How do experiments support causal claims?
They allow researchers to establish covariance, temporal precedence and internal validity
What designs are associated with experiments?
- Independent-groups designs - different participants are exposed to each level of the independent variable
- Within-groups designs - the same participants are exposed to all levels of the independent variable
What are the three potential threats to internal validity in an experiment?
- Design confounds - an experimenter’s mistake in designing the independent variable < only if unsystematically unvariable
- Selection effects - When the kinds of participants in one level of the independent variable are systematically different from those in the other < can be mitigated by random assignment or matched groups
- Order effects - Being exposed to one condition first changes how participants react to the later condition
How do you interrogate an experimental design?
- Construct validity - evaluating whether the variables were manipulated and measured in ways consistent with the theory
- External validity - whether the results can be generalised to other people or other situations
- Statistical validity - Effect size, precision and replication
- Internal validity - Involves looking for design confounds and seeing whether the researchers used techniques such as random assignment and counterbalancing.
What independent-groups designs are there?
- Posttest-only design - participants are randomly assigned to one of at least two levels of an independent variable and then measured once on the dependent variable.
- Pretest/posttest design - participants are randomly assigned to one of at least two levels of an independent variable and are then measured on a dependent variable twice—once before and once after
they experience the independent variable.
*Random assignments or matched groups can help establish internal validity in independent-groups designs by minimizing selection effects
What within-groups designs are there?
- Repeated-measures design - participants are tested on the dependent variable after each exposure to an independent variable condition
- Concurrent-measures design - participants are exposed to at least two levels of an independent variable at the same time and then indicate a preference for one level (the dependent variable).
*Within-groups designs allow researchers to treat each participant as his or her own control and require
fewer participants than independent-groups designs.
What are the advantages of within-groups design?
It ensures that the participants in the two groups will be equivalent and the researchers would not have to deal with selection effect, and requires less participants
What are the subgroups of order effects?
- Practice effects - where a long sequence might lead participants to get better at the task or to get tired or bored at the end
- Carryover effects - some form of contamination occurs from one condition to the next
How can you mitigate order effects?
By counterbalancing, where they present the levels of the independent variable to participants in different sequences. It can be full when exposing all of the independent variable or partial if not exposing all of the independent variable
What are the disadvantages of within groups design?
Within-groups designs present the potential for
order effects, permanent change and demand characteristics (a cue that can lead to a participants guessing a experiment’s hypothesis).
