From StudyingMed

< BGDB‎ | Tutorials
Jump to: navigation, search

Get lecture slides

Your lit review

  • Research into your domain and a focus topic (a public health intervention) within it
  • Broaden your understanding of the domain
  • Given a chance to learn more about a specific area
  • What you need to do now - is re-focus on your research

Your data

  • Is it cleaned are all the variables ready to analyse?
  • Make sure you have described all the section B and C data that is relevant to the final submission
    • I.e. all the useful data re your participants
    • Use tables and charts (pie charts for small numbers of categories)
      • Histogram = continuous data
      • Bar charts = categorical data

Where are you going?

  • Final product = BGDB submission - a journal style submission article
  • Describe your data and testing hypotheses on your domain
  • Research the science behind your hypotheses
  • Interpret your data analysis results and discuss these findings in the light of your research - draw from both the science background and the SH project
    • Eg do people with less immunisation have more infectious disease?
  • Start off with a few research questions (do some brainstorming). Come up with questions that are interesting and relevant.
  • Find some good journal articles about infectious disease
    • Can use journal article referencing, tables etc, not APA

What should you do next?

  • Review your SH lit review submission and research any errors highlighted by your marker
  • Review the formative feedback on your test analyses
  • Focus on one/two interesting...


  • List all the tasks you need to do
  • Set some key dates for key stages - e.g. starting/finishing of the: analysis, interpretation, research, drafting, editing, submitting
  • Plan your group work carefully
  • Use the strengths of your team and support each other!


Feedback from the formative and advice for this stage

  • Some groups missed out the interpretation altogether!
  • Correlation done very well on the whole - just make sure you mention r (direction and magnitude) as well as r squared. A fitted line is required.
  • Forcing an unsuitable variable into a calculation or test (e.g. calculating the mean of diet type!)
  • Poor choice of test for t-test/chi square. Basically mixing up the type of variable for the test required.
  • Mixing up T-statistic and Levene's test results
  • Poor interpretation of chi square test results
  • Using terms incorrectly e.g. incidence, prevalence, rate, causal relationship (can't show incidence - not showing rates over time)
    • Use association often

Data analysis

  • New QMP portal = the UNSW ResMap is available to help with your research process.
  • Understanding your research question and/or hypotheses that you are developing is vital for your sensible actions here - we will discuss in class - but bring along some ideas re: research questions and hypotheses that you wish to test!
  • Also be aware of:
    • The data types that you have
    • The numbers of variables that you wish to analyse
    • And can you assume normality holds for your data?

New QMP website gadget

Types of data

  • Variable
    • Quantitative
      • Discrete
      • Continuous
    • Qualitative
      • Dichotomous
      • Nominal
      • Ordinal (strongly agree -> strongly disagree)

Analysis of the data

  • The instructions ask for at least 4 variables - which variables should you use?
    • The dataset has 162 variables and 533 cases
  • Look at your research question, think about this

Possibly hypothesis

  • Are immunisation rates higher now than then? Does having immunisations protect against key diseases?