Chapter 11: Surveys & Questionnaires
Writing Good Questions
- Every question should be answerable by every subject - Questions should be easy to answer - Consider recall of information - Consider if respondents will be honest - Try to use a variety of question types - Questions should generate varied responses
Question Wording
- Purposeful language - Avoid bias "do you often go to your physician with trivial problems?" - Clarity - Avoid double barreled questions : using "or" or "and" to assess two things in the same question (confuses the subject) --"How many times a week do you run or weight train?" - Frequency and time-measures --The question "How often do you feel pain?" is vague and should be put in context --Better: "How many times did you feel pain for more than 24 hours in the past month?" - Sensitive questions
Measurement Scales Ex#1
-A scale is an ordered system based on a series of questions or items, resulting in a score that represents the degree to which a respondent possesses a particular attitude, value, or characteristic -Ex: Barthel Index (Ordinal Scale for ADLs)
Surveys
-A set of questions that elicits quantitative (numbers) or qualitative (open ended questions) responses Survey Formats -Questionnaires: A standard survey, usually, self-administered that asks individuals to respond to a series of questions -Interviews: The research asks respondents specific questions and records the answers
Planning the Survey
-In stage 1, planning includes delineation of the research question, the target population and hypothesis -In stage 2, development includes designing the survey, as well as creating and testing items to finalize format -Stage 3 includes IRB approval, selection of a sample, and distribution of the survey
Rank Order Questions
-Rank-Order: where the respondent is presented with a series of responses and is asked to rank them on an ordinal scale
Questionnaires
-Self-report: researchers do not observe the respondents behaviors or attitudes -Recall Bias: Do not accurately remember a past event or experience or leave out details when reporting -Longer recall period likely leads to more error --Does the subject have a motivation or concern to bias their response? ---e.g., work-related injuries, other consequences to a response
Numeric Rating Scale (NRS)
An assessment tool that measures a person's experience using a number rating (e.g. On a scale of 1-10, how severe is your pain?)
Analyzing Survey Results
Analysis of Questionnaire Data -The first step in the analysis of data is to collate responses and enter them into a computer. Each item on the survey is a data point and must be given a variable name, often an item number. Coding -Closed choice responses --Numeric codes that provide labels for data entry ---Ex: Code hospital size as 1 = less than 50 beds, 2 = 50-100 bets, and 3 = over 100 beds --Frequency counts and percentages -Open-Ended responses --Qualitative analysis Summarizing Survey Data -Quantitative Data -> descriptive statistics -Categorical data -> frequency of reports, cross-tabulations Missing Data -Providing a response choice of "don't know" or "I prefer not to answer" will eliminate missing data
Sensitive Questions
Any question that may cause some discomfort to the respondent Example: Elderly Patient --"Do you forget to take your medication?" Better: --How often do you miss taking your medication?
Checklist - Likert Scale: Exercise Testing, Perceived Exertion
Back extensor endurance test: 20 second hold 1 = very easy 2 = easy 3 = moderately hard 4 = hard 5 = very hard
Check All That Apply Questions
Can select from multiple answers and pair them up
Multiple-Choice Questions
Exhaustive -All possible responses that can be expected Mutually Exclusive -Each choice clearly represents a unique answer with no overlap
Ethics in Survey Research
Institutional Review Board (IRB) review and approval Informed Consent -Submission of questionnaire is considered informed consent -May be included In the cover letter -Face-to-face interviews provide a form to sign -For telephone interviews, provide full information at the start of the call to obtain verbal consent
Open vs Close-Ended Questions
Open-Ended: Ask respondents to answer in their own words -Advantages: Respondents own perspectives -Disadvantages: Difficult to interpret (becomes subjectiee) Close-Ended: Ask respondents to select an answer from among several fixed choices -Advantages: Easily coded, greater uniformity -Disadvantages: Respondents cannot express their individual viewpoints Ex: -Multiple choice questions -Dichotomous questions -Check all that apply -Measuring intensity -Checklists -Measurement scales -Visual analog scales -Rank-order questions
Questionnaires Advantages & Disadvantages
Pen and Paper or electronic format Advantages -No interview is needed -Standardized - everyone answers the same question -Complete on own time -Does it provide anonymity? Yes -Encourages honest answers -Large samples can be tested with minimal expenses Disadvantages -Questions on written questionnaires can be misinterpreted, or the responses misinterpreted -Unknown motivation of the respondent
Interviews
Subjects are asked specific questions and answers are recorded -Method: phone, face-to-face -Style/Type of Interviews --Structured (standardized set of questions) --Semi-structured (list of questions + discussion) --Unstructured (informal conversation) -Advantages: Opportunity for in-depth analysis -Disadvantages: cost, time, scheduling, and lack of anonymity of subjects
Dichotomous Questions
a directive form of questioning; these questions ask the customer to choose from two or more options -Which is better 2a or 2b
Checklists - Likert Scale
a numerical scale used to assess attitudes; includes a set of possible answers with labeled anchors on each extreme
Visual Analogue Scale
lets the patient make a mark along a 10-cm horizontal line from "no pain" to "worst pain imaginable"