ch 5 survey design Flashcards
*questionairre definiton
set of questions designed to generate the data necessary to accomplish objectives of research process
*components of questionairre
ID data, classification data (marital status, income), info sought, instructions (better if respondent doesn’t know goal), request for cooperation
*questionaire design process
- specify info needed 2. specify type of interview method 3. determine the content of individual questions 4. design the question to overcome respondents inability and unwillingness to answer 5. decide the question structure 6. determine the question wording 7. arrange questions in proper order 8. id form and layout 9. reproduce questionnaire 10. eliminate bugs by pretesting
*personal interview
adv: clarify misunderstandings, high degree of flexibility
*personal interview dis
dis: bias of interviewer, hard to replicate
*adv telephone
shorter
*dis telephone
no visual cues
*adv self admin survey
ask any question you want
*dis self admin survey
interviewer not present to exercise quality control
factors affecting response rate
length, characteristics, credibility, motivation
*response format
open ended
open ended*
free response adv: insights, exploratory research dis:time/cost
coding verbal data
two people read verbal response, reader forms useful list, merge list w 3rd person, 4th and 5th sort into categories, together sorters agree on final scheme
Mulit- response questions*
adv: reduce interview bias dis:, easy to code, easy to administer, dis: costly and time consuming, difficult to design
*dichatomous questions
binary, natural point dis: may miss grades/feelings, included natural response adv:easy to admin
*question wording
simple vocab? ambiguous meaning?double barreled?leading? confusing? applicable? appropriate length?
*filter question
measure familiarity and past experiences
*sensative questions
hide in middle, place domestic sensitive questions at end
*leading
clues respondent as to what answer should be
*doube barreled
two or more questions are designed into one
*implicit alternative
alternative that is not explicitly expressed
*implict assumptions
what will happen as a consequence of action
*question sequence
open question interesting and simple, basic info obtained first, difficult questions placed later, general precede specific, questions asked in logical order, branching questions off of branded question
*pretesting
all aspect testing including wording sequence, form, difficulty, and instructions, response in pretest similar to that in actual survey, conducted by mail or telephone if used in actual survey