Chapter 9: Survey Research PDF

Chapter 9: Survey Research A cognitive model of survey responding Survey responding is a psychological process Must take these processes into consideration for constructing a good survey questionnaire.  A cognitive model of survey responding Important considerations to answering this questi...

Chapter 9: Survey Research A cognitive model of survey responding Survey responding is a psychological process Must take these processes into consideration for constructing a good survey questionnaire.  A cognitive model of survey responding Important considerations to answering this question: 1) Respondent has to interpret what is average? 2) Respondent has to interpret what is a typical day? Not really a great question. Can be improved by: 3) Numerical values for the “average” 4) Different time period, like a week 5) Quantify the amount of alcohol (e.g. ounces)  Context effects on questionnaire responses Context effects: effects that are not related to the content but to the situation or circumstances in which the item appears E.g. rate political candidates  if the unpopular candidate comes up first, you might rate the following candidates more highly afterwards 1) Item-order effect: the sequence in which items appear affect people’s response e.g. Strack et al. (1988): General life satisfaction and dating frequency -The differing order of these questions might yield different Context effects on questionnaire responses 2) Response items can have unintended effects on responses e.g. How often do you get “really irritated”? “less than once a year” “more than once a month” Vs. “less than once a day” “several times a month” The time interval difference between the two options changes your interpretation of what meets the threshold for “really irritated”. Context effects on questionnaire responses -middle or centre option often perceived as being the “average/normal/typical” Solution: Counterbalance responses to avoid primacy or recency effects. e.g. first candidate listed on ballot receives a 2.5% boost simply by being listed in the first position Avoid too many responses that ask about extremes, and if you do have extremes, counterbalance these questions so everyone has a chance to score along the whole scale. Writing survey questionnaire items 1) Open-ended items: (IMPORTANT FOR EXAM) Purpose: to gain more intimate detail in responses Advantages: no bias or interference, more info can be gathered on a topic not well known, subjective experiences can be taken into account Disadvantages: may be difficult to quantify/code/measure/interpret Writing survey questionnaire items 2)Closed-ended items: (IMPORTANT FOR EXAM) Purpose: to use when the construct is well defined, or the variable is well defined Advantages: easier to code/interpret/measure, more practical for large-scale resources Disadvantages: you don’t get nuances between the select responses, no room for elaboration or subjective experiences  Writing survey questionnaire items ating scales: an ordered set of responses from which to choose from Writing effective items: BRUSO principle Writing effective items: BRUSO principle Avoid double barreled responses Response should be mutually exclusive and exhaustive Rating scales (5-7 response items is appropriate); Scales from 0-10 may be appropriate Extreme scores should be balanced around a neutral (modal) midpoint especially for bipolar dimension (the extreme ends of the scores) Modal point may be omitted