research methods instructor: cherisse seaton psychology 301 social psychology lecture 4, september...
Post on 21-Dec-2015
215 views
TRANSCRIPT
Research Methods
Instructor: Cherisse Seaton
Psychology 301Social Psychology
Lecture 4, September 11, 2008
Psychology Program Orientation Meeting
Day & Time: Monday Sept. 15 at 5:30 pmLocation: Bentley CentreWho: Undergraduate students wanting to learn more about the Psychology Program at UNBC
and free pizza
OverviewToday we will cover:
Introduction to research methodsCommon misinterpretations of research
evidence and fallacies in reasoningCritical thinking guidelines
ReadingsAronson et al. Chapter 2
ReviewChallenge to cognitive dissonance paradigm Important skills to make you ‘game show’
competitiveIntroduction to methods in Social Psychology
Research MethodsTheory
Definition: “An organized set of principles that can be used to
explain observed phenomena” (Aronson et al. p.32)
Explain a part of behaviourMust be testableParsimony
Research Design1.) Experimental Research
Random assignment to ‘groups’‘Quasi-experimental’
Looking at group differences, but you cannot assign individuals to groups
E.g., Gender2.) Correlational Research
Looking at the relationship between variables, but cannot determine cause and effect
3.) Observational ResearchObserving behaviour in natural habitat
Experimental Research1.) Experimental method (typically lab studies)
Definition: “The method in which the researcher randomly assigns
participants to different conditions and ensures that these conditions are identical except for the independent variable (the one thought to have a causal effect on people’s responses)” (p. 41)
Variables Independent variables (IVs) Dependent variables (DVs)
Conditions If variable has a number of levels, each is a different
condition Control vs. experimental group
Random Assignment vs. Random selection
Random selection“A way of ensuring that a sample of people is representative
of a population, by giving everyone in the population an equal chance of being selected for the sample” (Aronson et al. p.38)
E.g., Survey data – random digit dialingEnsures results of study will be generalizable
Random assignment“The process whereby all participants have an equal chance
of taking part in any condition of an experiment” (Aronson et al. p.45)
Once participants are selected, we randomly assign them to each ‘group’ or ‘condition’
Ensures there are no inherent difference between participants in each condition
Correlational Research2.) Correlational Method
Definition: “The technique whereby researchers
systematically measure two or more variables, and assess the relation between them (ie. how much one can be predicted from the other)” (p.36)
Determine relationships between variables
Correlation coefficientRelationship between two
variables. Positive correlation: as one variable goes up,
the other goes up Ranges from 0 (no correlation) to 1 (perfect
correlation)Negative correlation: as one variable goes up,
the other goes downRanges from 0 to -1
Correlation
Positive correlation
Negative correlation
No correlation
Research Approaches3.) Observational method
Definition: “The technique whereby a researcher observes
people and systematically records measurements of their behaviour” (p.33)
Observing behaviour in natural habitat
Archival analysis “A form of the observational method, whereby the
researcher examines the accumulated documents, or archives, of a culture (e.g., diares, novels, magazines, and newspapers)” (p.35)
Lab vs. field studiesLab studies
Advantages: Control – elimination of confounds
Disadvantages Artificiality - Generalization Simplicity Deception
Field studiesAdvantages:
‘Real world’ studies Can be experimental or observational
Disadvantages More confounds; less control
Interpreting ResearchImportance of “triangulation”
VariablesStudies/concepts
Common mistakesInterpreting causation from correlationTwo variables are related, but we cannot
claim that one causes the other3rd variableNeurological studies (e.g., fMRI) seem to be
particularly susceptible to misinterpretation E.g., Left-wing Vs. Right-wing Brains
Left-wing Vs. Right-wing BrainsAmodio et al. (2007) - Liberals and
Conservatives Inhibition task & neurological correlates.
Other misinterpretationsConclusion goes past the researchThe conclusion may be erroneous or is not
justified by the facts.
Are girls biologically predisposed to prefer pink?Hurlbert & Ling (2007)New York times:
“At last, science discovers why blue is for boys but girls really do prefer pink”
Sex differences in color preferencesConclusion: evolutionary adaptation
Men = blue sky, good weather (hunting) Women = Pink/red berry picking (gathering)
Study resultsDiscriminative
ability vs. preference
Cultural or genetic?
Evidence vs. hypothesis
Misinterpretations of ResearchLogical fallacies
1.) Fallacies of deception2.) Fallacies which use motive in place of
support3.) Fallacies which employ both deception and
emotion
1.) Fallacies of deceptionFalse dilemma/False alternatives -
"either/or" thinking Slippery slope/Adverse consequences
fallacy – The legitimacy of an argument is based on consequences
Appeal to ignorance – no evidence establishing x is false; therefore, x is true
Hasty generalizations/Availability bias - general conclusion based on an exceptional case, or on a very small sample
2.) Fallacies which use motive in place of supportAppeal to authority – Basing argument on
authority of it’s sourceAppeal to force (or fear) – You will be hurt if
you do not believe xAppeal to pity – I will be hurt if you do not
believe xAppeal to mass opinion /Ad populum fallacy
- validity of an argument should be evaluated by considering its acceptance among the general public
3.) Fallacies which employ bothAttacking the person – “Ad Hominem” –
Attempts to discredit source of an argument through circumstantial (guilt by association) evidence, or name calling Genetic fallacy - the error that the validity of
an argument should be evaluated by considering its source
Tu Quoque – Two wrongs – charge of wrongdoing is answered with retaliation
Example of many fallacies at workLilienfeld (2002) “When Worlds Collide”
Rind et al. (1998) meta-analysis: CSALink between CSA and psychopathologyAlthough the authors warned:
“…lack of harmfulness does not imply lack of wrongfulness (p.47)”
“…Society has, for instance, made laws against drinking by minors not because most adults state that underage drinking was a negative experience for them, but because legislators focus instead on concerns such as public safety (p.203).”
Rind et al. (1998) meta-analysisMeta-analysisDefinition:
“A statistical technique that averages the results of two or more studies to see if the effect of an independent variable is reliable” (Aronson et al. p.48)
Events Following ArticleDr. Laura
“If it’s science, why don’t they endorse it? If it’s not, why do they publish it?”
Attacked the technique of meta-analysisAttacked the legitimacy of the authorsAsserted that the studies results should not be
trusted as they contradicted conventional wisdom (common sense)
Events Following ArticleWhy didn’t people see the positive?
Not all victims of CSA are ‘doomed’ Possible to recover after experiencing CSA and victims can go
on to live full an happy lives; Possible mitigating variables, or experiences that allowed
them to move on with their lives (such as the support of a loving family)
Although the methods of the study were sound, it also in no way ‘proves’ CSA is not harmful (things are not proven in scientific research) Just as correlation cannot be taken to indicate causation,
(because there may be a 3rd variable, for example), lack of correlation cannot be taken to indicate lack of causation – could be some ‘mitigating’ variable accounting for lack of long term maladjustment
Dr. Laura’s ErrorsFundamental errors in logical reasoning:
False dilemmaAdverse consequences fallacyGenetic fallacyAd populum fallacy
Lay persons have a tendency to believe that if an argument runs counter to common sense, then it should not be trusted.
Why?Gap between academic and popular psycScientific evidence and the public’s
(mis)understanding“Common sense”
Social-cognitive errors: Motivated biases Belief perseverance (Confirmation bias)Pre-existing beliefs stemming from availability
biasCognitive dissonance Fear/threat promote defensiveness
Why?Social-cognitive errors: Motivated biases
Influence individual’s reasoning about information, even if that information is presented accurately.
If information challenges an individual’s self or their beliefs they may distort, deny or dismiss it. E.g., smoking
Introducing a complicated problem with no clear solution and highly political implications may be inviting denial
Avoiding MistakesUsing evidence based reasoning and avoiding
biasThinking critically about the information
you’re presented with (even when others aren’t)
Drawing reasonable conclusionsErrors in logic can be avoided by following 6
rules
Six rules of evidence based reasoning 1.) Falsifiability 2.) Logic3.)
Comprehensiveness
4.) Honesty5.) Replicability 6.) Sufficiency
1.) FalsifiabilityIt must be possible to conceive of evidence
that would prove the claim falseProblems with circularity
The use of multiple ‘outs’Original Freudian theory
Undeclared claimPredicting the future: “2009 will be a year
filled with changes”
2.) LogicAny argument offered as evidence in support
of a claim must be soundInvalid argument – counterexample:
All dogs have fleas; Mittens has fleas; therefore Mittens is a dog
Flea-ridden feline “Mittens” provides an effective counterexample – the argument is not valid
3.) ComprehensivenessThe evidence offered in support of a claim
must be exhaustive – that is, all the available evidence must be considered
Cannot simply discard the evidence that contradicts a theory, and consider only that which supports itEffectiveness of therapyPremonition: one prediction came true, but
hundreds did not
4.) HonestyThe evidence offered in support of any claim
must be evaluated without self-deceptionIf the overwhelming evidence falsifies your
belief, then you must come to the rational conclusion that your belief is false
Honestly seek out the ‘truth’
5.) Replicability If the evidence for any claim is based upon an
experimental result, or if the evidence offered in support of any claim could logically be explained as coincidental, then it is necessary for the evidence to be repeated in subsequent experiments or trials
Guards against implicit (or conscious!) experimenter bias and undetected error
Independent observers following the same procedures should achieve the same results
6.) SufficiencyThe evidence offered in support of any claim must be
adequate to establish the truth of that claim, with these stipulations:1. The burden of proof for any claim rests on the claimant2. Extraordinary claims demand extraordinary evidence,
and3. Evidence based upon authority and/or testimony is
always inadequateAbsence of disconfirming evidence is not confirmation
E.g., not all UFO sighting can be explained; therefore, they must be extraterrestrial
No amount of expertise can preclude human fallibility
Science Vs. Common SenseCommon sense is not especially commonScientific method:
Not merely formalized common senseTechniques designed to minimize believing
what is not trueDesigned to minimize the confirmation bias
Next Class…Attitudes Attitude Change
Subliminal messages: Past and presentPersuasionAdvertising