Background
This assessment task uses the Longitudinal Survey of ACU Student Health and Wellbeing, a (fictional) cohort study involving a survey administered to students on enrolment with ACU (first year baseline) and selected follow up data (during third year). For this task, you are presented with baseline data, along with selected follow-up data (where indicated) on:
1. Road traffic accidents
2. Depression and obesity
For Assessment Task 1, we will be considering all data.
Note: For Assessment Task 2, it will be necessary for you to select either 1. Road traffic accidents OR 2. Depression and obesity to “write up” and interpret. By completing Assessment Task 1, you will therefore be preparing for Assessment Task 2.
Assessment specifications
Due date 20 April 2018 (at 10:00am)
Weighting 40%
Length and/or format Annotated SPSS output plus summary (800 words).
Purpose To apply statistical theory and use various methods of statistical analysis to answer specific research questions relevant to public health.
Learning outcomes 1, 2, 3, 4
Submission Turnitin (via the PUBH620 LEO page). Upload a Word file containing your answers, interpretations, tables etc. You may cut and paste from the SPSS file as necessary (ensure that you also provide interpretation and other information where required)
Feedback Marks and feedback will be provided via LEO
Assessment criteria See marking rubric below
Datasets
The dataset is called ‘AT1 dataset 2018’.
Instructions
Please provide answers for each of the following six questions:
1. In your SPSS file, age (at time of enrolment) is given as a continuous variable.
a. What is the mean, standard deviation, minimum and maximum values for age?
b. Some students enrol before their 18th birthday, potentially confounding this analysis of longterm driving behaviour (for the RTA study) or generally creating ethical issues around recruitment of children and consent (for the depression/obesity study). It is therefore necessary to consider only those aged 18 and over, and to make data for those under 18 years of age “missing”. Please recode “age” into a new variable (named “age_category”) using this guide:
New value Description
0 Students aged 18 years at time of enrolment
1 Students aged 19 to 21 at time of enrolment
2 Students aged 22 to 25 at time of enrolment
3 Students aged 26 or more at time of enrolment
Hint: all other values ( 18) should be designated as “system missing” in the new variable
What is the frequency (% and number) of students in each of these new age categories?
2. Create table(s) of descriptive statistics for the demographic information of the students in this dataset (sample size n, mean and % values where appropriate). The demographic variables are cohort, state, age, gender living arrangement, faculty, degree type, metro, study mode and fee status. As a guide you may find the tables in the paper by Keijzers et al. (2011) helpful when presenting your demographic information. Then, provide a brief description of your findings that informs the reader about the ACU cohort.
3. Test if the mean for aggression, thrill seeking and risk acceptance scores differ for the following demographics and provide a brief interpretation of your findings:
a. Gender
b. Metropolitan background status
c. Study mode
d. RTA in past 12 months (from follow-up survey)
4. In terms of depression, is there a difference between the depressed and not depressed groups according to their:
a. Gender
b. Metropolitan background status
c. Study mode
d. Fee status
Provide a brief interpretation of your findings.
5. Using “RTA_one_crash” as the dependent variable, perform a binary logistic regression using the following predictor variables relating to the RTA study:
Type Variables
Demographic Age (using newly-created variable)
Gender
Living arrangements (enter as categorical variable) Domestic/international status
Driving Driving distance (average per week)
Behaviour Aggression
Thrill-seeking
Risk acceptance
a. Provide a table of odds ratios, confidence intervals and significance values for each predictor variable.
b. How do you interpret the findings for living arrangements?
c. Provide a preliminary interpretation and conclusion about the influence of the predictor variables on road traffic accidents. In particular, what is the effect of changes (increase or decrease) in “driver aggression” on the odds of experiencing a road traffic accident?
6. Using “OB” (obesity at third-year follow-up) as the dependent variable, perform a binary logistic regression using the following predictor variables:
Type Variables
Demographic Age (using newly-created variable)
Gender
Living arrangements (enter as categorical variable)
Baseline characteristics Overweight or obese at baseline
Depression at baseline
Parental factors Number of parents with university education
Presence/absence of obese parents
a. Provide a table of odds ratios, confidence intervals and significance values for each predictor variable.
b. Again, how do you interpret the findings for living arrangements?
c. Provide a preliminary interpretation and conclusion about the influence of the predictor variables on obesity at follow-up. In particular, what is the effect of changes (increase or decrease) in the number of parents with university education on the likelihood of later obesity?
General Variables
Name AGE
Definition Participant’s age at time of enrolment at ACU
Values NA – continuous data
Notes and related variables Age derived from date of birth (not shown).
Name GENDER
Definition Student gender
Values 0 Male
1 Female
Notes and related variables
Name LIVING_ARRANGE
Definition Self-reported initial living arrangements for students during first year of study
Values 1 At home
2 College or other student residential environment
3 Independently
Notes and related variables Self-reported data obtained from T0 survey information. Definitions:
• At home: students living in their family home.
• College or other residential environment: students living in specific ACU (or other University) managed accommodation (“colleges”) or in specific residential facilities operated by private companies exclusively for students (“student accommodation”).
• Independently: students living in private accommodation under standard rental arrangements or share-houses (not exclusively for students).
Name FACULTY
Definition Faculty student enrolled in
Values 1 Arts and Sciences
2 Education
3 Health Sciences
4 Theology and Philosophy
5 Business
Notes and related variables Data derived from original degree course student admitted to at time of enrolment.
For double degree students studying degrees from different faculties, the Faculty of the first-named degree is considered (e.g. Bachelor of Nursing/Bachelor of Business Administration = Faculty of Health Sciences).
Note 1: Faculty of Theology and Philosophy established in 2009; no data available prior to this date (courses previously available from Faculty of Arts and Sciences).
Note 2: Faculty of Business established in 2010; no data available prior to this date (courses previously available from Faculty of Arts and Sciences).
Name METRO
Definition Geographic background of students prior to enrolment
Values 1 Metropolitan
2 Non-metropolitan
Notes and related variables Data available for domestic students only.
Location (metropolitan or non-metropolitan) derived from address of students prior to enrolment using postcode assignment to either metropolitan or non-metropolitan areas. Does not reflect geographic location during study.
Name DEGREE_TYPE
Definition Type of degree undertaken by student
Values 0 Single degree
1 Double degree
Notes and related variables Data derived from original degree course student admitted to at time of enrolment
Name STUDY_MODE
Definition Full or part time study mode during first year
Values 0 Full time
1 Part time
Notes and related variables Data derived from original enrolment information
Name FEE_STATUS
Definition Status as domestic or international student
Values 0 Domestic student
1 International student
Notes and related variables Data derived from original enrolment information
1: RTA variables
Name driver_agg
Definition Driver aggression score
Values 0 to 15 (continuous scale)
Notes and related variables Scale and description as per example paper (Donovan scale of driving behaviour)
Values: 0 = low aggression; 15 = high aggression
Name thrill
Definition Thrill seeking behaviour score
Values 0 to 7 (continuous scale)
Notes and related variables Scale and description as per example paper (Donovan scale of driving behaviour)
Values: 0 = low thrill seeking behaviour; 7 = high thrill seeking behaviour
Name risk_accep
Definition Risk acceptance behaviour score
Values 0 to 16 (continuous scale)
Notes and related variables Scale and description as per example paper (Donovan scale of driving behaviour)
Values: 0 = low risk acceptance; 16 = high risk acceptance
Name dist_driving
Definition Self-reported average number of kilometres driven by students in a week (from follow-up survey conducted during third year)
Values 0 Less than 10 km per week
1 More than 10 km per week
Notes and related variables Information on average distance driven by student per week as measure of road use
Name RTA_one_crash
Definition Students who self-reported experiencing one or more road traffic accidents (as a driver) in the past year (from follow-up survey conducted during third year)
Values 0 No RTA
1 Experienced one (or more) RTAs
2: Depression and obesity variables
Name BL_owob
Definition Overweight or obese – status at baseline (first year)
Values 0 Normal weight or underweight
1 Overweight or obese
Notes and related variables Data derived from self-reported height and weight collected at baseline (enrolment in first year)
Name edu_par
Definition Number of parents with university education
Values 0 to 2 (continuous scale)
Notes and related variables Data from self-report during first year; indicates the number of parents (0, 1 or 2) with a university education (undergraduate or postgraduate)
Name owob_par
Definition Presence or absence of obese parent(s)
Values 0 No obese parents
1 At least one obese parent
Notes and related variables Data from self-report during first year; indicates whether parents are obese (based on simple classification by student, not parental height and weight)
Name depression
Definition Self-reported depression status during first year
Values 0 Not depressed
1 Depressed
Notes and related variables Data from baseline survey during first year
Name OB
Definition Obesity – status at follow-up (third year)
Values 0 Not obese
1 Obese
Notes and related variables Data derived from self-reported height and weight collected at follow-up (during third year)
Referencing
As this assessment task involves SPSS output and interpretation there will be no referencing required.
Turnitin: Turnitin is a tool used to assist in the detection of referencing problems and/or plagiarism. Turnitin generates a similarity index for a document: that is, what percentage of the document contains material that is matched to accessible sources. Presence of similarity does not necessarily indicate plagiarism: there are many reasons why similar text is discovered in student documents. Turnitin often classifies reference lists themselves as “similar”—this is similarity, but not plagiarism.
NOTE THAT IN STATISTICAL PROJECTS IT IS COMMON FOR TURNITIN TO IDENTIFY SIMILAR TEXT SO DO NOT WORRY IF TURNITIN IDENTIFIES HIGH SIMILARITY IN THIS ASSESSMENT TASK.
Marking rubric
In line with section 5.1 of ACU’s Assessment Policy, all assessment marking and grading must be criterion-referenced and use standards-based grading. Assessment criteria and standards are related to unit learning outcomes. Student performance on a task is evaluated against each criterion, and according to the set standards of achievement for that criterion.
Assessment criteria and standards for this task are provided in the following rubric. Each criterion is marked according to a five-point standard, from “poor” to “excellent”, with a descriptor for each standard. Within each standard there is a small marking range that further differentiates your final mark for the task.
Relevant PUBH620 learning outcome: Assessment task 1
3. Distinguish between different statistical tests, especially in terms of application and interpretation
4. Develop a sound statistical approach to the analysis and interpretation of public health data and communicate findings in an academic-standard output
5. Critique public health research on the basis of its statistical methods, analysis and interpretation
PUBH620: Assessment Task 1 marking rubric
Marking criteria and relevant unit learning outcome(s) Standard achieved
Excellent Very good Good Fair Poor
1. Demonstrated knowledge of statistical terminology, concepts and tests
Analysis demonstrates depth of knowledge of statistical concepts and terminology, and application of statistical tests to dataset Analysis demonstrates excellent depth of understanding of statistical terminology, concepts and tests Analysis demonstrates very good depth of knowledge of statistical terminology, concepts and tests Analysis demonstrates a satisfactory knowledge of statistical terminology, concepts and tests Analysis demonstrates limited knowledge of statistical terminology, concepts and tests Analysis demonstrates inadequate (or no) knowledge of statistical terminology, concepts and tests
LO1: demonstrate knowledge of statistical concepts 2 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
2. Use of descriptive statistics in analysis
Appropriate use of descriptive statistics to analyse the dataset
(particularly in relation to
graphical/tabular representations of data) Analysis applies descriptive statistics to the dataset in an entirely appropriate manner
(particularly in relation to graphical/tabular representations of data) Analysis applies descriptive statistics to the dataset in an appropriate manner
(particularly in relation to graphical/tabular representations of data) Analysis applies descriptive statistics to the dataset in a satisfactory manner
(particularly in relation to graphical/tabular representations of data) Analysis applies descriptive statistics to the dataset in an inadequate manner
(particularly in relation to graphical/tabular representations of data) Analysis applies descriptive statistics to the dataset in either a very limited manner, or not at all
LO3: perform statistical tests and interpret results 1 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
3. Use of inferential statistics in analysis
Appropriate use of inferential statistical tests to analyse the dataset Analysis applies inferential statistics to the dataset in an entirely appropriate manner Analysis applies inferential statistics to the dataset in an appropriate manner Analysis applies inferential statistics to the dataset in a satisfactory manner Analysis applies inferential statistics to the dataset in an inadequate manner Analysis applies inferential statistics to the dataset in either a very limited manner, or not at all
LO3: perform statistical tests and interpret results 1 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
4. Interpretation of results Appropriate interpretation of statistical tests and results of the analysis
Interpretation of results is extremely sound, and clearly communicated in a way that is easy to understand Interpretation of results is sound, and clearly communicated in a way that is easy to understand Interpretation of results is sound, but with some deficiencies Interpretation of results is not unsound and/or has numerous errors Very little (or no) interpretation of results
evident; and/or
interpretation communicated poorly
LO2: distinguish between statistical tests; LO3: perform statistical tests and interpret results 2 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
PUBH620: Assessment Task 1 marking rubric
Marking criteria and relevant unit learning outcome(s) Standard achieved
Excellent Very good Good Fair Poor
5. Overall conclusion(s) based on analysis
Appropriate discussion of results and conclusions drawn relative to the questions posed Conclusion(s) are detailed, sound, and directly supported by the statistical analysis. No irrelevant or inconsequential conclusion(s) are reported. Conclusion(s) are sound and directly supported by the statistical analysis; however, some points require clarification.
Conclusions are largely relevant. Conclusion(s) are sound and generally supported by the statistical analysis; however, several points require clarification. Some conclusions irrelevant or inconsequential. Conclusion(s) are adequate, but key elements are omitted or inaccurately interpreted and/or some conclusions not supported by statistical analysis Conclusion(s) are very limited (or absent) and fail to address key elements or are completely unsupported by the statistical analysis
LO3: perform statistical tests and interpret results 2 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
6. Coherence and presentation Overall coherence of discussion and presentation of results; correct use of terminology, statistical conventions etc Overall discussion and presentation of results is excellent; use of terminology is correct and appropriate Overall discussion and presentation of results is good; use of terminology is correct and appropriate Overall discussion and presentation of results is satisfactory; use of terminology is generally correct and appropriate but requires improvement in parts Overall discussion and presentation of results is inadequate in parts and/or some terminology use is inappropriate Overall discussion and presentation of results is very limited (or absent) and/or serious errors in the use of terminology
LO4: sound approach to analysis and interpretation 2 (4½–5 marks) (3½–4 marks) (2½–3 marks) (1½–2 marks) (0–1 marks)
Total marks available: 50 (LO1 = 10 marks; LO2 = 5 marks; LO3 = 25 marks; LO4 = 10 marks) Unit weighting: 40%