questions design

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:

http://www.education.gov.uk/complexneeds/modules/Module-4.4-Developing-inquiry-based-practice/All/downloads/m16p080c/(C)%20Questionnaire%20design.pdf

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