You are asked to analyse data from a recent observational study of ageing.
Data come from a single survey where people were asked a range of questions. A subset of variables is included in your dataset. You are asked to develop an explanatory model for cognitive function by
1) Identifying exposures which are relevant to cognitive function as measured by a memory score.
2) Testing whether the effects of those exposures are different for men and women.
3) Reporting your final model.
You are not expected to investigate any effect modification other than by the variable sex.
|
variable name
|
variable label
|
Label/coding/range
|
|
sex
|
sex
|
0= Male 1=Female
|
|
age
|
age
|
from 50 to 90
|
|
cohabiting
|
living with a partner
|
0=No, 1=Yes
|
|
diabetes
|
whether has diabetes
|
0=No, 1=Yes
|
|
cancer
|
whether has cancer
|
0=No, 1=Yes
|
|
memscore
|
CF memory score
|
from 0 to 27 (higher score for better memory)
|
|
wealth
|
quintiles of wealth
|
1=Poorest, 2nd , 3rd , 4th , 5=Richest
|
|
depression
|
whether has depression
|
0=No, 1=Yes
|
|
srhealth
|
self-rated health
|
1=excellent, 2=very good, 3=good, 4=fair, 5=poor
|
Some guidelines:
The report should be understandable to people who are not familiar with the data but are aware of the general aims of the study, and who need to know:
1) how you analysed your data, including strategies you used and reasons for decisions you made,
2) the results you obtained,
3) your assessment of the validity of your results,
4) your interpretation of your results.
Many of the principles involved in writing a good paper for a journal apply to this assignment but:
1) There should be more detail on the strategy and assessment you used to analyse the data than is found in most papers.
2) You should not explain the background to the study or the methods of data collection.
3) You should not use, cite or discuss relevant literature but you should consider the internal validity of the study (and hence your results).
4) You should discuss the interpretation of your results and how they relate to the question you were asked.
5) You may assume that the techniques you have used (e.g correlation, chi-squared test, t-test, confidence intervals, Mantel-Haenszel, regression, Wald test etc) are understood by the reader and do not need explanation, but you do need to say which techniques you used and for what purpose.
The following should be clear to the reader:
1) how you checked the data for errors
2) how you treated missing data (if necessary)
3) how you grouped categorical data (if necessary)
4) what analyses you performed (you may need to be selective in what you present) - your description of the results should be sufficiently clear that the reader could reproduce your results
5) how you assessed the validity of your results
Your report has a word limit of 1500 including tables. Your report should include a maximum of four tables (no figures).