Some people classify as Dichotomous
Questions, Questions Based on Level of Measurement and Filter or Contingency
Questions (William & Trochim, 2006)
Dichotomous Questions
|
Question has two possible responses
|
Questions Based on Level of Measurement
|
Measure occupation using a nominal
question
|
Contingency Questions
|
Ask the respondent one question in
order to determine
|
For designing a questionnaire, there are some
general principles that appeared from William & Trochim (2006) and
Loughborough University which made a good summarization:
Ÿ Put the most important items in first
half of questionnaire.
Ÿ Do not start with awkward or
embarrassing questions
Ÿ Start with easy and non threatening
questions.
Ÿ Go from the general to the particular.
Ÿ Go from factual to abstract questions.
Ÿ Leave demographic and personal
questions until last.
But in my opinion, the most useful principle
is from Oppenheim (1992):
By running backwards through the survey
stages we can try to ensure logic-tightness, so that nothing is left out and we
shall have measured everything that needs to be measured. Thus, consider the
result that we expected, than design the question. This is a efficient and a
good way to design questionnaire.
Reference:
OPPENHEIM, A.N., 1992. Questionnaire Design, Interviewing and Attitude
Measurement. London: Continuum.
LOUGHBOROUGH UNIVERSITY, Questionnaire Design [online]. Loughborough University.
Available at:
JEAN, M., 1986. Survey Questions:
Handcrafting the Standardized Questionnaire, Series: Quantitative Applications in the
Social Sciences, Volume 63, pp. 12.
WILLIAM, M.K. & TROCHIM, 2006. Types of Questions [online]. William, M.K. & Trochim.
Available at:
http://www.socialresearchmethods.net/kb/questype.php
没有评论:
发表评论