NRadjClass {PracTools}

Class-based nonresponse adjustments

Description

Compute separate nonresponse adjustments in a set of classes.

Usage

NRadjClass(ID, NRclass, resp, preds=NULL, wts=NULL, type)

Arguments

ID

identification value for a unit

NRclass

vector of classes to use for nonresponse adjustment. Length is number of respondents plus nonrespondents

resp

indicator for whether unit is a nonrespondent (must be coded 0) or respondent (must be coded 1)

preds

response probabilities, typically estimated from a binary regression model as in pclass

wts

vector of survey weights, typically base weights or base weights adjusted for unknown eligibility

type

type of adjustment computed within each value of NRclass. Allowable codes are 1, 2, 3, 4, or 5. (1 = unweighted average of response propensities, i.e., preds; 2 = weighted average response propensity; 3 = unweighted response rate; 4 = weighted response rate; 5 = median response propensity)

Details

The input vectors should include both respondents and nonrespondents in a sample. A single value between 0 and 1 is computed in each nonresponse adjustment class to be used as a nonresponse adjustment. Five alternatives are available for computing the adjustment based on the value of type. The value of the adjustment is merged with individual unit data and stored in the RR field of the output data frame.

Value

A data frame of respondents only with four columns:

NRcl.no

number of the nonresponse adjustment class for each unit

ID

identification value for a unit

resp

value of the resp variable (always 1)

RR

nonresponse adjustment for each unit

Author(s)

Richard Valliant, Jill A. Dever, Frauke Kreuter

References

Valliant, R., Dever, J., Kreuter, F. (2018, chap. 13). Practical Tools for Designing and Weighting Survey Samples, 2nd edition. New York: Springer.

See Also

pclass

Examples

require(PracTools)
data(nhis)
out <- pclass(formula = resp ~ age + as.factor(sex) + as.factor(hisp) + as.factor(race),
         data = nhis, type = "unwtd", link="logit", numcl=5)
    # unweighted average of response propensities within each class
zz <- NRadjClass(ID=nhis[,"ID"], NRclass = as.numeric(out$p.class), resp=nhis[,"resp"],
            preds=out$propensities, wts=NULL, type=1)