Purpose
Epidemiology seeks to identify patterns of health and risk/protective fa
Purpose
Epidemiology seeks to identify patterns of health and risk/protective factors for health in populations. Biostatistics provides the tools to allow us to analyse and interpret quantitative data to achieve these goals, and, in particular, to quantify the uncertainty that arises from inferring results from samples to populations.
The purpose of this assignment is to apply biostatistical methods and principles to a dataset to demonstrate you can:
Understand and describe fundamental biostatistical concepts
Identify, select and apply appropriate biostatistical tests (using statistical software) based on characteristics of the data and the research question
Interpret the output of biostatistical tests
Dataset
The data for this assignment is a modified dataset from the Youth Risk Behavior Surveillance System (YRBS) from the Centers for Disease Control and Prevention in the United States. This is an epidemiological surveillance system designed to monitor the prevalence of youth behaviours that most influence health. The YRBS is a repeat cross-sectional survey that uses a representative sample of high school students (public, private, and Catholic schools) from years 9 to 12. The dataset provided contains all of the information needed to complete the analysis for the assignment. You are not required to find additional references. The assignment is based on all content in this unit, but the focus is on the last half of the unit. You should use the topic guides, lecture recordings, computer practical materials, prescribed readings, and any additional resources provided to assist with your assignment. The data files are provided below.
Assignment Questions
Obesity and Weight Management among High School Students (70 marks total)
You are working as a Health Promotion Officer for a US State government and have been approached by the State School Board who want to understand more about how obesity and weight management practices are changing over time and some of the factors associated with obesity in US High School students. They ask you to undertake some exploratory analysis to provide the basis for further research and program development. You have access to a large dataset based on several surveys of students over a twenty year period with questions related to the issue of obesity and weight management. Note that some items were only added to the survey in the 2022 round (water consumption and physical activity).
Using the dataset provided, and using SPSS, answer the following questions (10 marks per question):
Was student weight different in 2002 and 2012? (note – do check and report on the distribution of weight, but for the purpose of the assignment, you can assume it is normally distributed)
Are their differences in how students perceive their weight in students from 2002, 2012, and 2022?
Was body mass index (BMI) different in students from 2012 and 2022 compared to those from 2002? (note – do check and report on the distribution of BMI, but for the purpose of the assignment, you can assume it is normally distributed)
[Please answer the following four questions using data from the 2022 cohort only ]
Are water consumption and weight correlated? (note – do check and report on the distribution of these variables, but for the purpose of the assignment, you can assume they are both normally distributed)
Is there a difference in the percentages of male and female students who are attempting to lose weight in 2022?
Is there an association between physical activity and BMI? Answer this question using a method that could also allow you to take confounding factors into account (note – this comment is just to help you choose the right method – you don’t need to actually adjust for confounding factors to answer this question).
Is water consumption associated with being classified as overweight (note, the term “overweight” here includes both overweight and obesity – see here for information on the classification of overweight and obesity using BMI)? Answer this question using a method that could also allow you to take confounding factors into account (this comment is just to help you choose the right method – you don’t need to actually adjust for confounding factors to answer this question).
Please note: For this assignment, you are only required to answer these 7 questions (i.e. no introductory section or closing discussion required). Also, please be aware that this dataset has been modified for this assignment, so your findings may not reflect the results of data analysis using the actual dataset.
Format
For each question, the following is required:
a) What test/analysis did you select? Why was it appropriate? (2 marks)
Think about the purpose of the test and how this matches the question
What were the requirements and assumptions of the test (this may include things like the type of variable, distribution of the variable etc.).
b) Which variables did you select and if relevant, which is the independent and dependent (outcome) variable? (1 mark)
c) Describe any data manipulation. Did you recode any variables? Did you select cases? Did you need to create new variables? Explain what you did and why. (2 marks)
d) Provide the results of the analysis. (2 marks)
Include the output of the tests you run including relevant tables, figures, or graphs. Provide these following the rules for displaying data (i.e. all appropriate elements included, only relevant information included)
e) Interpret the results. (3 marks)
Summarise the results, noting if a difference or association was found, the direction of the difference or association, and whether it was statistically significant. You should report p-values and confidence intervals correctly, to the most appropriate number of decimal places (see lecture notes and in particular the seminar answer sheets for examples of how to present values of statistical tests). Please also comment on the practical importance of the findings (and what considerations you have taken into account in commenting on this). Think about the considerations that need to be taken into account when deciding if something is a) statistically significant, and b) of public health importance.
The word limit for AT3 is 1500 words but there is no problem if your assignment is shorter than this.
Important: To show your working, you must include all the relevant statistical analysis output that you produced in SPSS to answer each question. Do not include any information in your output file that is not relevant to answering the questions.
Failure to include your output will result in a mark of zero for the assessment.
