CRJU 320 Study Guide - Midterm Guide: Clinical Urine Tests, Construct Validity, Factor Analysis
- Use when no criterion exists & want an objective estimate of validity
- Factor Analysis:
§ What are they?
§ When do we use them?
§ Pros and Cons
o The problems / threats to measurement construct validity
§ What are they?
§ How can we protect against them?
Types of Measures
o Verbal v. Non-verbal: Two Key Dimensions:
Verbal: Written or spoken information from subject
Non-verbal: observations of behavior, physiological measures, archival records (e.g. medical file, school file)
•Obtrusive v. Unobtrusive
–Obtrusive: Subject is aware that data are being collected
–Unobtrusive: Subject is unaware...
•Both dimensions are important for reactivity
–Def: Extent to which a measure causes a change in the behavior of the subject, thus affecting validity of measure
o Reactivity & Measurement Type (i.e. What is reactivity, and how do different
types of measures affect it?) Reactivity and Measurement Type:
–Verbal – Non-verbal: •Subjects can control verbal behavior more than non-verbal
–Therefore...»Urine test better than self-report
–Obtrusive – Unobtrusive: •Subjects less reactive to unobtrusive measures
–Therefore...»Unknown surveillance more accurate than known surveillance
o If given examples of different types of measures, be able to categorize them
according to the characteristics listed above
Questionnaire Construction
o Questions measuring key causal and outcome constructs
o Questions measuring demographics
o Questions measuring potential confounding variables
o Closed-ended versus Open-ended questions
o Mutually exclusive and exhaustive answer choices
o Biased questions
Survey Research
§ The purpose of survey research: Surveys are one of the most frequently used measurement approaches in
criminal justice research
•Ability to collect large amounts of information about many constructs at one time
BUT...They need to be done well if we want valid & reliable data
•The use of well designed survey instruments (questionnaires) improves the quality of data we collect
•The use of appropriate sampling strategies helps to increase the representativeness of a sample
•The use of appropriate surveying techniques increases the accuracy of information collected from our sample
§ General Terms associated with survey research (e.g. sample, representativeness,
response rate, etc.)
§ The Literary Digest Poll Example (discussed in lecture
Document Summary
Use when no criterion exists & want an objective estimate of validity. Pros and cons: the problems / threats to measurement construct validity. Types of measures: verbal v. non-verbal: two key dimensions: Non-verbal: observations of behavior, physiological measures, archival records (e. g. medical file, school file: obtrusive v. unobtrusive. Obtrusive: subject is aware that data are being collected. Unobtrusive: subject is unaware: both dimensions are important for reactivity. Def: extent to which a measure causes a change in the behavior of the subject, thus affecting validity of measure: reactivity & measurement type (i. e. what is reactivity, and how do different types of measures affect it?) Verbal non-verbal: subjects can control verbal behavior more than non-verbal. Obtrusive unobtrusive: subjects less reactive to unobtrusive measures. Therefore unknown surveillance more accurate than known surveillance: if given examples of different types of measures, be able to categorize them according to the characteristics listed above.
