Exercise Math Achievement

Using the dataset MathAchievein the MEMSS package carry out an analysis that treats School as a random effect. Answer the following questions:

-Which variables should be treated as fixed effects?
-Are differences between schools greater than can be explained by within-school variation?
-Is the relation between SES and MathAch different for different schools? Compare the models using deviance difference tests.
-Discuss how the decision whether to treat `School` as a fixed or a random effect might depend on the purpose of the study.

Exercise Nurses

The file nurses.sav holds data from nurses working in wards nested within hospitals (Hox, J.J. (2002). Multilevel analysis: Techniques and applications. Hove: Routledge). It is from a hypothetical study on stress in hospitals where wards were randomly assigned to and experimental or control condition. It contains the variables age(years), experience(years), gender (0=m, 1=f), wardtype (0=general care, 1=special care), expcon (0=control, 1=experimental), hospsize (0=small, 1=medium, 2=large), stress (likert scale 1-7) and the id variables hospital, ward and nurse.

Load this file into R using the function read_spss() from the haven package, grand mean center the variables age and experience and answer the following questions:

- Are differences between wards greater than can be explained by ward and nurse varying variables?
- Are differences between hospitals greater than can be explained by hospital, ward and nurse varying 
  variables?
- Is the relation between experimental condition and mean ward stress-score the same in all hospitals?
- If this relation is different, can the size of the hospital explain (a part of) these different 
  relations?

End of Practical.