Produce a publication quality bar chart of mean FEV1 values, indicating “error” bars of your
choice. The plot should have separate bars for each combination of GENDER and GROUP.
[5%]
ii. Produce an alternative, publication quality plot of your choice showing the distribution of
FEV1 for each of the four combinations of GENDER and GROUP.
[5%]
- Justify your choice of plot for 6.ii. and compare and contrast the plots from 6.i. and 6.ii. [5%]
- Carry out a suitable statistical test to compare FEV1 values between the two groups of
participants (GROUP), ignoring the other variables. Also compare BMI values between the
patient groups (GROUP) using suitable statistical methods. Present only the important
parts of these two analyses in a combined, publication quality, table and interpret the
results.
[15%] - Assess the relationship between FEV1CAT and GROUP. Is there a difference in the
proportion ofsubjects with poorlung function between the two groups? Comment on your
findings, including the precision of the estimated difference.
[10%] - Produce and briefly justify a suitable regression model with FEV1 as the dependent
variable and BMI as the independent variable. State the model equation. Present the
important parts of the results in a single table. Check the model assumptions. Interpret the
model and your findings.
[10%] - Write a statistical methods section that describes the statistical analyses conducted for this
assessment. This should include descriptive statistics, statistical tests, statistical software
and general statistical principals used. This should be a maximum of 250 words.
[5%] - Comment on the overall findings from the various analyses that you have run in the
previous questions.
[5%]
Presentation (including spelling, grammar and layout of tables and figures) will be taken
into consideration.
[5%]
The content and presentation of your syntax file will also be assessed. [5%]
You will be marked on the 12 tasks listed above (plus presentation and syntax file). The content of the
report should not exceed 2,000 words. The assignment should include figures and tables as requested or
necessary for illustration, which should be appropriately referenced. Figures, tables and references do not
count towards the total word count. Your use of SPSS will be evident from the SPSS data and syntaxfiles.
