Question 2. [11 marks] This question relates to the file diabet.sav.See the document "Description of datasets" on LMS for information on this dataset. Use SPSS (showing your output) and the file diabet.sav to produce requested output and answer the following questions.

1. [1 mark] Produce a scatterplot of duration of diabetes in years (DURATION) against glycosylated haemoglobin in mmol/L (HAEM) and comment on the relationship.
2. [2 marks] Produce a single table that shows the count and percentage of individuals who died within 5 years (DIED5) by eachduration grouping, where the duration grouping is:

DURATION 5 is short duration

5 < DURATION 10 is medium duration

DURATION > 10 is long duration.

Describe the relationship between duration group and DIED5. Note, you first need to create the duration grouping variable in SPSS.

1. [2 marks] Produce a single table that contains the mean of HEAM and PDW for each level of TREAT. Using this table, describe whether there is any evidence of associations between treatment and either HAEM or PDW.
2. [4 marks] Produce side by side box plots showing the distribution of DURATION for each level of TREAT and:
   1. Describe the distribution of DURATION for each level of TREAT noting any interesting features of these distributions.
   2. Provide a single sentence that summarises the relationship between DURATION and TREAT.
   3. In a single sentence provide an explanation of why this relationship could be reasonable.
3. [2 marks] Produce a single bar chart that shows the percent with DIED5=1 by TREAT and duration grouping (as created in (b)). Compare the difference in the proportion of individuals who die within 5 years across duration groups, separately for the three different treatments.