lecture 4 - causality & RD Flashcards

1
Q

necessary conditions causality

A

Stuart Mill:

  1. covariance
  2. temporal order
  3. temporal and spatial contiguity/link
  4. nonspurious relationship
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

two types/names of causality

A
  1. deterministic/nomothetic causality
  2. probablistic
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

intervening and anteceding factors

A

intervening = come between cause and effect

  • mediators: factors that come between IV and DV
  • moderators: factors that affect the relationship between IV and DV
    *suppresor vs distorter vs reinforcer

anteceding = alternative/correct DV

  • confound = completely explains observed relationship
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

different types of relationships

A
  1. zero-order relationship
  2. partial relationship: part of the relationship is direct, other part goes through the mediator
  3. spurious relationship / correlation = both IV and DV are caused by the confound, the real explanatory factor
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Research Design

A

= strategy/blueprint to investigate the research question

includes:

  1. type of research
  2. what data to collect
  3. method of data collection
  4. mode of analysis
  5. how threats to internal and external validity and measurement validity and reliability are minimized
    - internal validity = causal relationship without confounding
    - external validity = generalizability to other cases
    - measurement reliability = consistency and replicability of results
    - measurement validity = does it measure what it claims to measure?
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

types of research design

A
  1. experimental design = by definition randomized assignment treatment and control (-> ruling out confounding variable)
    - lab experiment
    - field experiment
    - survey experiment
  2. cross-sectional and longitudinal designs
    - cross-sectional = multiple cases in one point of time
    - longitudinal = longer periods of time (panel vs cohort designs)
  3. comparative design
  4. historical design
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

methods of data collection

A

data collection requires: clear boundaries of space/place, time, units/actors, factors/variables

  • questionnaires/surveys
  • interviews
  • participant observation / ethnographic research
  • textual/content/discourse analysis
  • statistical data
How well did you know this?
1
Not at all
2
3
4
5
Perfectly