CausalComparative Research Overview

1
Causal-Comparative Research Overview

• (Ex post facto Research
• Non-experimental research)

2
Causal-Comparative Research

• The aim of causal-comparative research is to
  determine the cause of existing differences among
  groups.
• Whereas correlational research involves
  collecting data on TWO or more variables on ONE
  group, causal comparative research involves the
  collection of data on ONE independent variables
  for TWO or more groups.

3
Causal-Comparative Research is Differentiated
from Experimental Reserarch

• In an experiment, the independent variable is
  manipulated by the researcher.
• In causal comparative research the independent
  has already occurred.
• Examples of independent variables include
  socioeconomic status, pre-school history, number
  of siblings, and so on.

4
Causal-Comparative DesignsSimilarities to
Experimental Designs

• Purpose
• Trying to determine cause-effect relation between
  variables
• Designs used
• Single-factor
• Two-factor
• Multi-factor
• Analysis of data

5
Causal-Comparative Designs vs Experimental Designs

• Start with effect, then seek causes
  (retrospective)
• Less often start with cause (prospective)
• No manipulation of variables
• Cannot be manipulated (SES, race, sex)
• Should not be manipulated ( cigarettes
  smoked/day)
• Were not manipulated (method of reading
  instruction)

6
Causal-Comparative Designs vs Experimental Designs

• Assignment of subjects to groups
• In experimental, assignment MUST be random
• In causal-comparative, assignment is based on
  preexisting characteristics
• Determination of cause is not as robust
• It is more that of a relationship, with a
  suggestion of cause

7
Causal Comparative Research

• Groups
• are classified according to common preexisting
  characteristic, and
• compared on some other measure
• There is NO
• intervention,
• manipulation, or
• random assignment

8
Example What causes lung cancer?

• Finding People with lung cancer smoke more than
  people without lung cancer. There are no other
  differences in lifestyle characteristics between
  the groups.
• Conclusion Smoking is a possible cause of lung
  cancer.
• Caution A third factor? Proper matching?

9
Value of Causal Comparative Research

• Uncovers relationships to be investigated
  experimentally.
• Used to establish cause-effect when experimental
  design not possible.
• Less expensive and time consuming than
  experimental research.
• Note if you conduct a quantitative research
  study it most likely will be a causal-comparative
  study.

10
Two Variations of Causal Comparative Studies

• IV presumed cause
• Groups formed on the basis of how much TV they
  watch, and compared on academic achievement
  (GPA).
• DV presumed effect
• Groups formed on the basis of gender, and
  compared on strength of career aspirations.

11
Two Variations (cont.)

• IV presumed effect
• Groups formed on the basis of whether they
  dropped out of high school, and compared on lack
  of mentoring relationship.
• DV presumed cause
• Groups formed on the basis of difficulty in
  learning to read, and compared on time parent
  spent reading to child.

12
Strengthening Causal Comparative Designs

• Strong inference (theory).
• Time sequence (presumed cause precedes presumed
  effect).
• Incorporate other, possible, causes in the design
  (measure common antecedents) .
• Use designs that control for extraneous causes
• matched group design
• Extreme groups design
• Statistical control (Analysis of Covariance)

13
More Examples of Causal Comparative Research

• A researcher measured the mathematical reasoning
  ability of young children who had enrolled in
  Montessori schools and compared the scores with a
  group of similar children who had not been to
  Montessori schools.
• A researcher measured the frequency of students
  misbehavior at schools which use corporal
  punishment and compared that to schools which did
  not use corporal punishment.

14
More Examples of Causal Comparative Research

• A researcher compared the high school dropout
  rate between students who had been retained (held
  back) in elementary school vs. similar students
  who had not been retained
• A researcher formed 3 groups of preschoolers
  those who never watched Sesame Street, those who
  watched it sometimes, and those who watched it
  frequently and then compared the 3 groups on a
  reading readiness test

15
Wide Variety of Statistical Procedures

• t tests, ANOVA, ANCOVA when two or more groups
  are being compared.
• Regression analysis when there are multiple
  independent variables.
• MANOVA, and multivariate regression, when there
  are multiple dependent variables.
• Path analysis and structural equation modeling
  when the theoretical causal paths are being
  investigated.

16
The END