Survey research: Ungraded review questions
Can you explain your answer? Use your fingers to indicate your answer: 1=A, 2=B, 3=C, 4=D. For “check all that apply,” use your both hands. After viewing the question, show me your answer in 15 seconds. Next, turn to your neighbor and you have one minute to convince him/her that you are right.
What is the problem of assigning “3” to “neutral” in a 5-point Likert scale? If a student wrote, “No idea” in the exam, the teacher should give him a zero. It is illogical to equate “neutral” or “no idea” to “3”. Neutral might mean different things to different people. All of the above
A 5-point Likert scale is ____ Nominal Ordinal Continuous It depends
What is(are) the advantage(s) of online survey (compared with hard copies or phone survey)? Lower cost More flexibility in survey design Higher response rate Less intrusive All of the above
Which one is NOT one of the purposes of pilot testing? Testing clarity of wording Testing user-interface Identifying logistic issues Increase the sample size by appending the pilot data to the data collected later.
It is a well-known fact that self-report data collected by survey research is not highly reliable. Today big data analysts overcome this problem by collecting behavioral data (i.e. what people do, not what they say) via the Internet (e.g. Google, Amazon, YouTube…etc.)? True. Both the quality (accuracy) and quantity (sample size) of the data are improved. False. There are ethical issues. Use of these data are prohibited by IRB.
Stephanie wants to conduct a nationwide survey Stephanie wants to conduct a nationwide survey. Which of the following sampling scheme should she use? Simple random sampling Convenience sampling Cluster sampling Stratified sampling Both C and D
Which of the following statement is FALSE? Some population segments are over- sampled and some are under-sampled. Sampling weights are needed for adjusting this disparity. We should use the raw sample weights to compute the mean and the standard deviation. We should use the normalized sample weight to compute the mean and the standard deviation.
Which of the following is NOT the right approach to address the concern that the sample may not represent the population in a large-scale survey research? Use simple random sampling Use multi-stage sampling. Check the sample properties against the population properties.