STA10003 FOUNDATIONS OF STATISTICS
Scenario
You are a new graduate researcher at a health sciences research institute. You have been given a
dataset which contains the data of 5000 Melbourne adults who participated in a health survey, and
asked to analyse the data to answer several questions of interest that are presented on the following
pages.
Data Preparation
For the purpose of your report, you must use SPSS to draw a random sample of 1500 cases of the
5000 observations. You will conduct your analysis on this sample of 1500. You should ideally use
the same 1500 cases you used for Assignment Part 1. However, you can draw a new random sample of
1500 cases if you need to from STA10003 Sem 2 2021 Assignment Data.sav.
Submission Instructions
• Your submission must be a single Word file or PDF file.
• Although a cover page is not required, you should include your name and student number within
the document [e.g., in footer].
• You must submit your file via the Canvas link by Monday 11 October at 11:59 pm
Australian Eastern Standard Time (AEST). Only the last document you submit will be
marked.
• Once submitted, please review your submission to ensure the correct file has been submitted.
• This is an individual assignment. Do not share your work with other students. They will have a
different random sample of data, so any copying will be detected.
For your Assignment Part 2, you are required to complete the first three questions by
producing the appropriate analyses using SPSS and writing the relevant report for each
question. Reports are to follow the style used in the examples. You are also required to
complete question 4, which contains short answer questions.
For each of the first three questions requiring SPSS, you should include the relevant output
immediately following your report.
Question 1
Researchers predicted that the average serves of vegetables and legumes/beans for Melbourne
adults is now higher than the average of three (3) serves recorded in 2018. Conduct a one-sample t-
test using the vegleg1n variable to test this prediction.
Produce the relevant SPSS output and write a one-sample t-test report based on your output. See
Supplement G. Include the relevant output with your answer.
Question 2a
Based on previous research, the investigators have also predicted that the average total time at work
in the last week (mins) is lower for males than for females. Conduct an independent samples t- test
using the totwkmin and sex variables to test this claim.
Produce the relevant output and write an independent samples t-test report based on your output.
See Supplement H. Include the relevant output with your answer.
Question 2b
Check and comment on the assumptions of the independent samples t-test. Include the relevant
output with your answer.
Question 3
The researchers have also hypothesised that the average number of serves of refined or lower fibre
cereals/grains is different to the average number of serves of wholegrain or high fibre
cereal/grains. Conduct a paired samples t-test using the wholgr1n and refgra1n variables to test
this claim.
Produce the relevant output and write a paired samples t-test report based on your output. See
Supplement J. Include the relevant output with your answer.
Question 3b
Check and comment on the assumptions of the paired samples t-test. Include the relevant output
with your answer.
Question 4
A researcher wants to investigate the percentage of public service employees who are currently
working more than 40 hours per week. The researcher hypothesises that more than 21% of public
service employees are currently working more than 40 hours per week. A random sample of public
service employees is obtained and each person is asked if they are currently working more than 40
hours per week (‘yes’ or ‘no’).
(a) What type of statistical test would be appropriate to investigate the hypothesis?
(b) Explain why this statistical test would be appropriate.
(c) The appropriate statistical test was conducted and a p-value of 0.143 was obtained. Based on
the results of this study, it was concluded that exactly 21% of public service employees
currently work more than 40 hours per week. Is this conclusion valid or not?
(d) Explain why the researcher’s conclusion is valid / not valid.
