Ch. 9: Experimental Designs: Within-Subjects Design Flashcards
synonym of within-subjects design
repeated-measures design
goal of within-subjects design
Use a single group of participants and test or observe each individual in all of the different treatments being compared
how are treatments administered in within-subjects designs?
The treatments can be administered sequentially or all together
groups in treatment conditions in within-subjects designs
are equivalent to the group in every other condition
within-subjects designs in nonexperimental research
they are well-suited to nonexperimental research that investigates changes occurring over time
two main threats to internal validity of within-subjects experiments
confounding from environmental variables and time-related variables
types of confounding from time-related variables
- history
- maturation
- instrumentation
- regression toward the mean
- order effects
history
the environmental events other than the treatment that change over time and may affect the scores in one treatment differently than another treatment
maturation
any systematic changes in participants’ physiology or psychology that occur during the research study and affect the participants’ scores
when is maturation particularly a concern?
for young children or elderly adults
instrumentation
changes in a measuring instrument that occur over time
when is instrumentation particularly a concern?
with behavioural observation measures
regression toward the mean
the tendency for extreme scores on any measurement to move toward the mean when the measurement procedure is repeated
why does regression toward the mean occur?
because an individual’s score is a function of stable factors and unstable factors, which change substantially from one measurement to another
order effects
occur when the experience of being tested in one treatment condition has an influence on participants’ scores in later treatment conditions
types of order effects
- carry-over effects
- contrast effects
- progressive error
- fatigue effesct
- practice effects
carry-over effects
occur when one treatment condition produces a change in participants that affects their scores in subsequent treatment conditions
contrast effect
the subjective perception of a treatment condition is influenced by its contrast with the previous treatment
progressive error
refers to changes in participants’ behaviour or performance that are related to experience but not specific treatment
examples of progressive error
Practice effects and fatigue
fatigue effects
progressive decline in performance as a participant works through a series of treatment conditions
practice effects
progressive improvement in performance as a participant gains experience through the series of treatment conditions
how can environmental factors be controlled?
Randomization
Holding them constant
Matching across treatment conditions
how can treatment effects be controlled?
Controlling time
Switching to a between-subjects design
Counterbalancing
controlling time
- The possibility that a study will be affected by a time-related threat is directly related to the length of time required to complete the study
- Shortening the time between treatments can reduce the risk of time-related threats
- But, this can increase the likelihood that order effects will influence the results
switching to a between-subjects design
In some situations, order effects are so strong that a researcher wouldn’t even consider using a within-subjects design
counterbalancing
changing the order in which treatment conditions are administered from one participant to another so that the treatment conditions are matched concerning time
goal of counterbalancing
to use every possible order of treatments with an equal number of individuals participating in each sequence
purpose of counterbalancing
to eliminate the potential for confounding by disrupting any systematic relationship between the order of treatments and time-related factors
how is counterbalancing usually discussed?
in terms of order effects, but it has the same effect on time-related threats
limits of counterbalancing
- It does not eliminate order effects entirely
- It adds the order effects to some of the individuals within each treatment, but not to all of the individuals
- It assumes symmetry of order effects, which isn’t always justified
- It is necessary to present the treatments in every possible sequence to completely counterbalance
- As the number of treatments increases, counterbalancing becomes more complex
how is the number of different treatment conditions identified?
as n! (n factorial)
solution to complex counterbalancing
use partial counterbalancing
partial counterbalancing
uses enough different orderings to ensure that each treatment condition occurs first in the sequence for one group of participants, second for another group, third in another group, and so on
latin square
a simple and unbiased procedure for selecting sequences that involves creating an n x n matrix and filling it with letters
advantages of within-subjects designs
- It requires relatively few participants
- It eliminates all problems based on individual differences
- Reduces variance
- More statistically powerful: it reveals treatment effects that might not be apparent in a between-subjects design
disadvantages of within-subjects designs
- There is an opportunity for time-related factors to influence participants’ scores
- Participant attrition
participant attrition
some of the individuals who start the research study may be gone before the study is completed
three main factors that differentiate within- and between-subjects designs
individual differences
time-related differences
number of participants
individual differences and choosing within- or between-subjects design
if you anticipate large individual differences, it is better to use a within-subjects design
time-related factors and order effects and choosing within- or between-subjects design
if you expect one or more treatment condition(s) to have large and long-lasting effects that may influence participants in future conditions, it is better to use a between-subjects design
fewer participants and choosing within- or between-subjects design
whenever it is difficult to find or recruit participants, a within-subjects design is better
matched-subjects design
each individual in one group is matched with a participant in each of the other groups.
how is matching done in matched-subjects designs?
so that the matched individuals are equivalent concerning a variable that the researcher considers to be relevant to the study
goal of matched-subjects designs
to duplicate all advantages of within- and between-subjects designs without the disadvantages of either one
advantages of matched subjects
- There are no order effects
- It removes the variance caused by individual differences
disadvantage of matched-subjects designs
matching can become very difficult as the number of matched variables increases and the number of different groups increases
two main applications of within-subjects designs
Two-treatment designs
Multiple-treatment designs
two-treatment designs
Within-subjects help evaluate the difference between two treatment conditions
advantages of two-treatment designs
- It is easy to conduct
- Increases the likelihood of obtaining a significant difference
- It is easy to counterbalance
disadvantages of two-treatment designs
- It only provides two data points
- Does not provide any indication of the functional relationship between the independent and dependent variables
statistical analyses for two-treatment designs
- In cases with two treatment conditions, a repeated-measures t-test or a single-factor ANOVA (repeated measures) can determine the statistical significance
- If the data are on an ordinal scale, a Wilcoxon Signed-Ranks test can be used
directionality and two-treatment designs
- Occasionally, a within-subjects study only produces data that shows the direction of the difference between the two treatments
- In this situation, you can statistically evaluate the data using a sign test
advantage of multiple-treatment designs
You are more likely to reveal the functional relationship between the two variables being studied
disadvantages of multiple-treatment designs
- With too many conditions, the distinction between treatments may become too small to generate significant differences in behaviour
- Multiple treatments increase the amount of time required for each participant to complete the full series of treatments, increasing participant attrition
- Counterbalancing becomes more difficult
statistical analyses for multiple-treatment designs
With data on an interval or ratio scale, repeated-measures ANOVA is typically used
