BetweenSubjects and WithinSubjects Experimental Design

1
Between-Subjects and Within-Subjects Experimental
Design

• Types of Experimental Design, Error Variance,
  Betwee-Ss, Within-Ss, Carryover Effects

2
Experimental Designs

• Between-Ss Design
• Different groups of Ss randomly assigned to
  levels of IV
• Scores are averaged and analyzed between levels
• Within-Ss Design
• Single group of Ss is exposed to all levels of IV
• Scores are averaged and analyzed between levels
• Single Ss Design
• Single group of Ss is exposed to all levels of IV
• Scores are individually analyzed between levels

3
Error Variance

• Statistical variability of scores caused by the
  influence of VARs other than Ivs
• subject-related variance or extraneous variables

4
Distribution Curves
5
Error Variance and Normal Curve
6
As variance increases
7
Distribution curves for 2 groups

• Two groups under two conditions with two
  different averages for the DV

8
Minimizing Error Variance

• Why is this necessary?
• Controlling for extraneous VARs as much as
  possible.
• Increase strength of IV
• get as much info about options to manipulate IV
  before conducting study
• Random assignment to different levels of IV
  (i.e., different conditions of exp.)
• Using more powerful statistical tools and/or
  using more conservative criteria for statistical
  significance.

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
47. Single-Factor Randomized Groups Designs

• Subjects are randomly assigned to treatment
  groups.
• Two groups (EXPERIMENTAL and CONTROL) are needed
  to constitute an experiment.
• The TWO-GROUP DESIGN is the simplest experiment
  to conduct, but the amount of information yielded
  may be limited.
• Additional levels of the independent variable can
  be added to form a MULTIGROUP DESIGN.
• If different levels of the independent variable
  represent quantitative differences, the design is
  a PARAMETRIC DESIGN.
• If different levels of the independent variable
  represent qualitative differences, the design is
  a NONPARAMETRIC DESIGN.

13
Conducting a Two-Group Matched Groups Experiment

• Obtain a sample of subjects.
• Measure the subjects for a certain characteristic
  (e.g., intelligence) that you feel may relate to
  the dependent variable.
• Match the subjects according to the
  characteristic (e.g., pair subjects with similar
  intelligence test scores) to form pairs of
  similar subjects.
• Randomly assign one subject from each pair of
  subjects to the control group and the other to
  the experimental group.
• Carry out the experiment in the same manner as a
  randomized group experiment.

14
(No Transcript)
15
Sources of Carryover

• Learning
• Learning a task in the first treatment may affect
  performance in the second.
• Fatigue
• Fatigue from earlier treatments may affect
  performance in later treatments.
• Habituation
• Repeated exposure to a stimulus may lead to
  unresponsiveness to that stimulus.
• Sensitization
• Exposure to a stimulus may make a subject respond
  more strongly to another.
• Contrast
• Subjects may compare treatments, which may affect
  behavior.
• Adaptation
• If a subject undergoes adaptation (e. g., dark
  adaptation), then earlier results may differ from
  later ones.

16
Dealing With Carryover Effects

• Counterbalancing
• The various treatments are presented in a
  different order for different subjects. May be
  complete or partial.
• The Latin Square Design for partial
  counterbalancing
• Used when you make the number of treatment orders
  equal to the number of treatments.
• Taking steps to minimize carryover
• Techniques such as pretraining, practice
  sessions, or rest periods between treatments can
  reduce some forms of carryover.
• Make treatment order an independent variable
• Allows you to measure the size of carryover
  effects, which can be taken into account in
  future experiments.

17
When to Use a Within-Subjects Design

• A within-subjects design may be best when
• Subject variables are correlated with the
  dependent variable.
• It is important to economize on participants or
  subjects.
• When you want to assess the effects of increasing
  exposure on behavior.

18
Within-Subjects Versus Matched Groups Design

• Both try to manage error variance associated with
  correlation between subject variables(SV) and
  DV.
• If you think carry over effects are serious, use
  matched groups design, even though it may be less
  economical.
• If correlation between SV and DV is weak, a
  randomized groups design would be more powerful
  than either a within-subjects design or a matched
  groups design.

19
Single factor vs. Multiple factor designs

• Single factor design (one IV)
• two level design simplest for within-Ss desgin
• multilevel design more than two levels of a
  single IV
• Ex Peterson and Peterson (1959) study on the
  effect of retention interval (3, 6, 9, 12, 15, 18
  seconds) on memory of three-consonant trigrams
  (e.g., JHK)
• Can be between Ss but more economical if
  Within-Ss
• If Within-Ss, requires counterbalancing.
• Is this a parametric or nonparamentric design?
• Multiple factor design or factorial design (more
  than one IV)
• Ex Peterson and Peterson (1959), adding one more
  factor (e.g., the time trigrams are displayed 2
  second 5 seconds)
• This is a 6 X 2 factorial design.

20
Factorial Designs

• Adding a second independent variable to a
  single-factor design results in a FACTORIAL
  DESIGN.
• The MAIN EFFECT of each independent variable and
  the INTERACTION between independent variables can
  be evaluated.
• Main effects are the separate effects of each
  independent variable. They are analogous to
  separate experiments involving those variables.
• An interaction exists if the effect of one
  independent variable changes over levels of a
  second.
• As factors are added, the complexity of the
  experimental design increases, with increasing
  numbers of possible main effects and interactions
  (higher order factorial designs).

21
(No Transcript)
22
(No Transcript)
23
Combining Experimental and Correlational design

• If you think a variable other than the IV is
  correlated with the DV, you can measure that
  variable to partial it out from the effects of
  the IV on the DV.
• The correlational variable is called a covariate.
• Ex In the Peterson Peterson study, you may
  want to partial out intelligence (by measuring
  this with an IQ test) to learn the impact varying
  lengths of intervals between exposure and recall
  have on the memory retention.
• Intelligence in this case is a quasi-independent
  variable.
• A quasi-independent variable is a correlational
  variable that resembles an independent variable.

24
Quasi-Experimental Research

• Using Quasi-Independent Variable (e.g.,
  demographic variables)
• Cannot make causal inferences.
• Time Series Design
• Make several observations of behavior before and
  after introducing your independent variable.
• Interrupted Time Series Design
• Make several observations before and after some
  naturally occurring event.
• Equivalent Time Samples Design
• Repeatedly introduce the treatment condition,
  alternated with periods of observation without
  the treatment.
• Ex studying operant conditioning with a variable
  ratio schedule.
• Nonequivalent Control Group Design
• Include a time series component and a control
  group that is not exposed to the independent
  variable.

25
(No Transcript)
26
(No Transcript)
27
Developmental Studies

• Changes in behavior as a function of ones
  chronological development.
• Three ways to study development
• Cross-sectional design (Between Ss)
• generation effect is a concern
• Longitudanal design (Within Ss)
• cross-generation effect is a concern
• Cohort-Sequential design (Mixed)
• To account for generation effect and
  cross-generation effect.

28
(No Transcript)
29
(No Transcript)
30
(No Transcript)