Create a Simple Random Sample Survey Design

Creates a survey design object for equal-probability simple random samples (SRS). Computes variances using the classical SRS formula var(ȳ) = (1 - f) × s² / n, where s² is the unweighted sample variance, n is the sample size, and f = n/N is the sampling fraction (0 when population size is unknown).

Usage

as_survey_srs(data, weights = NULL, probs = NULL, fpc = NULL)

Arguments

data

A data.frame containing the survey responses. Must have at least one row and unique column names.

weights

<tidy-select> Sampling weight column (a single column, values strictly > 0). Supply either weights or probs, not both. If both are NULL, uniform weights (= 1) are auto-assigned and a warning is issued; population totals will then equal sample totals, not estimated population totals.

probs

<tidy-select> Sampling probability column (a single column, values in (0, 1]). Converted to weights = 1/probs and stored internally as ..surveycore_wt... Supply either weights or probs, not both.

fpc

<tidy-select> Finite population correction column (a single column). Accepts either total population size (all values > 1) or sampling fraction (all values in (0, 1]). Cannot mix the two types. Cannot contain NA or non-positive values. If any FPC value (of population-size type) is less than the sample size, an error is thrown.

Value

A survey_srs object.

See also

as_survey() for designs with cluster or stratification structure, as_survey_rep() for replicate-weight designs

Other constructors: as_survey(), as_survey_calibrated(), as_survey_rep(), as_survey_twophase(), survey_calibrated(), survey_data(), survey_replicate(), survey_srs(), survey_taylor(), survey_twophase()

Examples

# Minimal: no weights, no FPC (uniform weights auto-assigned)
d <- suppressWarnings(as_survey_srs(data.frame(y = 1:5)))