Please enable JavaScript.

Coggle requires JavaScript to display documents.

BRM RESIT (Ethics (Nüremberg Code -> research with humans (Informed…

- - - - assent
      - vulnerable population
      - informed consent
      - voluntary participation
  - - - confidentiality
      - de-identification
      - Anonymity
- - - - predictive v.
      - concurrent v.
      - convergent
      - discriminant
    - - Content v.
      - Face v.
    - - Interactions by variables outside the construct (e.g. the program).
      - Mono-method bias: only using a single method of measurement (only relevant to your measures/observations, not the independent variable(s))
      - Mono-operation bias: relying on a single implementation of your independent variable.
      - Labeling issue: label isn’t a good description for what you implemented
      - Inadequate pre-operational explication of constructs: not defining the construct well enough.
      - Restricted generalizability across constructs (poor EV).
  - - - Source of noise
        
        poor reliability of treatment implementation
        
        random irrelevancies in the setting
        
        low reliability of measure
        
        random heterogeneity of respondents
        
        violated assumptions of statistical tests
  - - - Non-probability
        
        expert s.
        
        quota s.
        
        modal instances sampling
        
        heterogeneity s.
        
        purposive smiling
        
        snowball s.
        
        accidental/ haphazard or convince sampling
      - Probability
        
        systematic random
        
        cluster area random
        
        stratified random
        
        multistage random
        
        simple random
    - - systematic error
      - random error
      - wrongly generalising
- - - - pretest-posttest program-comparison-group quasi-experimental design
    - - pre-post two-group quasi-experimental design
    - - The pattern-matching NEDV design
  - - - covariance design
      - (randomised) blocking design
      - hybrid design
        
        switching replication design
    - - factorial design
        
        fully crossed factorial design
        
        incomplete factorial design