What are some solutions to nonresponse? Select all that apply. A. reduce undercoverage B. use stratified sampling C. use convenience sampling D. change wording of questions E. offer rewards and incentives F. reduce interview error G. attempt callbacks H. use cluster sampling

E. offer rewards and incentives

It is true when people receive a reward or incentive they would be more willing to respond.

F. reduce interview error

False the interview error is not associated directly to the non response bias

Step-by-step explanation:

The effect of Non response occurs when we have a "significant difference between those who responded to an specified survey and those who did not".

And we have a lot of factors that can create this effect, for example: type of survey, bad questions, unprobabilistic sample, no incentives and other possibilites.

For this case we need to se which of the following options can be considered solutions for this effect. Let's analyze one by one.

A. reduce undercoverage

By definition Undercoverage happens when some "elements of the population are inadequately represented in the sample". But is a type of bias different to the non response, so this one is not a solution properly for the non response bias.

B. use stratified sampling

False, no matter the sampling method used we can't ensure that an specific random sampling method will reduce the non response bias.

C. use convenience sampling

False, no matter the sampling method used we can't ensure that an specific random sampling method will reduce the non response bias.

And with this unprobabilistic method we don't have a solution for this case.

D. change wording of questions

False, the non-response bias is associated with the selected sample does not read or does not respond to a specific questionnaire, if we change the wording of the questions and people do not read it, we will have the same non-response bias.

E. offer rewards and incentives

It is true when people receive a reward or incentive they would be more willing to respond.

F. reduce interview error

False the interview error is not associated directly to the non response bias

G. attempt callbacks

True if we use an alternative method in order to try get the information we will reduce the non response bias.

H. use cluster sampling

False, no matter the sampling method used we can't ensure that an specific random sampling method will reduce the non response bias.