Group Comparison Research
CAUSAL-COMPARATIVE
(EX POST FACTO)
RESEARCH
BY:
DR RAKHI SAWLANI
Purpose
 Attempts to determine cause for
 Existing conditions
 Preexisting differences in groups
 Alleged cause and effect have already occurred
 Orientations
 Retrospective (basic): starts with an effect and seeks
possible causes
 Prospective (variation): starts with a cause and
investigates its effects on some variable
Features
 The independent variable (IV) is not
manipulated; it has already occurred
 Independent variables sometimes
called “attribute variables”
 Less costly and time-consuming to
conduct
 Establishing cause-effect relationships
is more difficult than in experiments
Procedures
 Identify an existing condition /event
 (e.g., differences in socialization among
1st grade students)
 Look “backwards” to see what may
have caused this difference /condition
to occur
 (i.e., some attended preschool, some
did not)
 Rule out other causal factors
Sometimes confused with
correlational research:
 Both lack manipulation of variables
 Both require caution in interpreting
results
 Both can support subsequent
experimental research
Causal comparative vs.
Correlational research
 Causal comparative
 Attempts to identify
cause-effect
relationships
 At least one
independent variable
 Two or more groups
 Involves a comparison
 Correlational
 No attempt to
understand cause and
effect
 Two or more variables
 Only one group
Sometimes confused with
experimental research:
 Both try to establish cause-effect
relationships
 Both can test hypotheses concerning the
relationship between an independent (X)
and a dependent variable (Y)
 Both involve group comparisons
Comparison to experiments
 Causal comparative
 Individuals already in
groups before study
begins
 Independent variable
has already occurred
 Independent variable
is not manipulated
 Cannot be
 Should not be
 Could be, but is not
 Experiment
 Individuals
randomly assigned
to groups (e.g.,
treatment or
control)
 Independent
variable
manipulated by
the researcher
Examples of non-manipulated
independent variables
 Age
 Sex
 Ethnicity
 “Learning style”
 Socioeconomic status (SES)
 Parent educational level
 Family environment
 Type of school attended
Design of causal-comparative
research
 Select 2 groups that differ on some IV
 One group possesses a characteristic that the
other does not
 Each group possesses the characteristic, but in
differing amounts
 Randomly sample Ss from each group
 Collect info on Ss to determine equality of
the groups
 Compare groups on the DV
Difficulty in interpreting
findings
 Establishing cause and effect requires
caution!
 Alternative explanations:
Different causal variable
Order of causation
Reverse causality
Order of occurrence
Evidence necessary to
demonstrate that X causes Y:
 Establish statistical relationship
between X and Y (i.e., correlational
research);
 determine that X precedes Y in time
(collect data over time, i.e.,
longitudinal research);
 demonstrate that other, unknown
factors did not determine the
dependent variable (i.e.,
experimental research).
Becker & Gersten (1982): “Effects of
Project Follow-Through…”
 Quasi-experimental study
 Ex post facto study
 Problem: Are the two groups in
this study comparable to one
another?
In order to make sure that the two groups are
comparable, and to ensure that the only post-test
differences between the groups are due to the
independent variable (the Follow-Through
intervention), data were obtained on students’:
family income
gender
language spoken in home
mother’s education
ethnicity
number of siblings.
Research Design
 FOLLOW-THROUGH
 Year 1 (1975) Gr 5
 Site 1
 Site 2
 Site 3
 Site 4
 Site 5
 Year 2 (1976) Gr 6
 Site 1
 Site 2
 Site 3
 Site 4
 Site 5
 NO FOLLOW-THROUGH
 Year 1 (1975) Gr 5
 Site 1
 Site 2
 Site 3
 Site 4
 Site 5
 Year 2 (1976) Gr 6
 Site 1
 Site 2
 Site 3
 Site 4
 Site 5
Dependent variables
 Wide-Range Achievement Test (WRAT)
reading
mathematics
 Metropolitan Achievement Test (MAT)
reading
mathematics
RESULTS
 A total of 180 comparisons of FT to
No-FT students. Of these, only 56
(31%) favored FT students!
 Largest differences between FT and
No-FT students were in basic skills
areas.
 FT students’ achievement declined
by grades 5 and 6 (2-3 years after
end of FT).
Critique of this research
 What are the strengths of the study?
 Groups are comparable to one another.
 Contrasted statistical with practical significance.
 Large sample size.
 Multiple “replications” of treatment effect.
 What are the weaknesses of the study?
 Lack of random assignment.
 Focus on standardized test performance.

causalcomp-DrRakhiS.ppt

  • 1.
    Group Comparison Research CAUSAL-COMPARATIVE (EXPOST FACTO) RESEARCH BY: DR RAKHI SAWLANI
  • 2.
    Purpose  Attempts todetermine cause for  Existing conditions  Preexisting differences in groups  Alleged cause and effect have already occurred  Orientations  Retrospective (basic): starts with an effect and seeks possible causes  Prospective (variation): starts with a cause and investigates its effects on some variable
  • 3.
    Features  The independentvariable (IV) is not manipulated; it has already occurred  Independent variables sometimes called “attribute variables”  Less costly and time-consuming to conduct  Establishing cause-effect relationships is more difficult than in experiments
  • 4.
    Procedures  Identify anexisting condition /event  (e.g., differences in socialization among 1st grade students)  Look “backwards” to see what may have caused this difference /condition to occur  (i.e., some attended preschool, some did not)  Rule out other causal factors
  • 6.
    Sometimes confused with correlationalresearch:  Both lack manipulation of variables  Both require caution in interpreting results  Both can support subsequent experimental research
  • 7.
    Causal comparative vs. Correlationalresearch  Causal comparative  Attempts to identify cause-effect relationships  At least one independent variable  Two or more groups  Involves a comparison  Correlational  No attempt to understand cause and effect  Two or more variables  Only one group
  • 8.
    Sometimes confused with experimentalresearch:  Both try to establish cause-effect relationships  Both can test hypotheses concerning the relationship between an independent (X) and a dependent variable (Y)  Both involve group comparisons
  • 9.
    Comparison to experiments Causal comparative  Individuals already in groups before study begins  Independent variable has already occurred  Independent variable is not manipulated  Cannot be  Should not be  Could be, but is not  Experiment  Individuals randomly assigned to groups (e.g., treatment or control)  Independent variable manipulated by the researcher
  • 10.
    Examples of non-manipulated independentvariables  Age  Sex  Ethnicity  “Learning style”  Socioeconomic status (SES)  Parent educational level  Family environment  Type of school attended
  • 11.
    Design of causal-comparative research Select 2 groups that differ on some IV  One group possesses a characteristic that the other does not  Each group possesses the characteristic, but in differing amounts  Randomly sample Ss from each group  Collect info on Ss to determine equality of the groups  Compare groups on the DV
  • 12.
    Difficulty in interpreting findings Establishing cause and effect requires caution!  Alternative explanations: Different causal variable Order of causation Reverse causality Order of occurrence
  • 13.
    Evidence necessary to demonstratethat X causes Y:  Establish statistical relationship between X and Y (i.e., correlational research);  determine that X precedes Y in time (collect data over time, i.e., longitudinal research);  demonstrate that other, unknown factors did not determine the dependent variable (i.e., experimental research).
  • 14.
    Becker & Gersten(1982): “Effects of Project Follow-Through…”  Quasi-experimental study  Ex post facto study  Problem: Are the two groups in this study comparable to one another?
  • 15.
    In order tomake sure that the two groups are comparable, and to ensure that the only post-test differences between the groups are due to the independent variable (the Follow-Through intervention), data were obtained on students’: family income gender language spoken in home mother’s education ethnicity number of siblings.
  • 16.
    Research Design  FOLLOW-THROUGH Year 1 (1975) Gr 5  Site 1  Site 2  Site 3  Site 4  Site 5  Year 2 (1976) Gr 6  Site 1  Site 2  Site 3  Site 4  Site 5  NO FOLLOW-THROUGH  Year 1 (1975) Gr 5  Site 1  Site 2  Site 3  Site 4  Site 5  Year 2 (1976) Gr 6  Site 1  Site 2  Site 3  Site 4  Site 5
  • 17.
    Dependent variables  Wide-RangeAchievement Test (WRAT) reading mathematics  Metropolitan Achievement Test (MAT) reading mathematics
  • 18.
    RESULTS  A totalof 180 comparisons of FT to No-FT students. Of these, only 56 (31%) favored FT students!  Largest differences between FT and No-FT students were in basic skills areas.  FT students’ achievement declined by grades 5 and 6 (2-3 years after end of FT).
  • 19.
    Critique of thisresearch  What are the strengths of the study?  Groups are comparable to one another.  Contrasted statistical with practical significance.  Large sample size.  Multiple “replications” of treatment effect.  What are the weaknesses of the study?  Lack of random assignment.  Focus on standardized test performance.