Zusammenfassung der Ressource
Questionnaire design
- Factors to consider
- Only use if
helps to meet
objectives
- Should be
easy to use,
process and
analyse
- Should limit
measurement
error
- Should
respect the
respondent's
dignity and
privacy
- Types of questionnaires
- Used in face to
face interviews
- Used in telephone interviews
- Self-administered
- For handheld
personal
computers
- Types of questions
- Open-ended questions
- Closed-ended questions
- Points to consider to avoid
threatening or objectionable
questions
- Using words
familiar to
respondents
- Using
open-ended
questions
- Establishing
context to
soften impact
- Loading question with
details in order to
minimise significance
of behaviour in itself
- Using a series of
questions on related
subjects, some
threatening and
some
none-threatening
- Asking in
borad
categories
- Including question in
sequence of questions
about non-threatening
bahavious
- Placing
question late
in
questionnaire
- Questionnaire layout
- Clear introduction to
encourage participation
- Interviews: Initial
introduction
should be
standardised
- Clear
instructions at
beginning
- Attractive,
easy and
convenient to
use
- Printing easy to
read and of suitable
size
- Questions should not
extend over more than
one page
- Questions should flow and be
ordered according to importance
and sensitivity
- Transition statements to
break up long
questionnaires
- Long enough to cover
all information without
being a burden
- Include all
possible
responses in
response lists
- Aids to recall
- Decide
how
questionnaire
will be coded
- Thank you's
at the end
- Self-administered:
invitation and
space for
comments
- Interviews: space
for starting and
finishing times and
comments
- Introduction or cover
letter to provide
adequate information on
study
- Who is doing study
- What study is about
- How their name was chosen
- What is expected
- Who will see
information
provided
- If
participation
voluntary
- Pre-test questionnaires
- Words understood?
- Questions
answered
correctly?
- Suggested bias?
- Closed-ended
questions responses
for all respondents?
- Questions interpreted similarly?
- layout inviting?
- etc.