Quantitative Research

1
Quantitative Research
Criteria
Experimental Designs

• Adequate experimental control
• Lack of artificiality
• Basis for comparison
• Adequate information from the data
• Uncontaminated data
• No confounding of relevant variables
• Representativeness
• Parsimony

• Pretest-posttest control group design
• Posttest only control group design
• Solomon four group design
• Counterbalanced design
• Factorial design

2
Criteria 1Adequate Experimental Control

• There are adequate restraints, or controls, on
  the conditions of the experiment.
• With controls in place the researcher should be
  able to accurately interpret the results.
• The experimental design should be structured in
  such a way that the experimental variable (or
  treatment) effect can be detected.

3
Criteria 2Lack of Artificiality

• The experiment must be conducted in such a way
  that the results will apply to the real
  educational world.

4
Criteria 3Basis for Comparison

• There must be some way to make a comparison to
  determine whether or not there is an experimental
  effect.
• In some cases a control group is used (control
  groups do not receive the experimental
  treatment).
• In some cases multiple treatments may be studied
  against an external criterion (i.e. Different
  reading techniques may be examined to assess
  their relative impact on student reading scores).

5
Criteria 4Adequate Information From the Data

• There must be adequate data for testing the
  hypothesis.
• The data must be such that statistical testing
  will be precise enough for the researcher to make
  a decision about the hypothesis.

6
Criteria 5Uncontaminated Data

• The data should adequately reflect the
  experimental effects.
• The data should not be affected by poor
  measurement or errors in the experimental
  procedures.
• Example subjects in an experiment should not
  communicate across groups (control and
  experimental).

7
Criteria 6No Confounding Variables

• Confounding of variables occur when two or more
  variables cannot be separated.
• Experimental controls should be such that
  variables can be isolated.
• Translation the relationship between independent
  variable and the dependant variable can be
  measured with confidence in that another variable
  is not contributing to the effect.

8
Criteria 7Parsimony

• Keep it simple!
• This criterion means that, with all other factors
  being equal, a simple design is preferred to a
  more complex one.
• Note simple designs are usually easier to
  implement, and may yield data that is easier to
  interpret.

9
QuantitativeExperimental Designs
This family of research designs is characterized
by random selection of subjects. The
following experimental designs are based on the
problem statement The effects of teaching
method(s) on history achievement.
10
Experimental Design 1Pretest-Posttest Control
Group Design
Diagram
O O
O O
X1 X
11
Experimental Design 2Posttest-Only Control
Group Design
Diagram
O O
X1 X
12
Experimental Design 3Solomon Four Group Design
Diagram
O O
X1 X
O O
O O
X1 X
13
Experimental Design 4Counterbalanced Design
Diagram
X1 X
O O
O O
X X1
14
Experimental Design 5Factorial Design
Diagram
Independent Variable 1 Teaching Method
Lecture Only Reading/Lecture/Small Groups
Randomly assigned 3rd graders scoring above 85 on an aptitude test. Randomly assigned 3rd graders scoring about 85 on an aptitude test.
Randomly assigned 3rd graders scoring below 60 on an aptitude test. Randomly assigned 3rd graders scoring below 60 on an aptitude test.
Independent Variable 2 Aptitude