Research System

advertisement
Research System
Presentation of
results
Generalization
Culmination
Modification
Theory
Theory building
Hypothesis generation
Verification
Hypotheses
Measurement issues
Research design
Sampling issues
Statistical analysis
Interpretation
Data
Research Strategies
“All research strategies are
seriously flawed…”
McGrath, J. E. Martin, J., & Kulka, R. A. (1982). Judgment
Calls in Research. Beverly Hills, CA: SAGE Publications
Inc.
A Three-Horned Dilemma
 Different methods have different strengths:
• Rigor
• Relevance
• Generalizability
Every research strategy either avoids two of the horns by
an uneasy compromise but gets impaled, to the hilt, on
the third horn; or it grabs the dilemma boldly by one
horn, maximizing on it, but at the same time “sitting
down” (with some pain) on the other two horns.
(McGrath, 1982: 74)
Dilemmatics: McGrath Taxonomy
 The study of research choices and tradeoffs
 There is no perfect study. Research involves tradeoffs
 Research looks at Actors emitting Behaviors in a
Context
 We want
 Generalizability from Actors population
 Precise measure and control of Behavior
 Realistic Contexts for observation of actor behavior
RIGOR
II
Lab
Experimental
Experiment Simulation
Judgment
Task
Field
Experiment
III
I
J
Sample
Survey
Field Study
Formal
Theory
Computer
Simulation
GENERALIZABILITY
IV
RELEVANCE
Q I: Field Strategies
RELEVANCE
 Field study
– No deliberate manipulation; everything is measured
– Naturally occurring setting
– Example: Survey of QWL in 100 organizations
 Field experiment
– Deliberate manipulation of one or more variables
– Naturally occurring setting
– Example: Hawthorne studies; Greenberg’s work
Q II: Experimental Strategies
 Laboratory experiment
RIGOR
– Deliberate manipulation of variables
– Contrived setting
– Example: Effects of communication channels on team
performance, effects of feedback and goal-setting on
individual performance
 Experimental simulation
– Deliberate manipulation of variables
– Contrived realistic setting
– Example: Center for Creative Leadership Looking Glass
simulation; Zimbardo prison experiment
– http://www.youtube.com/watch?v=JxGEmfNl-xM
Q III: Respondent Strategies
 Judgment tasks
GENERALIZABILITY
– Emphasis on task/judgment selection, often with a
limited number of participants
– All variables are measured
– Example: “policy capturing” studies, creation of the
competing values framework
 Sample Survey
– Emphasis on sample selection
– All variables measured
– Example: National Survey of Organizations
Q IV: Theoretical Strategies
GENERALIZABILITY
 Formal theory/literature reviews
– No actual research participants
– Summarize the literature to create new models for
testing (inductive process)
 Computer Simulations
– No actual research participants
– All outcomes are computer- generated
– Example: garbage can decision-making processes,
monte carlo studies
Integrative & Hybrid Strategies
 Because every methodological choice is flawed
or incomplete, you can decrease the effects of the
trade-offs by:
 Using different methods across studies
 Using multiple methods within a single study
 Packaging different studies with different methods
together
Alternatively, Three Study Types
 Experimental
 Quasi-Experimental
 Correlational or passive observational study (field)
 Single subject (case study)
Important Concepts: The Building Blocks of
Research Methods
 Independent Variable
 Dependent Variable
 Extraneous Variable
 Hypothesis
 Experiment
Causal Inference
 Conditions needed for causality
 Covariation of cause and effect
 Temporal precedence (cause must come before effect in
time)
 Control to rule out alternative interpretations
 True experiments are best suited to infer causality
because the include the greatest degree of control
 Apply MAXIMINICON principle
 Maximize relevant systematic variance
 Minimize irrelevant systematic variance
 Control extraneous sources of variance
What is a TRUE experiment?
 There must be manipulation
 Manipulation of a cause results in an effect
 There must be random assignment to experimental
conditions
 There must be control of extraneous variable
Experiment
 Advantages
 high degree of control
 strong inference of causality
 measurement of behavior is precise
 often laboratory experiments can be replicated easily
 Disadvantages
 low realism
 low external validity in general
 some phenomena cannot be analyzed in a laboratory
 some variables may have a weaker (or stronger)
impact in the lab than they would in a natural
environment
Quasi-Experiment
 Advantages
 high realism
 greater external validity
 moderate degree of control
 moderate to high inference of causality
 Disadvantages
 internal validity may be compromised
 external validity may be compromised
 measurement may be imprecise
 it may be difficult to get people to agree to
participate
 it is often difficult to get access to many field settings
Correlational Field Study
 Advantages
 realistic
 data on a large number of variables can be collected
 the researcher's impact on the study is often lower
 allows exploration of contextual effects
 Disadvantages
 causality is difficult to assess
 internal validity may be compromised
 external validity may be compromised
 organizations may not agree to participate
 measurement of variables less precise than lab
 low response rate common
Single Subject/Case Study
 Advantages
 good for low base rate occurrences
 provides source of rich and descriptive data
 good for generating new ideas
 Disadvantages
 low internal and external validity
Threats to Validity
 Validity = the confidence we can have that our findings
from any study are “true”
 All research has threats to validity – that is, things that
minimize the degree to which we can embrace a
particular finding as “true”
 Many sources of validity threats but two common
ones:
 Research participants
 Researchers themselves
Threats due to Participants
 Roles
 Good subject
 Faithful subject
 Negativistic subject
 Apprehensive subject
 Role Multiplicity and Conflict
 Attributes of Participants
 Comprehension artifact
 They misunderstand
Threats due to Researcher





Attributes of researchers
Expectancies
Designer, observer, and interpreter effects
Data analyst
Tester

Poor measurement decisions
Week 2 Assignment
 Describe a research topic that you are interested in
 Write three hypothesis statements about relationships




you might expect
Identify what the IVs and DVs are in these
relationships
If there are mediators or moderators in your
hypothesis statements, identify what these are
Indentify the type of research strategy you would use
to study this research question
Use the article you identified last week as your
reference
Download