Q uestion 2

USE SPSS to answer this question and include key parts of your SPSS output in your answer. You must also explain your answer clearly and you are limited to a maximum of 5 sides .

In the provision of health care there is considerable concern that quality standards across the country should be in line with current best practice. It is therefore seen as undesirable, and a cause for further investigation, if certain parts of the country or particular hospitals are found to perform less well than others. One measure of quality is the number of unforeseen ‘complications’ that arise when patients are operated upon.

A health services monitoring organisation is therefore interested in estimating the number of complications it should expect from hospitals, given the numbers and types of patients they operate on. Data has been collected for a particular speciality in 100 hospitals, recording the number of complications arising last year, the numbers of adult patients and the numbers of child patients. The data is stored as ‘complics’, ‘adults’ and ‘children’ in file Complications3.sav, also on the web board.

1. Plot graphs that you think shed some light on what relationships might exist between number of complications and the numbers of adult and child patients.
2. Use linear regression to determine which, if either, of adults and children show evidence of a significant linear relationship with the number of complications at the hospitals. Justify your answer.
3. Comment on the apparent quality of the two models in part (b). Justify your answer. Carry out a common-sense based examination of the residuals for the two models and comment on your findings.
4. Using your preferred model advise the health services monitoring organisation about the number of complications to expect in an average hospital with 1200 adult patients and 400 child patients and in an individual hospital with these numbers of patients. Comment on the level of accuracy of your predictions.
5. In a moment of inspiration someone suggests that the number of complications at the hospital might be best explained by the total numbers of patients (i.e. adults plus children). Investigate this possibility and suggest a practical explanation of your findings.

1. Explain clearly how your preferred model could be used by the health services monitoring organisation to identify hospitals which seem to have unusually high or low complication rates, given their numbers of the types of patient.

Price: $16.59

Solution: The downloadable solution consists of 8 pages, 859 words and 16 charts.
Deliverable: Word Document