LECTURE NOTES - QUESTION WORDING AND OTHER ISSUES
MINIMIZE Validity (systematic) - degree to which measure actually measures intended measurement.
MEASIREMENT Reliability (consistent) - degree measurement yields consistent results (same responses)
ERROR: Consistent - common measure of accuracy of measurement tool.
CONSTRUCTION RELIABILITY ASSESSMENT:
OF AN INDEX 1. Test-Retest Method - same measurement on same unit more than once and comparing results.
TO CHECK More similar results the measurements are = reliable. (eg: How old are you?)
RELIABILITY: 2. Internal Consistency Method - assess extent to which index related to individual items.
Several items w/ underlying concept = positively correlated with each other.
3. External Validation - indicator forms its relationship to indicators of other concepts to which the concept
measure should theoretically be related.
SOURCES OF 1. RESPONDENT CHARACTERISTICS 2. SURVEY CHARACTERISTICS
INCONSISTENC Non-attitudes: lack of knowledge / interest Interviewer effects
Y Lack of attention to question Order effects
Discomfort / reluctance to provide true opinion Method effects - format questions
Question wording effects
Answer set effects (wording of possible responses, forced /
unforced answers, # of possible responses)
3. CONTEXT EFFECTS
ways of manipulating something into yoaffecting answers to questions make certain aspects / criteria to judge salient issues.
Framing - formulating truth in such a way that you push ppl to respond
Specific definition - 2 logically equivalent statements of a problem lead people to choose different options.
Eg: give drug to 100 cancer patients & 80 cured OR give drug to 100 cancer patients & 20 dead.
Framing can cause:
Presentation of same problem, emphasis on subset of considerations = individuals to focus on these considerations
Issue frames= alternative definitions, constructions or depictions of a policy problem.
How framing works:
makes certain beliefs more / less imporpossible outcomes / consequences more or less lik