1. What is the deviance for

    • the full/saturate model
    • the fitted model

  2. Calculate the log-likelihood and deviance values of this fitted model using information from R output.

  3. State the hypotheses being tested in a goodness of fit test by the analysis of deviance approach.

  4. How to make conclusion from a goodness of fit test based on an R output?

  5. Calculate the odds and 95% confidence interval for the calculated odds given a set of covariates.

  6. What is the Hessian matrix of the log-likelihood function?

  7. Explain whether the log-likelihood function is a concave function given the link

  8. ROC, sensitivity, specificity

  9. Matched case-control study

  10. Total number of parameters in different GLMs, e.g., multinomial model

  11. Link functions/corresponding latent variable distributions

  12. Exponential family distributions

    • What is the canonical parameters?
    • Mean and variance.
    • Canonical link function