theory Testable - Missouri State University
Transcript of theory Testable - Missouri State University
Unit 1Unit 1
Scientific StudyScientific Study of Educational PsychologyPsychology
What are the Goals of Scientific Study?y
Describe
Understand
Predict
Control
Prediction makes the theory Testable
Theorizing
That’s just a theory. Give me the facts!
Just the facts, Ma’am!
THEORYTHEORYTHEORYTHEORY
A Th i f lA Theory is a useful statement, or orderly, i t t d t fintegrated set of statements that allows one t d ib l i dto describe, explain, and predict events or l ti hi b trelationships between
events.
Truth in Science
Aristotle believed all heavenly bodies revolved
d h h faround the earth. Basis of navigation!
C i th i dCopernicus theorized revolution around the sun.
Einstein and relativityEinstein and relativity.
Building Formal TheoriesBuilding Formal Theories
FormalTheories
Evidence Hypothesis TestingEvidenceData
Hypothesis TestingResearch
Building Personal TheoriesBuilding Personal Theories
P lPersonalTheories
E idEvidenceData Observations
Issues that distinguish theories
• Goal directed/deterministicactive/passive– active/passive
– Free will vs. mechanistic• Quantitative/qualitative• Quantitative/qualitative
– Continuous vs. stages• genes/environment• genes/environment
– Universality/Diversityt ti / i i l• constructive/principle
Psychoanalytic Theoryy y yalso called Psychodynamic
• FreudFreudID, SUPEREGO, EGO • stagelike
• qualitative
• constructive• constructive
Behavioral LearningBehavioral Learning
• deterministicStimulus Response
• deterministicpassive
• Continuous
• quantitativequantitative
• environment
• principle
B.F.Skinner
Organismic/Adaption TheoriesOrganismic/Adaption Theories
• Constructivist– Note different use • Goal‐directed
ti• active
• Stages
• qualitative
• constructive
Piaget
Theory and the Scientific Method
Theory Hypothesis
Multiple
Theory yp
Evaluate Evaluate
MultipleWorkingHypotheses
Hypothesis Testing
relative to relative to other theory other theory Testing
Research
yyand researchand researchMultiple pstudies lead to conclusions
Research MethodsT h titiTeacher practitioner
Applying personal theories/beliefs
Teacher TheoristAction research
h dSophisticated Consumerof Information
Research DesignsResearch Designs
• DescriptiveDescriptive
• Correlational
i l• Experimental
• Developmental
Multiple studies lead to conclusionslead to conclusions
Descriptive Research
Data Collection methodsData Collection methods
Observational Methods
Case Histories
Survey DataySelf‐report
Archival DataArchival Data
Clinical Interview
Is changing your answer a good idea?idea?
3
3.5
2
2.5
%
1
1.5
0
0.5
wrong right right wrong wrong wrong
Correlational ResearchCorrelational Research
Leaves fall
Coldtemperature
Observation: Each year the leaves fall off of the tree at the
i hsame time the temperature declines
Interpreting CorrelationsInterpreting Correlations
Cold Leaves
Alternate hypothesis
Coldtemperature
Leavesfall
Interpreting CorrelationsInterpreting Correlations
Third variable alternate hypothesisThird variable alternate hypothesis
sun
Leaves Coldfall temperature
ba b
Alt t h th iAlternate hypothesis: Alternative explanations f h fi difor a research finding; e.g., a relationship between
i blvariables
Plausible alternate hypothesis: Believable alternate hypothesis
Interpreting CorrelationsInterpreting Correlations
Direction of causality
a b
y
Interpreting CorrelationsInterpreting Correlations
Third variable alternate hypothesisyp
c
a ba b
Reciprocal CausationReciprocal Causation
a b
Correlation CoefficientsCorrelation Coefficients
• Describes the extent of theDescribes the extent of the relationship between variables
• A perfect correlation is 1 or ‐1
• At or near zero means that variables are unrelated
F th f• Farther from zero means a stronger correlation
Correlation Coefficients100
R=.5890
80
Gra
de
70
6060
Days in class
45403530252050
Days Present
Correlation Coefficients100
R=‐.5890
e
80
Gra
de
70
6060
50
Days Absent
30252015105050
Days absent
88Correlation Coefficients
50
r =.8845
e
40
Gra
de
35
30
13108525
Attendance
131085
Days Present
Correlation Coefficients100
r = 34
100
90 r =.3490
80
Gra
de
80
70
60
50
Attendance
5040302050
Days Present
What is a good correlation?What is a good correlation?
• Test Reliability
• Decision MakingDecision Making
• Main predictor of variable/eventvariable/event
• One of many predictors fof variable/event
Problems with CorrelationalhResearch
Over inference of causality
Shotgun Correlational ResearchResearch
Grand conclusions from ll l ti hismall relationships
Strength of CorrelationalhResearch
• Naturalistic settingNaturalistic setting
• Can study variables you can’t manipulatep
• Theory building through strong inferenceg
ExperimentsExperimentsIndependent v. Dependent v.
• Independent Variable: varied controlledvaried, controlled manipulated by the experimenter.
• Dependent Variable: Measured to determine if its value is dependent on the independent variable.independent variable.
ExperimentsExperimentsTeaching Method Achievement
Independent variable
Dependent variable
Operational Definitions: The method by which a variable is translated into
b i hnumbers in a research study
Operational Definitions of Aggression:Are the valid?Are the valid?
Independent i bl Dependent variable
Aggression
variable Dependent variable
• Fights in dorms• Fights in dorms
• Bus’s Shock Machines
• Horn Honking
• Balloon popping• Balloon popping
Operational Definitions of Independent Variables:Are they valid?Are they valid?
TV violence Aggression
Independent variable Dependent variable
TV violence Aggression
p variable
E t i bl• Extraneous variables: Variables which vary other than the I.V. and D.V.
• Are there extraneous variables confounded with the I.V. that create alternate hypotheses?
Quasiexperimental Research
d d d
Teaching method Learning
Independent variable
Dependent variable
1 00 l i t l• 1:00 class: experimental
• 11:00 class: control
• What alternate hypotheses are created?
• How does randomizationcontrol extraneous variables?
Quasiexperimental Research
d d d
Teaching method Learning
Independent variable
Dependent variable
•Pretest-posttestPretest posttest designs
Factorial Studies: Schacter, 1971
Influence of clock speed on eating by obesity status of subjectobesity status of subject.
2.5
1 5
2
1
1.5normalobese
0.5
0normal speed fast speed
Problems with experimentsProblems with experiments• Cannot independently vary
d hage, and many other important variable.
• Generalizability: Manipulation of variables
t fl t t i lmay not reflect typical variation in the environmentenvironment.
• Manipulation of human behavior may createbehavior may create ethical risks.
Strengths of experimentsStrengths of experiments
• Opportunity to controlOpportunity to control extraneous variables that confound variables.
• Ability to draw causal inferences.
Single Subject Design (ABAB design)1010
Return to8
Baseline
Return toBaseline
6Baseline
4
2 Intervention ReintroduceIntervention
01 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7
Intervention
The Age VariableThe Age Variable
Can we Independently vary age?
• Cross-sectional Research
Cross‐sectional Study of Reading abilityCross sectional Study of Reading ability
80100
e
406080
ng sc
or
2040
read
in
010 20 30 40 50 60 70 80
ageCohort effects
Cross‐sectional Research
• comparability of groupsgroups,
• cohort effects • can't measure
change,g ,
The Age VariableThe Age Variable
• Cross-sectional Research
• Longitudinal Researchg
Cross‐sectional + longitudinal scores of bili d d id d iability to understand ideas expressed in
words
60
50
55
ore
Longitudinal
45
50
Sc
gCrossectional
4020 25 30 35 40 45 50 55 60 65 7020 25 30 35 40 45 50 55 60 65 70
Age
LongitudinalLongitudinal
• constancy or stability of traits
• more potential to consider the causal nature of relationships between variables
Cross Lagged Correlational AnalysisCross Lagged Correlational Analysis
Age 16Age 9
Violent TV
gAge 9
Violent TV
r
ra1b2
r ra2b2ra1b1
Aggressiveness Aggressiveness
If ra1b2 > ra2b1 and > ra1b1 and > ra2b2 then A → B
LongitudinalLongitudinal
• history• testing instrumentation• testing, instrumentation,
carryover• mortality
Cross‐sequentialq(time lag)
Time 1Time 1 Time 2Time 21010 20201010 2020
2020 3030
3030 4040
4040 5050
5050 6060
Common Plausible Alternate H hHypotheses
• Generalizability:
GeneralizabilityGeneralizability
• External Validity
– Naturalistic Experiments
• Representative sample:
– 70% of Psychology research with intro. Psychology students.
Common Plausible Alternate HypothesesHypotheses
• Generalizability:
• Chance• ChanceStatistical Significancep < 05p < .05
Effect sizesEffect sizes
• Interpreting small diffmean differences
• Diversity– Overconfident amateur psychologists
Common Plausible Alternate H hHypotheses
• Generalizability:
• ChanceChance
• Effect size
D d Ch i iD d Ch i i• Demand CharacteristicsDemand CharacteristicsPlaceboPlacebo
Experimenter BiasExperimenter Bias