From StudyingMed

< BGDB‎ | Lectures
Jump to: navigation, search


  • See the lecture slides
  • use terms "internal and external validity", "generalisatbility"
  • 1. selection bias 2. confounding factors 3. experimental problems 4. measurement of the outcomes 5. loss to follow up
  • measurement problems = consists of how you ask the questions in the questionnaire
    • e.g. diet didn't ask "did you overeat" etc
  • Selection bias can adversely effect your validity because the population you're testing on might not actually reflect the population you're interested in
    • Think about inclusion bias and who we should have excluded
    • What were our exclusion criteria and what could we have added?
  • For experimental trials, you need to randomise people twice and apply inclusion/exclusion criteria to minimise confounding factors
    • For our experiment, we're interested in the confounders as these have affected health over this period
    • Want to know about all the factors that might influence our exposure variables