DS350 - Quantitative Methods

Lecture Reviews #5-7 - Survey Practice

REVIEW QUESTIONS:

1) What is sampling error? What is nonsampling error? To what extent is each of these present in surveys? To what extent is each quantifiable?

2) What is self-selection bias in surveys?

3) What can be done to reduce nonresponse in surveys? Why is this important?

4) What are the three primary modes of survey administration, as given in class? What are the principle advantages and disadvantages of each?

5) What is meant by "coding" a survey? How is it done?

6) What is a pivot table in Microsoft Excel?

7) What is a random sample? Why is random sampling important?

COMPUTATIONAL EXERCISE:

Use the "Titanic" data for this exercise.

Your research question is: "Did the various classes of persons on board (crew, first, second, third) survive at different rates?" Another way of phrasing this is, "Is there a relationship between CLASS and SURVIVE?"

Construct an appropriate pivot table. Compute the survivorship rates for each class of passenger.