Measurement and Data Collection - PowerPoint PPT Presentation

1 / 12
About This Presentation
Title:

Measurement and Data Collection

Description:

Reliability - the consistency of an instrument in measuring what it is designed ... Educational Testing Service (ETS) Test Collection Catalog (Oryx Press) ... – PowerPoint PPT presentation

Number of Views:28
Avg rating:3.0/5.0
Slides: 13
Provided by: Computing53
Category:

less

Transcript and Presenter's Notes

Title: Measurement and Data Collection


1
Chapter 15
  • Measurement and Data Collection

2
Measurement Scales
  • Nominal
  • Ordinal
  • Interval
  • Ratio

3
Reliability
  • Reliability - the consistency of an instrument in
    measuring what it is designed to measure
    (similarity of results over time)
  • Reliability Coefficient - a numerical indicator
    of the consistency of a measurement instrument
    (from 0.00 to 1.00)

4
Reliability Measures
  • Parallel (Alternative Forms)
  • Test-Retest
  • Split-Halves
  • Kuder-Richardson
  • Cronbach Alpha
  • Inter-rater
  • require only one administration of instrument

5
Validity
  • the extent to which an instrument measures what
    it is designed to measure (internal validity)
  • Content-related evidence
  • Criterion-related evidence
  • concurrent
  • predictive
  • Construct-related evidence

6
Commercial Instruments
  • eliminate need to develop own instrument
  • require little/no pilot testing (if used with
    appropriate population
  • have available reliability/validity statistics
  • overall - reduces effort in preparing for data
    collection
  • But more expensive and must be appropriate for
    the purpose of the study

7
Types of Instruments
  • Achievement tests
  • Attitude measures
  • Aptitude tests
  • Personality measurements
  • Rating scales
  • Observation systems

8
Holistic Scoring
  • Dependent variable - qualitative in nature (e.g.
    overall quality of an essay)
  • holistic - scores the response as a whole, rather
    than in parts
  • focuses on global quality
  • subjective
  • often ordinal - places responses in one of
    several ordered categories

9
Holistic Scoring
  • Summarize the qualitative aspects desired in the
    response (a model response)
  • Score responses anonymously and in random order
  • Have two/more equally competent scorers score
    responses independently
  • Use a third scorer to resolve discrepancies in
    the scores of two scorers
  • Score all responses to one item before beginning
    the next item
  • Make scorers aware of irrelevant factors that
    might affect scoring
  • Be aware of factors such as fatigue and boredom

10
Where to Find Test Information
  • Buros Tests in Print III
  • Buros Tenth Mental Measurement Yearbook
  • Index to Tests Used in Educational Dissertations
    (Fabiano, 1989)
  • Educational Testing Service (ETS) Test Collection
    Catalog (Oryx Press)
  • Volume 1 - Achievement Tests and Measurement
    Devices
  • Volume 2 - Vocational Tests and Measurement
    Devices
  • Volume 3 - Tests for Special Populations
  • Mirrors for Behavior (Simon Boyer, 1974)
  • Test Publishers
  • Measurement Journals

11
Ethical Legal Considerations
  • protect participants' dignity and welfare
  • prevent undue intrusion of privacy
  • maintain ethical standards of coworkers/assistants
  • inform participants of the purpose of the
    research
  • inform participants of their role in the project
  • inform participants of the research results
    (optional)

12
Procedures (Fowler, 1984)
  • Gather written confidentiality of all having
    access to data (who can identify respondents)
  • Minimize links between identifiers and answers
  • Prevent access to completed instruments by
    nonproject members
  • Remove identifiers before showing to nonproject
    members
  • Do not show data to those able to identify
    respondents by their demographic profiles
  • Do not allow general users of the data file to
    link identifiers and responses
  • During analysis, present data for small
    categories of respondents very carefully
  • The researcher is responsible for the storage or
    eventual destruction of instruments when the
    research project is completed
Write a Comment
User Comments (0)
About PowerShow.com