Designing Questionnaires Chapter 15 Selecting Question Type Open ended questions: no response options provided. Pros: access concepts the researcher has not considered, gain insight into respondents’ vocabulary regarding the topic, gain insight into relative importance (based on ordering) Cons: difficult to record, to code, to enter data resulting in greater expense and time Closed ended questions (fixed alternative): responses are provided. Pros: cheaper, faster, easier data collection and entry Cons: miss “rich” data, only measures what researcher has conceived. Fixed Alternative Questions Dichotomous: choose one of two , e.g. yes/no Multiple choice alternatives: determinant choice: select only one frequency determination: assesses general frequency, select one checklist: select all that apply Scale responses Likert, Semantic differential, sum scales, distance, etc. Question “Shoulds” Single issue per question Brief Common interpretation Comprehensible vocabulary Simple Exhaustive categories Mutually exclusive categories Question “Should nots” No unclear assumptions Why do you download music? (You may check off more than one box.) ____ CDs are too expensive. ____ CDs only contain a few good songs. ____ It is more practical (no need to leave home). ____ Rather spend money elsewhere. ____ Do not believe in supporting the music industry. ____ Convenience. ____ Other. This assumes that they download. Not beyond experience/ability Who was the best U of L president in the last 3 decades? Question “Should nots” No specific examples (unless needed) Don’t ask info they can’t remember What is your favourite brand of chocolate bar? (e.g. Coffee Crisp, Snickers, etc.) __________________ Stating a few limits ability to remember others (part-list cueing effect) How much did you spend on groceries in the last month? ____________ No ambiguity How many advertisements promoting music sharing can you recall? A)0 B)1-2 C)3-4 D) more than 4 Does this mean different ads, or different showings of an ad? Question “Should nots” No double-barreled questions I believe that all that is legal is ethical and all that is illegal is unethical. Strongly disagree 1 2 3 4 5 6 Strongly agree What if they believe all that is illegal is unethical, but all that is legal isn’t necessarily ethical? No leading questions Do you support the school board’s innovative new math curriculum? Not at all 1 2 3 4 5 6 Strongly support “innovative, new” suggests what the “correct” answer is Question “Should nots” No loaded questions Do you support helping child victims of AIDS? Not at all 1 2 3 4 5 6 Strongly support The “morally correct” answer is obvious Not overly complex Do you feel that the contractual clause 59B section II should be revised to indicate the new funding options available to employees in the temporary worker category when workers are eligible for maternity/paternity leave, conditional upon their years of service and job classification? Yes__ No__ Dillman's "Total Design Method“ for Survey Research (mail surveys) Use attractive single unit (e.g. one nice survey booklet) Offer incentives (don’t have to be large) Use postcard to warn Use postcard to remind Questionnaire Order “Cover letter” Instructions Get the right people Warm-up (if necessary) How to proceed (e.g. don’t skip ahead/back) Screening Who, purpose, explanation, incentive, screening (as needed) Easy questions Interesting questions Transition statement (if necessary) Key content/difficult questions Demographics Sequencing Gain attention, pique interest Avoid priming concepts, directing responses Response order; Question order Place similar questions together General to specific; Open to closed Randomize or counterbalance Start easy and interesting mind frame Clearly distinguish question and responses CAPS versus lower case, etc. Clearly distinguish question from responses: LAYOUT A (bad): Do you agree, disagree or have no opinion that this company has: A good vacation policy - disagree/not sure/agree. Good management feedback - disagree/not sure/agree. Good medical insurance - disagree/not sure/agree. LAYOUT B (better): Does this company have: Disagree Not Sure Agree A good vacation policy 1 2 3 Good feedback 1 2 3 Good medical insurance 1 2 3 Pretesting Pretest the questionnaire Test on “experts” Test on target market Pretesting approaches Complete the survey Interpret the questions: what do they “mean” Define key words Comment on confusing items, missing items Address response options as well as questions Response Issues Halo effect Positivity/negativity bias Extremeness aversion/extremeness seeking Framing Priming The End