Non-experimental Designs
Scientific explanation in psychological experiments: An experiment is a controlled procedure
- set specific antecedent conditions
- manipulate and compare treatment conditions
- ensure equivalence across groups (similar groups in all aspects)
There are occasions when these requirements cannot be or should not be met
- certain characteristics of the participants may be of interest to us but we are unable to
manipulate them (e,g. gender, IQ)
- there are also times when we want to study something but cannot because of ethical reasons
(e.g. effect of damage to frontal lobe on moral decisions)
- some behaviors are better studied in a natural environment rather than in a manipulated and
controlled one
Non-experimental designs – used when an experimental design is not desirable or feasible
Non-experimental research designs:
1. Phenomenology
2. Case studies
3. Field studies
4. Archival studies
5. Qualitative research
6. Surveys and interviews
7. Correlation
8. Quasi-experimental designs
Dimensions for describing research designs:
1. Degree of manipulation of antecedent conditions
- how much we control conditions
- low manipulation: we just let things happen naturally
- high manipulation: we control the environment; environment is artificial
2. Degree of imposition of units
- how much we limit the responses of participants
- by limiting responses, data can be controlled
- low imposition of units: we do not limit the responses of the participants
- high imposition of units: we limit the responses of the participants
- middle: either the design is a field experiment or using a quasi-experimental design (quasi-
experimental designs are high in manipulation of antecedent conditions but is not considered as
a true experiment because some conditions cannot be addressed)
Dimensions for evaluating research designs: (determined by how much we control on our study)
1. Internal validity
- refers to the degree of our confidence that it is the antecedent conditions that produce the
effect on behavior and not some other factor
- there is a cause-and-effect relationship
- we are sure that the IV is responsible for the change in the DV
- based on antecedent condition
- low internal validity: we cannot establish causality
- high internal validity: we are confident that we can establish causality
Scientific explanation in psychological experiments: An experiment is a controlled procedure
- set specific antecedent conditions
- manipulate and compare treatment conditions
- ensure equivalence across groups (similar groups in all aspects)
There are occasions when these requirements cannot be or should not be met
- certain characteristics of the participants may be of interest to us but we are unable to
manipulate them (e,g. gender, IQ)
- there are also times when we want to study something but cannot because of ethical reasons
(e.g. effect of damage to frontal lobe on moral decisions)
- some behaviors are better studied in a natural environment rather than in a manipulated and
controlled one
Non-experimental designs – used when an experimental design is not desirable or feasible
Non-experimental research designs:
1. Phenomenology
2. Case studies
3. Field studies
4. Archival studies
5. Qualitative research
6. Surveys and interviews
7. Correlation
8. Quasi-experimental designs
Dimensions for describing research designs:
1. Degree of manipulation of antecedent conditions
- how much we control conditions
- low manipulation: we just let things happen naturally
- high manipulation: we control the environment; environment is artificial
2. Degree of imposition of units
- how much we limit the responses of participants
- by limiting responses, data can be controlled
- low imposition of units: we do not limit the responses of the participants
- high imposition of units: we limit the responses of the participants
- middle: either the design is a field experiment or using a quasi-experimental design (quasi-
experimental designs are high in manipulation of antecedent conditions but is not considered as
a true experiment because some conditions cannot be addressed)
Dimensions for evaluating research designs: (determined by how much we control on our study)
1. Internal validity
- refers to the degree of our confidence that it is the antecedent conditions that produce the
effect on behavior and not some other factor
- there is a cause-and-effect relationship
- we are sure that the IV is responsible for the change in the DV
- based on antecedent condition
- low internal validity: we cannot establish causality
- high internal validity: we are confident that we can establish causality
