panels: Panel Participation Rates
In leeper/responserates: AAPOR Survey Response Rates

Description Usage Arguments Details Value References See Also Examples

AAPOR panel participation formulae

recr(i, r = 0, nc = 0, o = 0, uh = 0, uo = 0, e = 1)

pror(i, p = 0, r = 0, nc = 0, o = 0, FUN = rr5)

comr(i, p = 0, r = 0, nc = 0, o = 0, FUN = rr5)

`i`	The total number of completed interviews.
`r`	The total number of refusals.
`nc`	The total number of non-contacts.
`o`	The total number of “other” units.
`uh`	The total number of unknown if occupied (housing) units.
`uo`	The total number of other unknowns.
`e`	Optionally, the (estimated) proportion of units of unknown eligibility (`uh` and`uo`) that are eligible. Otherwise assumed to be 1 (all eligible).
`p`	The total number of partial interviews.
`FUN`	A function specifying the calculation of the rate. Default is `rr5`.

These functions calculate response/participation metrics relevant to panel surveys, especially probability-based online panels. Specifically, three functions are exposed based upon definitions provided in AAPOR's “Standard Definitions”:

recr provides a “recruitment rate”, used to calculate the proportion of invited respondents who agree to empanelment.
pror provides a “profile rate”, used to calculate the proportion of empaneled respondents who complete an initial panel profile questionnaire.
comr provides a “completion rate”, used to calculate the response rate for a given survey.

pror and comr are simply wrappers for rr5 (by default) or another response rate function (specified in FUN. One possible calculation of e suggested by the Standard Definition's calculator is the number of eligible e = \frac{i + p + r + nc + o}{i + p + r + nc + o + uo}. The default used here, however, is the more conservative value of 1.

A numeric value indicating the response rate.

AAPOR. 2016. “Standard Definitions Final Dispositions of Case Codes and Outcome Rates for Surveys.”

rates, rr1, coop1, ref1, con1

# recruitment/empanelment rate
recruitment <- recr(1000, r = 1000, nc = 1000)

# profile rate
profile <- pror(i = 900, nc = 100)

# single-study completion rate
completion <- comr(i = 200, p = 50, r = 100)

# cumulative response rate
recruitment * profile * completion