here is some R code to help you with task 5