Non-Experimental Observational Designs Pt1

• • Many naturally occurring variables can’t be manipulated
  • Many variables shouldn’t be manipulated (ethics)
  • The research question does not require manipulation of variables (descriptive/exploratory)
  • Experiments are simply not feasible
• Researchers might want to:

  • DESCRIBE THE CHARACTERISTICS OF A CERTAIN GROUP
  • DRAW COMPARISONS BETWEEN DIFFERENT GROUPS
  • ANALYSE RELATIONSHIPS BETWEEN NATURALLY OCCURRING VARIABLES(CORRELATIONAL DESIGN)
  • INVESTIGATE THE INFLUENCE OF VARIABLES THAT CANNOT BE EXPERIMENTALLY MANIPULATED
  • INVESTIGATE THE NATURAL COURSE OF A CONDITION
  • INVESTIGATE RISK FACTORS FOR A CERTAIN OUTCOME
• Hierarchy of study design:

  • (from lower to higher hierarchy) generate hypotheses:
  • observational studies
  • cohort
  • case control
  • cross-sectional
  • case-report
  • cohort
  • (from lower to higher hierarchy) establish causality:
  • experimental & quasi-experimental studies
  • controlled clinical trials
  • randomised controlled trials (RCTs)
  • systematic reviews (+/- meta-analysis)
  • establish causality is higher up the hierarchy than generate hypotheses
• • observational studies can be categorised into 2 main types:
  • descriptive:
  • Describe the sample characteristics, behaviours, conditions
  • Exploratory/Analytic:
  • Systematically investigate relationships between variables
  • Hypothesis driven
  • Correlational design
• Correlation designs:

  • Analysis of relationships:
  • correlation research to understand:
  • Whether patterns exist within data
  • Whether and how variables are relate
• Correlation:

  • Variables often co-vary
  • A variable rises or lowers in relation to the behaviour of another variable
  • These variables might be described as “related” or “associated” with one another
  • We measure these associations statistically
• • correlation does not equal causation
• Different types of observational study design:

  • all observation study designs start with a study population
  • if we took data from a study population in the present it would be called cross-sectional
  • if we took a study population in the present and continued taking this data into the future, it would be called prospective
  • if we took data a study population in the present and looked back into their past, it would be called retrospective
  • retrospective and prospective studies are both classified as longitudinal studies
• Basic model of observational studies:

  • includes observing the relationship between exposure and outcome i.e. how exposure (risk factors, determinant, treatment) lead to outcome (disease, mortality, health)
  • aim is to measure (quantify) the relationship between exposure and outcome
  • cannot determine whether causal link - need experimental design