Survey of Statistical Methods

1
Survey of Statistical Methods

• Wednesday, January 19, 2005

2
Group 1 Discussion points

• Paths to Knowledge
• Personal experience
• Authority and/or consensus
• Intuition
• Common sense tradition
• Rationalism reasoning

• Scientific Norms
• Universalism
• Organized skepticism
• Communalism
• Disinterestedness
• Honesty

• Empirical Methods
• Empiricism
• Rationality
• Rationalism

3
Group 2 Discussion Points

• Sources of Uncertainty
• Topical
• Investigative
• Methodological
• Design
• Internal
• External
• Statistical
• Descriptive statistics
• Inferential statistics
• Interpretive

• Research of any kind has three things in common
  that are both friend and foe
• Dealing with ignorance uncertainty
• Disputation and conflict within a framework
  shared by other researchers
• Learning and discovery through errors

• Kinds of uncertainty
• Distortion
• Confusion
• Inaccuracy
• Incompleteness
• Absence
• Uncertainty
• Probability
• Ambiguity or vagueness

4
Group 3 Discussion Points

• Two common strategies in research
• Begin with data reduction and then enhance as
  much as necessary
• Begin with data enhancement and then reduce as
  much as possible.

• Quantifying or even counting should not be
  undertaken without first considering whether
  these are sensible given possible arguments to
  the contrary.

• Counting
• vs
• Quantification

• Data reduction
• vs
• Data enhancement

5
What you must know before conducting attempting
any type of statistical analysis

• What is the purpose of the research?
• What type of research design is being used?
• What variables are measured?

6
Hypotheses, Purposes, and Questions - Variations
of the same thing.

• Hypothesis
• Children who were shown a video with mild
  violence were more aggressive on the playground
  than those who were shown a similar video without
  violence.

• Purpose
• The purpose was to explore the differences in
  aggressive playground behavior between children
  who were shown a video with mild violence vs.
  children who were shown a similar video without
  violence.

• Question
• What effect did watching a video with mild
  violence have on aggressive playground behavior
  in children?

7
Overview of Research Designs

• Quantitative
• Research in which the results are analyzed and
  presented as quantities or numbers. Utilizes the
  laws of probability.
• Qualitative
• Research in which the results are trends or
  themes that are described in words.

• Experimental
• Researchers give treatments and observe to see if
  they caused changes in behavior.
• Non-Experimental
• Researchers do not give treatments. Create
  treatment groups based on pre-existing
  qualities.

8
Research Designs
Qualitative
Quantitative
1
2
Experimental
Non-Experimental
3
4
9
Research Designs
Quantitative
Qualitative

• Pre-Experimental
• One-shot case study
• One-group pre-post test
• Static Group Comparison
• True Experimental
• Post-test only control group
• Pre-post test control group
• Solomon 4-group
• Factorial Designs
• Quasi-Experimental Designs
• Interrupted Time Series
• Equivalent Time Samples
• Non-equivalent Control Group
• Counterbalanced Design
• Case Study

Experimental
Non-Experimental
10
Research Designs
Quantitative
Qualitative

• Pre-Experimental
• True Experimental
• Factorial Designs
• Quasi-Experimental
• Case Study

Experimental
-None-
Non-Experimental
11
Research Designs
Quantitative
Qualitative

• Survey
• Epidemiology
• Reliability/Validity
• Meta-Analysis
• Historical Research
• Program Evaluation
• Case Study

-None-

• Pre-Experimental
• True Experimental
• Factorial Designs
• Quasi-Experimental
• Case Study

Experimental
Non-Experimental
12
Research Designs
Quantitative
Qualitative

• Phenomenological
• Symbolic Interaction
• Ethnography or Culture
• Ethnomethodology
• Case Study

• Pre-Experimental
• True Experimental
• Factorial Designs
• Quasi-Experimental
• Case Study

-None-
Experimental

• Survey
• Epidemiology
• Reliability/Validity
• Meta Analysis
• Historical Research
• Program Evaluation
• Case Study

Non-Experimental
13
Research Designs
Quantitative
Qualitative

• Pre-Experimental
• True Experimental
• Factorial Designs
• Quasi-Experimental
• Case Study

-None-
Experimental

• Phenomenological
• Symbolic Interaction
• Ethnography or Culture
• Ethnomethodology
• Case Study

• Survey
• Epidemiology
• Reliability/Validity
• Meta Analysis
• Historical Research
• Program Evaluation
• Case Study

Non-Experimental
14
Effects of Instructors Clothing on College
Students Evaluations of Instruction

• Background Notes
• Two female and two male graduate students
  alternated the way they dressed while giving the
  same lectures to undergraduates in 12 sections of
  introductory psychology. The types of clothing
  were formal professional (i.e. suit and dress
  shoes), casual professional (i.e. skirt and
  sweater, or slacks and dark sports shirt), and
  casual (worn jeans, T-shirt, open flannel
  shirt, and athletic shoes). Among other things,
  students rated the guest lecturers on
  competence and interesting presentation.

15
How does a researcher decide what design to use?
Curiosity
Internal Validity
Problem
Report
Assumptions
Conclusions
Literature
Delimitations
Hypotheses
Data Analysis
Limitations
Research Design
Data Collection
External Validity
Variables
Subjects
16
Assignment for Friday

• Complete online homework assignment (linked on
  the Lecture Schedule page)
• Download and read the material for Fridays lab
  (linked on the Lecture Schedule page)
• Meet in the CHS Computer Lab _at_ 8am