PSYC-1115: Applied Epidemiology & Statistics in
the Global Context
1) Introduction to the assessment
This
assessment involves analysis and interpretation of a public health related data
set, and the preparation of workbook in a format of a technical report.
Choosing and applying statistical tests to a dataset provided and interpreting
the output will increase students’ familiarity with statistical methods and
their critical interpretive skills.
This assessment
has he following aims:
• To
increase your understanding of statistical techniques by applying them to data
• To build
your confidence in using statistical software
• To gain
skills in sourcing and retrieving health and epidemiological information
• To help
you acquire the skills necessary to write and present a technical report/
workbook
2) Background
Task: Each
student will be provided with a dataset and will analyse the data, interpret
the results and prepare a technical report/workbook of the findings.
The data are
derived from the Health Survey for England 2003 dataset. Student will be
provided with different versions of the dataset; thus, study findings may
differ slightly between students.
Because we
are not using the complete survey database and some changes have been made to the
description of the methodology to simplify comprehension of the data, we will
refer to the country as Pinkland instead of England for the purpose of this
assessment.
3)
Instructions
You will be provided with a dataset
(Pinkland.SAV) in SPSS format. Make sure you can open the database in SPSS.
Analysing the data
It is for you to decide what are the appropriate methods using the knowledge
you’ve learned from the material studied in this module.
Analysis:
•
Descriptive statistics & descriptive
epidemiology of the sample & main outcome variable (BMI)
•
Inferential statistics & analytical
epidemiology (association between BMI and other variables)
Write an
extended technical summary about your findings. At a minimum, the
workbook/technical report should provide estimates for men and women of the
prevalence of overweight and obesity and identify which population groups are
most at risk.
You must
carefully consider which of the output from SPSS is necessary to include in the
workbook/ technical report. Do NOT cut and paste tables directly from the SPSS
output files without deleting superfluous text and figures. Please edit the
charts to make them reader friendly.
Round values
for your data to no more than four significant figures. (For the same number of
significant figures, different variables will have different numbers of decimal
places because they are measured using different units. For example, mean Z
scores may have three decimal places, while mean weight in kg might only have
one or two.) Also, except for very small p-values, values of most test
statistics should be rounded to two decimal places.
Please remember:
For the purposes of
writing your report/workbook, the data are from Pinkland even though you
know that the data are actually from England.
Completing the assessment
While you
may wish to discuss ideas with other students about how to analyse the data, it is absolutely essential that you write up
your results individually. Working together on your written work is
considered a form of cheating and is an assessment offence.
You will
most likely choose to analyse your data and present your findings in different
ways from your classmates and there is no single correct approach.
a)
Minimum recommended process for
data analysis and for reporting of the findings
You should always
clearly state the objectives of your analyses. For example, ‘A paired t-test
was performed to assess the mean difference in x between the two sets of observations.’
Descriptive statistics
·
Summarise
the demographic characteristics of the sample in terms of age, sex, ethnicity,
and marital status. You could treat age as a continuous variable, and/or group
it into appropriate categories.
·
You
can also summarise variables such as car ownership, family size and limiting
longstanding illness.
·
Create
a new continuous variable, BMI, from the values of weight and height. Remember
BMI is measured as weight in kilograms (kg) divided by the square of height in
metres (m). The units for BMI are kg/m2.
·
Summarise
the data relating to BMI. You will need to include measures of location (or
central tendency) and measures of spread (variation) and to report confidence
intervals.
·
In
order to give prevalence rates of overweight and obesity, create a new
categorical variable from BMI using the threshold values of 18.5, 25 and 30 as
follows:
o
BMI <18.5
= underweight
o
BMI from
18.5 to 24.99 = normal weight
o
BMI ≥
25.00 = overweight
o
BMI ≥ 30.00
= obese
·
If the
data allows you can further classify the obese group into: Obese class 1 (BMI from 30 to 34.99); Obese
class 2 (BMI from 35.00 to 39.99); Obese class 3 (BMI ≥ 40.00).
Inferential
statistics and analytical epidemiology
·
Investigate
if and how BMI is associated with age, sex, and educational attainment.
·
You
can choose whether you use BMI as a continuous variable or as the derived
categorical variable.
·
You
can choose how to use age (continuous or categorical) and educational
attainment (if you prefer to condense education into fewer categories or use the
number of categories originally defined).
b) Additional Analysis (Optional)
Investigate the association of BMI with ethnicity, car ownership, occupation
and presence of long-standing illness. You can also look at the association
between presence of long-standing illness and ethnicity and check if the burden
of disease is distributed evenly across ethnic groups or not. Again, you are
free to re-group the variables. For example, ethnicity could be treated as a
binary variable (white vs non-white).
c) Report writing: Guidance on the
style of a technical report/ workbook
Technical/ executive summaries are briefing documents written by
technical experts on specific topics for decision-makers (often civil servants,
programme managers or administrators.) They are commissioned to provide
information on specific questions or issues and to provide a basis for
decision- making and action. As such they should put forward all the relevant
facts and set out the relevant issues. The aim is to inform the reader
sufficiently to enable her/him to understand the reasons for and implications
of any decisions and subsequent actions that she/he takes. Information presented
in the report should not include personal views that are not supported
by the data or by other evidence/literature. You should assume that the person
you are writing for is intelligent and proficient, but busy, and not an expert
in relation to the issue in hand. The report should include a short background,
aim(s) of the report, key results a discussion and conclusions.
Below are suggestions of content that
should be covered in each section:
Introduction
·
Why
overweight/obesity is an important issue in this particular country. (You can
use data and evidence from England and the UK to support this section.)
·
Why
this survey is needed
Objectives
·
Clear
statement of the aims of the report
Methods
·
Brief
description of data collection and sampling procedure
·
Important
features of the study design and quality control
·
Description
of how the variables used in the analysis were defined
·
Description
of the analysis plan for descriptive and inferential statistics (for
categorical and continuous variables) and of the software used for analyses
Results
·
Description
of the sample (e.g. age distribution, gender, socio-economic status, demographics,
etc), descriptive statistics for the nutritional variables (BMI) and for other
health outcomes used in the analysis
·
Presentation
of the results of analytical analyses (associations between BMI and other
variables).
·
At
least one table and one graph
Discussion
and conclusions
·
Compare
your findings to the WHO values (for example) which indicate a crisis and to
other relevant literature on the topic
·
Suggested
reasons for the patterns and trends in the data based on the research and
literature
·
Identify
any limitations of your data analysis and the survey methodology
·
Identify
the need for intervention to address the problems identified and make recommendations
References
·
Keep a
list of all references in the Harvard format. Preferably use appropriate
software for this.
d)
Length
(max 2500 words +/- 10%)
The maximum length of the workbook is
not limited to the text only, this includes tables and figures/charts, but excludes
the reference list. To simplify the length estimation, each table (regardless
of the size) will count as 100 words and each graph as 50 words.
For example, if you include 4 tables (4
x 100 = 400) and 2 graphs (2 x 50 =100) in your workbook you have used 500
words. The remaining 2000 will be distributed as text across the workbook.
IMPORTANT: The reference list is NOT
included in the word count.
Marking criteria
Please see
the attached rubric for details of the marking criteria and grading scale.
The criteria
for passing this assessment include:
Data
analysis and interpretation (Accounts for 75% of the mark)
•
Use of appropriate tests
•
Appropriate use and display of tables and
graphs
•
Presentation of results (key prevalence rates,
identification of high-risk groups and main associations)
•
Adequate interpretation of key results
•
Reasons for the patterns and trends in the data
•
Explanations clear and understandable
•
Degree of synthesis / creative thought
demonstrated
•
Limitations of survey methodology and of your
data
•
Use of statistical software package (SPSS) to
conduct data analyses and present results.
Academic writing and referencing (Accounts for 25% of mark)
•
Clarity and logical organisation of the
report/workbook
•
The style of the text is clear, simple,
concise, logical and systematic
•
Page style / font /
margins appropriate
•
Reference list and in text references
consistent
•
References using Harvard style
•
Reference list complete and without errors
•
Supplementary items cross referenced and
appropriate
•
Appropriate text explaining tables and graphs
•
Clear English with coherent flow and correct
grammar
•
Appropriate length
Submission
The
assessment must be submitted electronically on Monday the 11th of
January 2021 by 11:30 pm GMT using the Coursework Submission section on Moodle.
You can use the Originality Report section on Moodle to check the originality
of your assignment before submitting.
