All College Day January 26, 2018 Inge Bond, Brian Goo, Erik Hou Office of Research, Planning, and Institutional Effectiveness Session Objectives: • Steps to survey design • Getting your questions right • Response options • Survey resources at Mission • Survey resources, external Steps to Survey Design: Rationale • What problem are you trying to solve? • Ask yourself: Can I take action based on what I learn? • What question(s) are you trying to answer? • Is there another way to get this information? Steps to Survey Design: Respondents • Who has the information needed to answer your research question(s)? • When thinking about your audience: • Consider equity issues – is your group representative of the population? • Are there questions you can ask that will help you to identify pertinent subgroups? • Demographic information • Habits (e.g. drivers versus cyclists) Steps to Survey Design: Deployment • Online (electronic) surveys: • Benefits: respondents can take their time, feel less recognizable in their responses • Drawbacks: harder to gather responses • In person (especially in-class): • Benefits: captive audience – participation is likely to be high • Drawbacks: respondents may not be as candid as they would be in an electronic survey Steps to Survey Design: Survey Format • Explanation and Consent: • Respondents must be competent and willing to answer the questions • Make the purpose of your survey clear to the respondents • Describe your anonymity/confidentiality policies and procedures • Keep it short and simple • Shorter surveys = higher response rates • Arrange questions in a logical order • Incentives • Increase response rates • Pre-test your survey • Have 3-5 colleagues test your survey out to look for grammatical errors, branching issues, and overall clarity Steps to Survey Design: Questions • Make sure that every question is necessary • Identify the data you need to collect, then write your questions • Use language that is simple and familiar to your respondents • Avoid acronyms or other “edu-speak” (know your audience) • Ask one question at a time • Expecting a single answer to a question that has multiple parts – muddies the responses and can confuse/frustrate respondents • and is a red flag of a double-barreled question • Example question: “Rate the effectiveness of Basic Skills STEM courses in student success” Steps to Survey Design: Questions • Avoid leading or biased questions • Look at your adjectives and adverbs – if not needed, take them out • “How much do you like…” • Avoid prestige bias • Some will exaggerate where they think a response is desirable (such as working long hours). • Avoid negative questions • Easily misinterpreted • not, prohibit, impossible are red flags of negative questions Steps to Survey Design: Questions, cont’d • Open-ended questions: • Enable respondents to provide information beyond what a fixed set of response options would • Entail more complex data analysis • Close-ended questions: • Greater uniformity of responses • Enable scaled responses • Simplify data analysis Steps to Survey Design: Likert Scale Questions • Use response scales whenever possible: • Allows for nuance and a more accurate data than binary response • Ideally, between five and seven points is best • In general, providing a middle category provides better data • Only caveat is if you want to make respondents give a non- neutral answer Steps to Survey Design: Likert Scale Questions • If possible, avoid agree/disagree questions, as respondents can be biased toward the “agree” option(s) (alternatives include measures of likelihood or satisfaction) • Example scaling options • https://www.uc.edu/content/dam/uc/sas/docs/Assessment/likert- type%20response%20anchors.pdf • Types of scales • Satisfaction • Awareness • Importance • Support/Opposition Steps to Survey Design: Creating the Survey/Survey Platforms • Survey Monkey – The college has established accounts • Advanced data analysis and survey export • Office365/One Drive Forms – New-ish. Comes with the college’s subscription to Office365 • For surveys of people with WVM id’s this can be a good way to limit participation • Google Forms – Most established open survey platform Survey Resources at Mission College • ORPIE is here to help! • Survey schedule: http://missioncollege.org/research/documents/surveys/Surveys_Scheduled_ 2017_2018.pdf • Avoiding survey fatigue • Instrument review • Survey Monkey accounts: • ORPIE (no public access – used for surveys with high need of confidentiality) • Office of Instruction (access through Patty E.) • Office of Student Services (access through Zita M.) Survey Resources, External: • Harvard University Program on Survey Research Tip Sheet on Question Wording • Survey Monkey survey guidelines • Qualtrics Research Success Kit (requires registration) • FERPA compliance Questions? Comments?