The Solomon four-group design
o Allows you to test for any possible sensitization of the pre test (works against carryover) Group 1 pretest treatment posttest Group 2 Pretest no treatment posttest Group 3 no pretest treatment posttest Group 4 no pretest no treatment posttest
A B A B desgign
A design which involves two different conditions:
o Baseline phase: asses behaviour in absence of the treatment
o Intervention phase: asses behaviour during application of the treatment
→Intersubject replication: When you replicate the experiment with other subjects which helps you to increase the external validity
Disadvantages of A B A B
- Drifting baselines: It can happen that it is impossible to stabilize a baseline so you have to subtract it out
- Unrecoverable Baselines: if baseline levels of performance can not be recovered during reversal (carryover effect)
- Unequal baseline between subjects: when subject react differently to a certain stimulus
- Inappropriate Baseline level: you can only increase the baseline decreasing is inappropriate
Factorial design
- Incorporates two or more independent variables in a single experiment
- Each variable is referred as a factor
- A two-factor design with two levels for A and three levels for B is labelled as 2 x 3 (three-factor design which each three levels would be labelled as 3 x 3 x 3)
Factorial within-subjects design
Each subjects exposed to every combination of all levels of all the factors
Higher-order factorial design
A factorial design including more than two independent variables
o The number of cells e.g. 2 x 2 with 5 participants would mean 4 x 5= 20 or e.g. you require 8 groups 2 x 2 x 2 with 5 participants in each group means 8 x 5= 40
→indicates the minimum of participants
o With tree factors you will get three main effects
→two-way interactions: A x B, A x C, and B x C
→three-way interaction: A x B x C (occurs when A x B interactions changes depending on the level of C)
Main Effect
When two have two lines in the graph
Interaction effect
when the lines are not parallel so they will cross at one time that is the interaction effect
Quasi-experimental designs
Resemble experimental designs but use quasi-independent variables rather than true independent variables
Time series designs
o Collect a series of observations of behaviour across time
o You make several observations (O1-4) before introducing the treatment and several after introducing (Q5-8)
Interrupted time series design
o You chart changes in behaviour as a function of some naturally occurring event (e.g. introduction of a new law) rather than manipulate an independent variable
o You observe the behaviour of the participants still before and after the event
Equivalent time sample design
o In this design you administer and withdraw a treatment repeatedly
o First the treatment is introduced than you observe after that you repeat the observation but without the treatment. You can repeat this sequence as often as necessary
o Most appropriate when the effects of the treatment are temporary or transient
Non-equivalent control group design
o You include a time series component along with a control group that is obviously not exposed to the treatment
o The control group is non-equivalent because it comes from a different community
Baseline design
- Focuses on the behaviour of a single subject both within and across the experimental treatments and does not rely on averaging to deal with uncontrolled variability
- Behavioural baseline: Within a treatment condition, the behaviour of interest is sampled repeatedly over time and plotted to create a behavioural baseline
- This is done until the baseline meets a stability criterion which means that the baseline has stabelized
