DESIGN

1
DESIGN

• The question drives the design

2
QUALITATIVE

• What is the purpose of the study?
• What is the question?

3
QUALITATIVE APPROACHES

• Meaning? . . . . . . . . . . . . .Phenomenology
• Culture? . . . . . . . . . . . . . Ethnography
• Process?. . . . . . . . . . . . .Grounded Theory
• Communication?. . . . . . Discourse Analysis

4

• Open-ended questions do NOT a qualitative study
  make!

5
Rigor and Fidelity

• to philosophical underpinnings are hallmarks of
  good qualitative research design

6
GOOD DESIGN

• Increases likelihood that research questions will
  be accurately answered
• Decreases number of competing explanations

7
RIGOR AND CONTROL

• are hallmarks of good quantitative design
• Validity of findings increased with fewer threats
  to validity

8
MAXMINCON

• MAX research variance
• experimental conditions are as different as
  possible
• e.g., nursing intervention different from usual
• e.g., wide variations in scores

9
MIN

• MIN error/random variance
• e.g., if fatigue a problem, measure at a
  different time
• e.g., ensure reliability of instruments

10
CON

• CON extraneous variables
• 1. random selection/random assignment
• 2. control groups
• 3. match participants
• 4. build variable into design as I.V.
• 5. statistical procedures (e.g., covariance)

11
THREATS TO VALIDITY

• Internal threats interfere with meaning of
  results with this sample
• Most important
• External threats interfere with ability to
  generalize results beyond study sample

12
INTERNAL THREATS

• History
• Selection
• (are those who participate different?
• Maturation
• (Would time have changed participants anyway?)

• Control group
• Control group, matching Dilman method
• Control group

13

• Testing
• (did it change participant?)
• Mortality/attrition
• (are those who dropped different?)
• Instrumentation
• (did everyone get the same? Was it
  accurate/precise?)

• Solomon 4/control gr
• Add to sample analyze drop-outs
• Train tester, inter-rater reliability calibrate

14
EXTERNAL THREATS

• Hawthorne
• Experimenter effects
• Tight control (artificiality of findings)

15
Experimental Designs

• Cause and Effect

16
Experimental Design

• Strongest external validity
• Randomization
• Manipulation of IV
• Control groups

17
Quasi-Experimental

• Lacking one of the big 3 of Experimental
• Control group call Comparison

18
Other Experimental Designs

• Pretest/Postest
• Posttest only
• Solomon 4

19
TERMS

• Cohort/panel ?threat
• Nonequivalent Groups not randomized therefore
  not equal
• Pre-experimental lacks 1/2 of big 3

20

• Participants as their own controls
• Secondary Analysis
• Meta-analysis
• Triangulation

21
Nonexperimental Designs

• May infer relationships
• May NOT infer cause and effect

22

• Ex post facto
• Retrospective
• Preexperimental/Nonexperimental

23
TIME

• Retrospective
• Prospective
• Cross Sectional
• Time Series

24
No study is an island . . .

• Science builds
• No single study proves . . .
• Findings support or fail to support
• What is the preponderance of evidence?