cs comp&experiemental2011 12
DESCRIPTION
CauseandComparative AND ExperimentalTRANSCRIPT
Educational Research: Causal-Comparative&
Experimental Studies
ELT-718 Research Methods
Asst. Prof. Dr. Hasan BEDİR
Research...
The systematic application of a family of methods employed to provide trustworthy information about problems
…an ongoing process based on many accumulated understandings and explanations that, when taken together lead to generalizations about problems and the development of theories
The basic steps of research...
Scientific and disciplined inquiry is an orderly process, involving:
description and execution of procedures to collection information (“methodmethod”)
objective data analysisanalysis statement of findings (“resultsresults”)
recognition and identification of a topic to be studied (“problemproblem”)
Research methods...
QuantitativeQuantitative……collects and analyzes numerical datadata
obtained from formal instrumentsinstruments
Quantitative methods...
descriptive research (“survey research”) correlational research causal-comparative research (“ex
post facto research”) experimental research
Research MethodologiesA continuum rather than “either/or”
Qualitative Goal: To Understand,
Predict Descriptive accounts Similarities and
Contrasts Applied and Theoretical Research Questions Field study
Natural conditions
Quantitative Goal: To Predict and Control
Measure and Evaluate Generalize to population,
reproduction Basic and Theoretical Hypothesis testing Emprical study
Controlled, contrived
Data Collection
Quantitative Emphasis on numerical data, measurable variables Data is collected under controlled conditions in order to
rule out the possibility that variables other than the one under study can account for the relationships identified
Qualitative Emphasis on observation and interpretation. Data are collected within the context of their natural
occurrence.
Causal-Comparative Research
The Purpose
Purpose of explaining educational phenomena through the study of cause-and-effect relationships. The presumed cause is called the independent variable and the presumed effect is called the dependent variable. Designs where the researcher does not manipulate the independent variable are called ex post facto research.
Causal-Comparative Research (Continued)
Causal-Comparative research is also a type of non-experimental investigation in which researchers seek to identify cause-effect relationships by forming groups of individuals in whom the independent variable is present or absent and than determining whether the groups differ on the dependent variable.
causal-comparative research (“causal-comparative research (“ex post factoex post facto research”) research”)
…at least two different groups are compared on a dependent variabledependent variable or measure of performance (called the “effect”) because the independent variableindependent variable (called the “cause”) has already occurred or cannot be manipulated
Research variables...
IndependentIndependent……an activity of characteristic
believed to make a difference with respect to some behavior
…(syn.) experimental variable, cause, treatment
dependent variabledependent variable……the change or difference occurring
as a result of the independent variable
…(syn.) criterion variable, effect, outcome, posttest
Data analysis and interpretation…
…researcher uses a variety of descriptive and inferential statistics:
meanmean
standardstandard deviationdeviation
t-testt-test
analysis of analysis of variancevariance
chi squaredchi squared
meanmean
…the descriptive statistic indicating the average performance of an individual or group on a measure of some variable
standard deviationstandard deviation
…the descriptive statistic indicating the spread of a set of scores around the mean
t-testt-test
…the inferential statistic indicating whether the means of two groups are significantly different from one another
analysis of variance (“ANOVA”)analysis of variance (“ANOVA”)
…the inferential statistic indicating the presence of a significant difference among the means of three or more groups
chi squared (chi squared (ΧΧ22))
…the inferential statistic indicating that there is a greater than expected difference among group frequencies
Research Designs “True” Experimental Design : The Researcher actually manipulates the independent variable
Non-Experimental Design: Passive Observation by Researcher
Quasi-Experimental Design: Assignments to experimental conditions are not random
Experimental DesignsExperimental research design: The researcher has control over the experiment in terms of sample selection, treatment, environment, etc.Experimental designs are typical in psychology, medicine, education, etc.
Experimental Designs
Experiments often discuss pre and post test observations
POST-TEST ONLY
X O 1
Where:
0t = Observation in time t of experimental group
X = Treatment
0c = Control group
experimental researchexperimental research
…the researcher selects participants and divides them into two or more groups having similar characteristics and, then, applies the treatment(s) to the groups and measures the effects upon the groups
Types of experimental comparison…
1. comparison of two different approaches (A A versus BB)
2. comparison of an existing approach to a new approach (AA and ~~ A A)
3. comparison of differing amounts of a single approach (AA and a a or aa and AA)
where:
A – experimental (“treatment”) group
B – control (“no treatment,” “nonmanipulated”) group
Group experimental designs…
1. single-variable
2. factorial
types of pre-experimental designstypes of pre-experimental designs
one-shot case studyone-shot case study
X O
…a single group exposed to a treatment ( X ) and then posttested ( O )
one-group pretest-posttest designone-group pretest-posttest design
O X O
…a single group is pretested ( O ), exposed to a treatment ( X ) and, then, is posttested ( O )
static group comparisonstatic group comparison
X1 O
X2 O
…involves at least two groups ( X ), one receiving a new, or experimental treatment ( X1 ) and another receiving a traditional, or control treatment ( X2 ) and, then, are posttested ( O )
“True” experiments defined An experiment that utilizes random assignment to
conditions in an effort to ensure that the participants in each condition are statistically identical. In doing so, any differences observed in the dependent variable are attributable only to the presence/absence of the independent variable.
Campbell & Stanley’s taxonomy
RO1 X O2
RO3 O4
where R = random assignment, O = observation,X = treatment
types of true experimental designstypes of true experimental designs
pretest-posttest control group designpretest-posttest control group design
R O X1 O
R O X2 O
…at least two groups are formed by random assignment ( R ), administered a pretest ( O ), receive different treatments ( X1, X2 ), are administered a posttest, and posttest scores are compared to determine effectiveness of treatments
posttest-only control group designposttest-only control group design
R X1 O
R X2 O
…at least two groups are formed by random assignment ( R ), receive different treatments ( X1, X2 ), are administered a posttest, and posttest scores are compared to determine effectiveness of treatments
Solomon four-group designSolomon four-group design
R O X1 O
R O X2 O
R X1 O
R X2 O
…four groups are formed by random assignment ( R ) of participants, two groups are pretested ( O ) and two are not, one pretested and one unpretested group receive the experimental treatments ( X1, X2 ), each group is are administered a posttest on the dependent variable, and posttest scores are compared to determine effectiveness of treatments
factorial designsfactorial designs
…involve two or more independent variables with at least one independent variable being manipulated by the researcher
Experimental DesignFactorial Design
Diagram
Independent Variable #1: Teaching MethodIndependent Variable #2:Aptitude
Randomly assigned 3rd
graders scoring below 60 on an aptitude test.
Randomly assigned 3rd
graders scoring below 60 on an aptitude test.
Randomly assigned 3rd
graders scoring about 85 on an aptitude test.
Randomly assigned 3rd
graders scoring above 85 on an aptitude test.
Reading/Lecture/Etc.Lecture Only
High
Low
Independent Variable #1 Teaching Method
How many possible Teaching Methods are there? Which will be the methods used in the study? If more than one will be used, each method may be considered a factor of the variable known as Teaching Method.
Teaching Method
Lecture onlyLecture & SmallGroup Discussion
Independent Variable #2 Aptitude
How many possible levels of aptitude are there? How many may be represented in the group of subjects participating in the study? Once identified, levels of aptitude may be considered factors of the variable known as Aptitude.
Low
High
Aptitude
Lecture onlyLecture & Small
Group Discussion
Low High
examples of factorial designsexamples of factorial designs
two-by-two factorial design two-by-two factorial design (four cells)
2 X 2
…two types of factors (e.g., method of instruction) each of which has two levels (e.g., traditional vs. innovative)
A 2 X 2 factorial design…
Independent Variable
A B
Dep
ende
nt V
aria
ble
O
O
Group #1 Group #2
Group #3 Group #4Cells
not manipulatedmanipulated
A 2 X 2 factorial design…
A
No interaction between factors
B
A 2 X 2 factorial design…
ANo interaction between factors
B
A 2 X 2 factorial design…
A
Interacting factors
B
A 2 X 2 factorial design…
A
Interacting factors
B
two-by-three factorial design two-by-three factorial design (six cells)(six cells)
2 X 3
…two types of factors (e.g., motivation; interest) each of which has three levels (e.g., high, medium, low)
Single-subject experimental designs…
1. A – B – A withdrawal
2. multiple baseline designs
3. alternating treatments designs
simple A – B designsimple A – B design
…baseline measurements ( O ) are repeatedly made until stability is established, then the treatment ( X ) is introduced and an appropriate number of measurements ( O ) are made during treatment implementation
simple A – B designsimple A – B design
O O O X O X O X O
baseline treatment phase phase
A | B
A – B – A withdrawal designsA – B – A withdrawal designs…baseline measurements ( O ) are repeatedly made
until stability is established, then the treatment ( X ) is introduced and an appropriate number of measurements ( O ) are made during treatment implementation, followed by an appropriate number of baseline measurements ( O ) to determine stability of treatment ( X )
A – B – A withdrawal designsA – B – A withdrawal designs
O O O X O X O X O O O
baseline treatment baseline phase phase phase
A | B | A
multiple-baseline designsmultiple-baseline designs
…used when a return to baseline conditions is difficult or impossible since treatment effects oftentimes do not disappear when a treatment is removed
…“multiple” refers to the study of more than one behavior, participant, or setting
…instead of collecting baseline data on one specific behavior, data are collected on: (1) several behaviors for one participant, (2) one behavior for several participants, or (3) one behavior and one participant in several settings
…then, over a period of time, the treatment is systematically applied to each behavior (or participant, or setting) one at a time until all behaviors (or participants or settings) have been exposed to the treatment
multiple baseline designmultiple baseline design
Example: one treatment for three behaviors in three settings
behavior 1 O O OXOXOXOXOXOXOXOXOXOXOXO setting #1
behavior 2 O O O O O OXOXOXOXOXOXOXO setting #2
behavior 3 O O O O O O O O OXOXOXO setting #3
A B
the baseline remains the same…
…while the treatment is applied at other
settings
Threats to validity…
…internalinternal: factors other than the independent variable that affect the dependent variable
…externalexternal: factors that affect the generalizability of the study to groups and settings beyond those of the experiment
Threats to internal validity…
1. history
2. maturation
3. testing
4. instrumentation
5. statistical regression
6. differential selection of participants
7. mortality
8. selection-maturation interaction
historyhistory
…the occurrence of events that are not part of the experimental treatment but that occur during the study and affect the dependent variable
maturationmaturation
…the physical, intellectual, and emotional changes that occur naturally in a study’s participants over a period of time
testingtesting
…refers to improved scores on a posttest as a result of having taken a pretest
instrumentationinstrumentation
…the unreliability or lack of consistency in measuring instruments that can result in an invalid assessment of performance
statistical regressionstatistical regression
…the tendency of participants who score highest on a test to score lower on a second, similar test and vice versa
differential selection of participantsdifferential selection of participants
…the outcome when already formed groups are compared raising the possibility that the groups were different before a study even begins
mortalitymortality
…the case in which participants drop out of a study which changes the characteristics of the groups and may significantly affect the study’s results
selection-maturation interactionselection-maturation interaction
…if already-formed groups are used in a study, one group may profit more (or less) from a treatment or have an initial advantage because of maturation, history, or testing factors
Threats to external validity…
1. pretest-treatment interaction
2. selection-treatment interaction
3. multiple treatment interference
4. specificity of variables
5. treatment diffusion
6. experimenter effects
7. reactive effects
pretest-treatment interactionpretest-treatment interaction
…the situation when participants respond or react differently to a treatment because they have been pretested
multiple-treatment interferencemultiple-treatment interference
…the situation when the same participants receive more than one treatment in succession
selection-treatment interferenceselection-treatment interference
…the situation when participants are not randomly selected for treatments
specificity of variablesspecificity of variables
…the situation when a study is conducted with (1) a specific kind of participant; (2) is based on a particular operational definition of the independent variable; (3) uses specific dependent variables; (4) transpires at a specific time; and, (5) under a specific set of circumstances
treatment diffusiontreatment diffusion
…the situation when different treatment groups communicate with and learn from each other
experimenter effectsexperimenter effects
…the situation when the researchers present potential threats to the external validity of their own studies
reactive arrangementsreactive arrangements
…the situation when a number of factors associated with the way in which a study is conducted interacts with or shapes the feelings and attitudes of the participants involved
Types of reactive arrangements…
…Hawthorne effectHawthorne effect: any situation in which participants’ behavior is affected not by the treatment per se but by their knowledge of participating in a study
…compensatory rivalrycompensatory rivalry: the control group is informed that they will be the control group for a new, experimental study (“John Henry John Henry effecteffect”)
…placebo effectplacebo effect: the situation in which half of the participants receive no treatment but believe they are
…novelty effectnovelty effect: the situation in which participant interest, motivation, or engagement increases simply because they are doing something different
Controlling for extraneous (confounding) variables…
1. randomization
2. matching
3. comparing homogeneous groups or subgroups
4. using participants as their own controls
5. analysis of covariance (ANCOVA)
randomizationrandomization
…the process of selecting and assigning participants in such a way that all individuals in the defined population have an equal and independent chance of being selected for the sample
matchingmatching
…a technique for equating groups on one or more variables, usually the ones highly related to performance on the dependent variable (e.g., pairwise matching)
comparing homogeneous groups or subgroupscomparing homogeneous groups or subgroups
…a technique to control an extraneous variable by comparing groups that are similar with respect to that variable (e.g., stratified sampling)
using participants as their own controlsusing participants as their own controls
…exposing a single group to different treatments one treatment at a time
Data analysis and interpretation…
for single-subject research
…a visual inspection and analysis of graphical presentations of results
…focuses upon: adequacy of the design; an assessment of treatment effectiveness (clinicalclinical vs. statistical significancestatistical significance)
Mini-Quiz…
True and false…
…causal-comparative studies attempt to identify the cause-effect relationships; correlational studies do not
True
…causal-comparative studies typically involve two (or more) groups and one independent variable, whereas correlational studies typically involve two (or more) variables and one group
True
…causal-comparative studies involve relation, whereas correlational studies involve cause
False
…oftentimes, causal-comparative research is undertaken because the independent variable could be manipulated but should not
True
…one of the most important reasons for conducting causal-comparative research is to identify variables worthy of experimental investigation
True
…“lack of control” means that the researcher can and should manipulate the independent variable
False
…each group in a causal-comparative study represents a different population
True
…the more similar two groups are on all relevant variables except the independent variable, the stronger the study is
True
…there is random assignment to treatment groups from a single population in causal-comparative studies
False
…lack of randomization, manipulation of the independent variable, and control are all sources of weakness in a causal-comparative design
True
…matching, comparing homogenous groups or subgroups, and covariate analysis are strategies that enable researchers to overcome problems of initial group differences on an extraneous variable
True
…interpretation of the findings in a causal-comparative study requires considerable caution because the cause may be the effect and the effect may be the cause
True
…extraneous variables or confounding factors may be the real “cause” of both the independent and dependent variables
True
Fill in the blank…
…groups selected for a causal-comparative study which differ on some independent variable and comparing them on some dependent variable
comparison groups
Fill in the blank…
…unexplained variables that influence a dependent variable
confounding factors
extraneous variables
Fill in the blank…
…a method for controlling extraneous variables by comparing groups that are homogeneous with respect to the extraneous variable
comparing homogeneous groups
Fill in the blank…
…a method for controlling extraneous variables by forming subgroups within each group that represent all levels of the control variable
comparing homogeneous subgroups
Fill in the blank…
…a statistical tool to determine the effects of the independent variable and the control variable on the dependent variable, both separately and in combination
factorial analysis of variance
Fill in the blank…
…the descriptive statistic indicating the average performance of a group on a measure of some variable
mean
Fill in the blank…
…the descriptive statistic indicating how clustered or spread out around the mean a set of scores is
standard deviation
Fill in the blank…
…the inferential statistic determining whether there is a significant difference between the means of two groups
t-test
Fill in the blank…
…the inferential statistic determining whether there is a significant difference between the means of three or more groups
analysis of variance
Fill in the blank…
…the inferential statistic determining whether there is a greater than expected difference among group frequencies
chi squared
Fill in the blank…
…activities by which a researcher endeavors to ensure that the results of a causal-comparative study are not tainted by extraneous variables
control