Research Methods 2 /
Methodologie 2
Psychology
Vrije Universiteit Amsterdam
Jayne Brink
Jayne Brink
, Week 1: Lecture 1 + 2
Concept Explanation
Theory Set of high-level statements that
describes how constructs/variables
relate to each other based on previous
observations. So, it describes established
relationships based on existing
data/previous research. Often
conceptual.
Components:
- Assumptions
- Variables: conceptual and
operationalized constructs
- Propositions: relationships
between variables
Criteria for good theories:
- Logically consistent
- Falsifiable
- In agreement with data
- Clear (specific)
- Parsimonious
- Consistent with other theories
- Useful/predictive (fertile)
Conceptual construct Abstract, cannot be measured directly
Operationalized construct Concrete, can be measured directly
Model Theory but less abstract, quantifiable
Orthogonal Every one of them has an independent
influence
Hypotheses Describe hypothesized relationships that
have not yet been tested. Often
operationalized. *Hypotheses must
always be stated before the research is
carried out.
Typical Mediator analysis Multiple regression
Typical Moderator analysis ANOVA or multiple regression with a
categorical moderator
Internal validity Do the cause-effect relationships hold,
are there alternative explanations?
Construct validity Do the operationalizations capture the
construct?
Statistical validity Do the statistics support the claim?
External validity Are the results generalizable?
HARK-ing Hypothesizing after the results are
known (increase of type 1 error)
Pre-registration Involves writing down your experimental
design, your hypotheses, and your
Methodologie 2
Psychology
Vrije Universiteit Amsterdam
Jayne Brink
Jayne Brink
, Week 1: Lecture 1 + 2
Concept Explanation
Theory Set of high-level statements that
describes how constructs/variables
relate to each other based on previous
observations. So, it describes established
relationships based on existing
data/previous research. Often
conceptual.
Components:
- Assumptions
- Variables: conceptual and
operationalized constructs
- Propositions: relationships
between variables
Criteria for good theories:
- Logically consistent
- Falsifiable
- In agreement with data
- Clear (specific)
- Parsimonious
- Consistent with other theories
- Useful/predictive (fertile)
Conceptual construct Abstract, cannot be measured directly
Operationalized construct Concrete, can be measured directly
Model Theory but less abstract, quantifiable
Orthogonal Every one of them has an independent
influence
Hypotheses Describe hypothesized relationships that
have not yet been tested. Often
operationalized. *Hypotheses must
always be stated before the research is
carried out.
Typical Mediator analysis Multiple regression
Typical Moderator analysis ANOVA or multiple regression with a
categorical moderator
Internal validity Do the cause-effect relationships hold,
are there alternative explanations?
Construct validity Do the operationalizations capture the
construct?
Statistical validity Do the statistics support the claim?
External validity Are the results generalizable?
HARK-ing Hypothesizing after the results are
known (increase of type 1 error)
Pre-registration Involves writing down your experimental
design, your hypotheses, and your
