weights_for_population: Calculates survey weights given respondent information and...

In sate: Scientific Analysis of Trial Errors (SATE)

Calculates survey weights given respondent information and target population demographics

Description

Calculates survey weights given respondent information and target population demographics. Respondent demographic info must be properly encoded in respondentdata to work with the target.demographics. If respondent demographic info is missing, the respondent's weight will be coded 1. Weight values trimmed so that no weights are greater than 6 or less than .1.

Usage

weights_for_population(respondentdata, targetdata)

Arguments

respondentdata	Dataset with encoded respondent demographic information (use encode.cloud.respondent.variables to prepare respondentdata) must have a ParticipantId variable.
targetdata	A one row data.frame (or named vector) with the following statistics: black, ba_or_more, hhincome_over50k, age35plus, woman, hispanic (use target.population.demographics to obtain)

Value

Returns respondentdata with raked sampling weights encoded.

Examples

   library(sate)

   example_n <- 100
   example <- data.frame(Race = sample(x=c("Black or African American", "Other"),
                                       size=example_n, replace=TRUE),
                         Education = sample(x=c("Bachelor's degree (for example: BA, AB, BS)",
                                            "Other"), size=example_n, replace=TRUE),
                         Household.Income = sample(x=c("$70,000-$79,999", "Other"),
                                                   size=example_n, replace=TRUE),
                         Age = sample(x=18:80, size=example_n, replace=TRUE),
                         Gender = sample(x=c("Woman", "Man", "Prefer not to say"),
                                         size=example_n, replace=TRUE),
                         Ethnicity = sample(x=c("No, not of Hispanic, Latino, or Spanish origin",
                                            "Other"), size=example_n, replace=TRUE),
                         ParticipantId = 1:example_n)
   respondents.encoded <- encode.cloud.respondent.variables(dataset=example)

   pop.targets <- target.population.demographics(state="FL")

   respondents.weighted <- weights_for_population(respondentdata = respondents.encoded,
                                                  targetdata = pop.targets)

sate documentation built on April 3, 2025, 11:14 p.m.