CausalComparative and Experimental Studies - PowerPoint PPT Presentation

1 / 28
About This Presentation
Title:

CausalComparative and Experimental Studies

Description:

Ex Post Facto Study: Study after the fact. ... Improved scores on a post-test as a result of taking the pre-test. ... Forms of pre-test and post-test are not the same. ... – PowerPoint PPT presentation

Number of Views:88
Avg rating:3.0/5.0
Slides: 29
Provided by: asn
Category:

less

Transcript and Presenter's Notes

Title: CausalComparative and Experimental Studies


1
Causal-Comparative andExperimental Studies
  • Dr. Barrick
  • ED 250

2
Definition of Causal-Comparative
  • Research that attempts to determine the cause or
    reason for pre-existing differences in groups or
    individuals
  • Ex Post Facto Study Study after the fact.
  • ExampleHaving your father/mother tutor in math
    at home may be a factor in higher achievement.

3
Purpose of Causal-Comparative Research
  • Attempts to identify a cause-effect relationship
    between variables correlational studies do not.
  • Causal-Comparative research involves comparison
    not relationship.
  • Individual students are not randomly assigned as
    in experimental studies.
  • There is no treatment as experimental studies.

4
Examples of Independent Variables in
Causal-Comparative Studies
  • Orgasmic Variables
  • Age
  • Sex
  • Ethnicity
  • Ability Variables
  • Intelligence
  • Scholastic Aptitude
  • Specific Aptitude
  • Perceptual Aptitude

5
More Variables
  • Personality
  • Anxiety
  • Introversion
  • Aggression
  • Self-concept
  • Self-esteem
  • Aspiration
  • Learning styles
  • Family
  • Income
  • SES status
  • Employment status
  • Marital status
  • Environment
  • Birth order
  • Number of siblings

6
More Variables
  • School
  • Preschool attendance
  • Size of school
  • Type of school
  • Per pupil expenditure
  • Type of curriculum
  • Leadership style
  • Teaching style
  • Peer pressure

7
Notes
  • Definition and selection of comparison groups is
    very important. Selection bias is a danger.
  • Lack of randomization, manipulation, and control
    are all sources of weakness.
  • There may be another variable which is the real
    cause of the observed difference between groups.

8
Ways to control
  • Matching
  • Homogeneous groups
  • Analysis of covariance

9
Experimental Studies
  • Dr. Stan Barrick
  • ED 250
  • Educational Research

10
Definition of Experimental
  • In an experimental study, the researcher
    manipulates at least one independent variable,
    controls other relevant variables, and observes
    the effect on one or more of the dependent
    variables.
  • Independent variable is called the experimental
    variable.

11
Steps in Experimental Study
  • Selection and definition of the problem.
  • Selection of the sample.
  • Selection of the instruments.
  • Selection of design.
  • Execution of procedures.
  • Analysis of data.
  • Formulation of conclusions.

12
Properties of Experimental Studies
  • Has at least one hypothesis that states an
    expected causal relationship between two
    treatment variables.
  • Randomly selects the samples for the two or more
    treatments.
  • Controls for external variables.
  • Has a control group and manipulates a variable.
  • Looks for significant differences due to the
    treatment.

13
Experimental Validity
  • Any uncontrolled extraneous variables that
    affect performance on the dependent variable are
    threats to the validity of an experiment.
  • Internal validity ensuring that the observed
    differences of the dependent variable are due to
    the independent variable.
  • External validity ensuring that the results are
    generalizable.

14
Threats to Internal Validity
  • History
  • Maturation
  • Testing
  • Instrumentation
  • Statistical regression
  • Differential selection of sample
  • Mortality
  • Selection-maturation interaction.

15
History
  • The occurrence of events that affect the outcome
    but are not part of the design.
  • The longer the experiment lasts the more likely
    that history will interfere.

16
Maturation
  • The natural physical, intellectual, and emotional
    changes that occur in the participants over a
    period of time and which affect the dependent
    variable.
  • Most likely to be a threat during the
    developmental stages of growth.

17
Testing
  • Improved scores on a post-test as a result of
    taking the pre-test.
  • Usually a threat when factual information is to
    be recalled.

18
Instrumentation
  • The lack of reliability and consistency in the
    measuring instruments.
  • Forms of pre-test and post-test are not the same.
  • Observations are not normed between observers.
  • Mechanical devices which collect data are not
    reliable.

19
Statistical Regression
  • Statistical regression occurs when a participant
    scores extremely high or low. On successive
    score will tend toward the mean.
  • Can occur when testing gifted or LH students on a
    standardized test.

20
Differential Selection
  • Usually occurs when preformed groups are
    compared.
  • Example- comparing the students at a community
    college with the students at a university.

21
Mortality
  • Attrition of participants affects the study.
  • Example- the study of a teaching method with one
    group having a high mobility rate.

22
Selection Interactions
  • Selection interacts with maturation.
  • Selection interacts with history.
  • Selection interacts with instrumentation.
  • Etc.

23
Threats to External Validity
  • Pre-test treatment interaction.
  • Selection treatment interaction.
  • Multiple treatment interaction.
  • Specificity of variables.
  • Treatment diffusion.
  • Experimenter effect. Pygmilan Effect
  • Reactive effect. Hawthorne Effect

24
Pre-Experimental Designs
  • One shot case X O
  • One group pre/post-test 0 X O
  • Static group comparison

25
Experimental Designs
  • Pre/Post test control group
  • Post test only control group

26
Experimental Designs
  • Solomon four group design

27
Quasi-Experimental Designs
  • Non-equivalent control group
  • Time Series

28
Quasi-Experimental
  • Counterbalanced
Write a Comment
User Comments (0)
About PowerShow.com