Ps1007 - Questionnaire PDF

Preview

Summary

-What is your domain?
-Do your questions adequately cover your domain
......

Description

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

QUESTIONNAIRE DESIGN WHY USE A QUESTIONNAIRE? Advantages: • Useful for surveys of behaviour / attitudes / beliefs / demographics etc. • Easy to administer and score for a large number of people Disadvantages: • Low response rate • Desirability bias • Data is not as detailed as interview

ACHIEVING CONSTRUCT VALIDITY  Choosing question formats  Writing well-worded questions  Encouraging accurate responses Construct validity: to what extent are our Questions actually tapping into the psychological construct we are interested in? Closed questions: -Present respondents with a limited range of options to choose from -For example, do you enjoy research methods classes? They are the best, they are ok, they are the worst Coding the responses….Error? Open questions: -Provide spaces for respondents to explain and elaborate answers

 would then need to code the responses (as for the closed questions) Collecting different types of data: Questions about the same topic could be asked in such a way as to provide different data • Nominal  unordered categories • Ordinal  ordered categories • Interval+  continuous scale Often no single correct type of data to collect but, • Questions designed to elicit a particular type of data could limit answers • Reflect an underlying assumption which you may not be aware of

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

QUESTION FORMAT/WORDING Avoid ambiguity, imprecision and assumption Are you asking retrospective questions that rely on memory? Are you asking questions about knowledge respondents may not have? Avoid double-barrelled questions Avoid leading questions Writing well-worded questions: Leading questions “Would you agree that the government’s policies on health is unfair?” Double-barreled questions “did you believe the training programme was a good one and effective in teaching you new skills?”

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN Hidden assumptions “When did you stop beating your wife?” Value judgements “Do you agree followers of backwards religions should be excluded from holding public office roles?” Double negatives (vs. negatively worded items) “do you agree your lecturer is not inarticulate?” Question order Put sensitive questions near the end (e.g., death, sex, religion) Response sets Yea-saying / nay-saying, Fence sitting, Reverse coding / negatively worded items

Appearance and layout: Should always be word processed (font 12+) Instructions should be clear Clear spaces between questions

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN Piloting the questionnaire: How long does it take to complete? Were the instructions clear? Were any questions unclear or ambiguous? Did you object to answering any questions? Were any topics omitted? Was the layout clear? Any other comments? Administration of questionnaires: Decide target population and sampling frame?  list of all members of your population e.g., Electoral Register (available most libraries), Postcode address file (PAF) a list of all addresses to which mail can be sent (13% non-domestic addresses). Households, not people – so would need to contact the individuals Stratified random sampling  E.g., you know how many males/females are in your actual population, you try to sample randomly whilst maintaining a male/female ratio of participants that is consistent with your Decide type of sample: opportunity, random, stratified, etc. Model of administration: face-to-face, postal or phone, online Check for missing data (99)  missing data coded as 99 so you can tell SPSS to ignore it

ASSESSING RELIABILITY: CRONBACH’S ALPHA Reliability  is important because without reliability we can’t have validity associated with the scores (which should reflect what we want to measure) We need to do this type of analysis before we do any other type of analysis (e.g., Pearson’s r) – what’s the point of looking for significant correlations if the data is bad!

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

CRONBACH’S ALPHA

Example fear of computers and statistics:

Into SPSS:

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

-these values generally need to be above .3 -if any of those values were greater than the alpha value then we would delete question one (if below .3), then we would delete that question that goes above overall alpha value.

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN

PS1007 – QUANTITATIVE RESEARCH QUESTIONNAIRE DESIGN...