The Scientific Method

The Scientific Method

1. steps

   • pick topic
     • read
       • concepts
         • theories
           • operationalize
             • hypothesis
               • research design
                 • gather data
                   • draw conclusions

2. variables

   • operationalization
     • move from abstract concept to variable
   • types
     • nominal
       • many options
         • can't be ordered
       • political parties
     • indicator
       • categorical
       • yes/no
     • ordinal
       • can be ordered
       • differences aren't meaningful
       • strongly agree - strongly disagree
     • interval
       • data can be ordered
       • differences are meaningful
       • GDP
   • independent variable
     • X
     • cause
   • dependent variable
     • Y
     • effect

3. causal mechanism

   • links phenomena to variable

4. correlation

   • reverse causality
     • is it possible Y causes X
     • temporal order
   • spurious relationships
     • relationship that only appears to exist
   • confounding variables
     • omitted variable bias
       • Bias = B + (A x D)
         • + x + & - x -
           • overestimate
         • - x +
           • underestimate
       • only will affect relationship if impacts X
     • need to be controlled
       • alters relationship between X and Y
   • positive if same direction
     • negative if opposite

5. literature review

   • summarize knowlege
   • find gaps

6. ethics

   • participants
     • confidentiality
       • anonymity
         • protect data
     • informed consent
       • nature of research
         • use of findings
           • risks
       • barriers
         • age
           • disability
         • social pressure
     • right to withdraw
       • can't be any barriers
         • includes disincentives
     • welfare
       • stigma on participants
   • researchers
     • personal safety
     • legal action
     • conflict of interest
   • colleagues
     • plagiarism

7. sampling

   • population
     • group being studied
       • location
     • unit of analysis
       • what is studied
     • time period
   • samples
     • subsection of population
       • make generalizations using sample
   • probability
     • number of ways event can happen
       • divided by possible events
       • frequentist approach
         • P(A) = [N(A)]/n
           • # of times event occurs
   • representativeness
     • sampling frame
       • list of population
     • selection method
       • must all be equally likely to be selected
     • sample size
       • n
         • greater n = greater accuracy
   • sampling error
     • standard error of mean
       • SEM = SE/√n
         • n
           • observations
             • higher means more accurate
       • SE
         • variation in data
         • lower SE means more accurate
   • non random
     • accidental
       • self selection
         • harder to reach populations
     • purposive
     • snowball