40
Research Questions & Research Questions & the “Language” of the “Language” of Variables & Hypotheses Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly) 4 (Mostly)

Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Embed Size (px)

Citation preview

Page 1: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Research Questions & the Research Questions & the “Language” of Variables & “Language” of Variables &

HypothesesHypotheses

Research Questions & the Research Questions & the “Language” of Variables & “Language” of Variables &

HypothesesHypotheses

Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Page 2: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Recall: Research QuestionsRecall: Research Questions

Questions researchers ask themselves, not Questions researchers ask themselves, not the questions they ask their informantsthe questions they ask their informants

Must be Must be empirically testableempirically testable NotNot

too vaguetoo vague too generaltoo general untestable (with implicit, untested assumed untestable (with implicit, untested assumed

outcomes)outcomes)

Page 3: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Relationship of Theory & Empirical Observation (Wheel of Science)Relationship of Theory & Empirical Observation (Wheel of Science)

Page 4: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Conceptualization & Operationalization Conceptualization & Operationalization

ConceptualizationConceptualization Conceptual, abstract (or theoretical) definition - a Conceptual, abstract (or theoretical) definition - a

careful, systemic definition of a construct that is careful, systemic definition of a construct that is explicitly written to clarify one’s thinkingexplicitly written to clarify one’s thinking

OperationalizationOperationalization linking a conceptual definition to specific measurement linking a conceptual definition to specific measurement

technique(s) or procedure(s)technique(s) or procedure(s)

operational definition - the definition of a variable in operational definition - the definition of a variable in terms of the terms of the specificspecific activities to measure or indicators activities to measure or indicators in the in the empiricalempirical world world

Page 5: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Matching Theoretical Concepts & Empirical (Operational) MeasuresMatching Theoretical Concepts & Empirical (Operational) Measures

Example: (Which county had “worst” damage from bad weather?)Example: (Which county had “worst” damage from bad weather?)

Page 6: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Conceptualization Issues: Distinguishing between Theory & IdeologyConceptualization Issues: Distinguishing between Theory & Ideology

SimilaritiesSimilarities Set of assumptions or starting pointSet of assumptions or starting point System of concepts/ideasSystem of concepts/ideas Specifies relationships between concepts (usually “causes”)Specifies relationships between concepts (usually “causes”)

But social scientific theoriesBut social scientific theories Recognize uncertainlyRecognize uncertainly Process orientedProcess oriented Based on evidenceBased on evidence Seek logical consistency etc..Seek logical consistency etc..

Page 7: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Elements of TheoryElements of Theory

ConceptsConcepts AssumptionsAssumptions Propositions/HypothesesPropositions/Hypotheses

about relationships, associationabout relationships, association

Page 8: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Which Theory is Best?Which Theory is Best?

Fewest assumptions (parsimony)Fewest assumptions (parsimony) Covers widest range of phenomenaCovers widest range of phenomena More accurate predictionsMore accurate predictions # 1

Page 9: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Measurement?Measurement? systematic observation systematic observation can be replicated (by someone can be replicated (by someone

else)else) Measures:Measures:

Concepts (constructs), Concepts (constructs), theoriestheories

measurement measurement instrument/toolsinstrument/tools

Need to recognize concept Need to recognize concept in observations (measures)in observations (measures)

??(# of library holdings as a measure of quality ??(# of library holdings as a measure of quality of university?)of university?) MacLeans Magazine survey

results, 2000.

Page 10: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

ConceptsConcepts Symbol (image, words, Symbol (image, words,

practices…)practices…) definitiondefinition must be shared to have social must be shared to have social

meaningmeaning Some only have one value Some only have one value

(homelessness)(homelessness) Concepts with Concepts with more thanmore than one one

possible value or attribute possible value or attribute sometimes called sometimes called variablesvariables

Concept clustersConcept clusters (ex. ethnic (ex. ethnic minorities)minorities)

Constructs Constructs (in operational stage-- (in operational stage-- use multiple measures or use multiple measures or indicators)indicators)

Page 11: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

AssumptionsAssumptions not necessarily explicit (may not necessarily explicit (may

be implied-- implicit)be implied-- implicit) not tested through not tested through

observation in the context observation in the context usedused

concepts and theories build concepts and theories build on assumptionson assumptions

Example: Some communication Example: Some communication research “deconstructs” research “deconstructs” assumptions in everyday life– assumptions in everyday life– can do the same with scholarly can do the same with scholarly researchresearch

Page 12: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Classification as conceptualizationClassification as conceptualization

typologytypology intersection of simple concepts forms new intersection of simple concepts forms new

conceptsconcepts broader, abstract concepts that bring broader, abstract concepts that bring

together narrower, more concrete conceptstogether narrower, more concrete concepts

ex. ex. Emile Emile Durkheim’sDurkheim’s 4 types of suicide4 types of suicide Varies by degree of integration to and Varies by degree of integration to and

regulation by societyregulation by society Altruistic (+I), Anomic (-I), Egotistical (-Altruistic (+I), Anomic (-I), Egotistical (-

R), Fatalistic (+R)R), Fatalistic (+R)

Photo: R. Drew, AP

Page 13: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

PropositionsPropositions

logicallogical statement about a statement about a (usually causal) (usually causal) relationshiprelationship between between two variablestwo variables

i.e. “Increased television i.e. “Increased television watching leads to more watching leads to more shared family time and shared family time and better communication better communication between children & their between children & their parents”parents”

Page 14: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

From Concept to MeasureFrom Concept to Measure

Neuman (2000: 162)

Page 15: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Examples of Developing Conceptual & Operational Definitions

Examples of Developing Conceptual & Operational Definitions

Construct = alienationConstruct = alienation if you theorize 4 components (family, work, if you theorize 4 components (family, work,

community, friends)community, friends)then operational definition must take all into then operational definition must take all into account & measuresaccount & measures

Construct= green consumer?Construct= green consumer?

Page 16: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Rules forCreating MeasuresRules forCreating Measures

Measures must be:Measures must be: mutually exclusivemutually exclusive

possible observations must only fit in one possible observations must only fit in one categorycategory

exhaustiveexhaustive categories must cover all possibilitiescategories must cover all possibilities

composite measures must also be:composite measures must also be: uni-dimensionaluni-dimensional

Page 17: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Operationalization Issue: Choices in Level of MeasurementOperationalization Issue: Choices in Level of Measurement

Based on Based on purposes of the study & conceptual purposes of the study & conceptual

definitionsdefinitions What is range in variation of “attributes” is What is range in variation of “attributes” is

necessary for measuring your concept?necessary for measuring your concept? Practical constraintsPractical constraints

Page 18: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

VariableVariable Must have more than one possible “value” Must have more than one possible “value”

or “attribute”or “attribute” context important, ex.context important, ex.

Religion (variable)Religion (variable)Possible Attributes: protestant, catholic, muslim, Possible Attributes: protestant, catholic, muslim,

jewish, etc…jewish, etc… Protestant (variable)Protestant (variable)

Possible attributes: baptist, united, presbyterian, Possible attributes: baptist, united, presbyterian, anglican etc...anglican etc...

Page 19: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

*Types of variables**Types of variables*

dependent variable (effect)dependent variable (effect) independent variable (cause)independent variable (cause) intervening variableintervening variable control variablecontrol variable

Page 20: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Causal RelationshipsCausal Relationships

proposed for testing (NOT like proposed for testing (NOT like assumptions)assumptions)

5 characteristics of 5 characteristics of causal hypothesiscausal hypothesis at least 2 variablesat least 2 variables cause-effect relationshipcause-effect relationship can be expressed as predictioncan be expressed as prediction logically link to research question+ a theorylogically link to research question+ a theory falsifiablefalsifiable

Page 21: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Examples of 2 possible Relationships between Two Variables (p.52)Examples of 2 possible Relationships between Two Variables (p.52)

Page 22: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Types of Hypotheses (note plural form)Types of Hypotheses (note plural form) null hypothesisnull hypothesis

predicts there is no relationshippredicts there is no relationship if evidence support null hypothesis then????if evidence support null hypothesis then????

Direct relationship (positive correlation) Direct relationship (positive correlation) Indirect relationships (negative correlation)Indirect relationships (negative correlation)

Page 23: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Ways of stating causal relationshipsWays of stating causal relationships

causes, causes, leads to, leads to, is related to ,is related to , influences, influences, is associated with, is associated with, if…then…, the higher….the lower if…then…, the higher….the lower etc…etc…

Page 24: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Hypothesis TestingHypothesis Testing

Page 25: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Possible outcomes in Testing Hypotheses (using empirical research)

Possible outcomes in Testing Hypotheses (using empirical research)

support (confirm) hypothesissupport (confirm) hypothesis reject (not support) hypothesisreject (not support) hypothesis partially confirm or fail to partially confirm or fail to

supportsupport avoid use of PROVEavoid use of PROVE

Page 26: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Causal diagramsCausal diagrams

X Y

X Y

Direct relationship (positive correlation)

Indirect relationship (negative correlation)

Page 27: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Types of Errors in Causal ExplanationTypes of Errors in Causal Explanation ecological fallacyecological fallacy reductionismreductionism tautologytautology teleologyteleology SpuriousnessSpuriousness

Page 28: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Double-Barrelled Hypothesis & Interaction EffectDouble-Barrelled Hypothesis & Interaction Effect

OR

Means one of THREE things

1

2

Page 29: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Interaction effectInteraction effect

Page 30: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Ecological Fallacy & ReductionismEcological Fallacy & Reductionismecological fallacy--wrong unit of analysis

(too high)reductionism--wrong unit of analysis (too low)

reductionism--wrong unit of analysis (too low)

Page 31: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Teleology & TautologyTeleology & Tautologytautology--circular reasoning (true by definition)teleology--too vague for testing

Neuman (2000: 140)

Page 32: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Spurious RelationshipSpurious Relationship

spuriousness--false relationship (unseen third variable or simply not connected)

Neuman (2000: 140)

Page 33: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

ExamplesExamples Storks and babiesStorks and babies

Lots of storks seen around an apartment buildingLots of storks seen around an apartment building An increase in number of pregnanciesAn increase in number of pregnancies ??????

?

Page 34: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

But...But...

The relationship is The relationship is spuriousspurious.. The storks liked the heat coming from the The storks liked the heat coming from the

smokestacks on the roof of the building, and so smokestacks on the roof of the building, and so were more likely to be attracted to that building.were more likely to be attracted to that building.

The tenants of the building were mostly young The tenants of the building were mostly young newlyweds starting families.newlyweds starting families.

So…the storks didn’t bring the babies after all.So…the storks didn’t bring the babies after all.

Page 35: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Causal Diagram for StorksCausal Diagram for Storks

Stork = SStork = S Baby = BBaby = B

S B+

Newlywed = NNewlywed = N Chimneys on Building Chimneys on Building

= C= C

N B+

C S+

Page 36: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Examples (cont’d)Examples (cont’d)

The larger the number of firefighters, the The larger the number of firefighters, the greater the damagegreater the damage

Page 37: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

But...But...

A larger number of firefighters is necessary A larger number of firefighters is necessary for a larger fire. Of course, a larger fire will for a larger fire. Of course, a larger fire will cause more damage than a small one.cause more damage than a small one.

Page 38: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Causal DiagramCausal Diagram

Firefighter = FFirefighter = F Damage = DDamage = D

Size of Fire = SSize of Fire = S

F D+

F

S+

+ D

Page 39: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

Examples from research (cont’d)Examples from research (cont’d) tall 15 yr. olds like shopping more than tall 15 yr. olds like shopping more than

basketballbasketball

Page 40: Research Questions & the “Language” of Variables & Hypotheses Baxter & Babbie, 2003, Chapters 3 & 4 (Mostly)

But...But...

Fifteen year old Fifteen year old girlsgirls are likely to be taller, are likely to be taller, since they are having a growth spurt at that since they are having a growth spurt at that age. age.

Fifteen year old girls are more likely to Fifteen year old girls are more likely to prefer shopping to sports. prefer shopping to sports.

Thus, it is gender, not height, that is the Thus, it is gender, not height, that is the deciding factor.deciding factor.