What are the sources of survey error?
- Total survey error
- Systematic error
- Random error
- Sample design error
- Measurement error
What are systematic error?
- This is a measurement error and results from a variation between information sought and information collected
- This is potentially more serious than sample design error
What are the rules of the questionnaire design?
Survey creator’s job to avoid influencing respondents and build a survey that needs to have honest and thoughtful responses
What are question constructions?
Looking at the questions and ensuring its measuring what needs to measure
What are the rules of the questionnaire design?
Survey creator’s job to avoid influencing respondents and build a survey that needs to have honest and thoughtful responses
What are question constructions?
Looking at the questions and ensuring its measuring what needs to measure
What are open response questions
Encourages answers with full answers
What are closed response questions
yes or no questions
What are some considerations for closed response questions?
- Number of response categories
- Order of response categories
- Proactively managing uncertainty and ignorance
What are some considerations for closed response questions?
- Number of response categories
- Order of response categories
- Proactively managing uncertainty and ignorance
What are ambiguous questions?
Broad questions
What are some examples of ambiguous questions?
Who’s your banker
What are double barreled questions?
Questions that include more than one topic
What are some examples of double barreled questions?
How much do you enjoy collecting and analyzing data
What are the should of question wording?
- Question should be focused on a single issue or topic
- Question should be brief
- Questions should be interpreted in the same way by all the respondents
- Questions should use the respondents core vocabulary
- Questions should be applicable to all that are completing the instruments
- Question responses should be applicable to all who are completing the instrument
What are the shorten version of should of question wording
- One variable per questions
- avoid ambiguous wording
- consider your audience
- Respondent selection and survey flow
What are the should nots of question wording
- Should not give limited alternatives
- Should not have ambiguous wording
- Should not be double barreled
- Should not lead the respondent to a particular answer
- Should not have loaded wording or phrasing
- Should notes for details that cannot be related
- Should not use a specific example to represent a general case
- Should not be beyond the respondents ability to experience
- Should not require the respondents to guess a generalization
What does the Questionnaire layout look like?
- Introduction/Cover letter
- Identifying surveyor/sponsor
- Purpose of the survey
- Explanation of respondent selection
- Request for participation
- Anonymity
- Questions
- Open with screening or easy/non threatening questions
- Proceed from broad to specific
- Flow smoothly from one topic to another
- Demographics at end
- Thank you statement
What are the question order
- Level of specificity:
- Questions flow from general to specific
- Order of questions/answers
- Randomize questions related to concepts that your are testing
- Randomize order of answers
What are the categories of sensitive questions
- Categories
- Middle/End of instrument
- Momentum
- Counter biasing
- Everybody approach
- How others feel or third person
- Anonymity
What are the role of the pretest?
Attempts to minimize respondent and interviewer error
What are the types of pretesting the questionnaire
- Skip patterns
- Respondent interest and attention
- Logical flow
What are two types of error?
Random error and systematic error
What is chance variation?
Difference between the sample value and the two value of the population mean that can’t be eliminated
