svyplan

svyplan helps you determine survey sample sizes, allocate samples across strata and clusters, and assess expected precision or power. Start with anticipated outcome variability and design assumptions, then compare plans under different precision targets, budgets, and response rates.

The results describe a planned design. svyplan reports stratum-level design quantities such as sizes, sampling fractions and base weights. It never draws a sample and never attaches selection probabilities or weights to units. When the plan is settled, sondage draws it, and analysis of the realized sample belongs to survey or srvyr.

Installation

Install the development version from GitLab:

install.packages("pak")
pak::pkg_install("gitlab::dickoa/svyplan")

A quick example

Suppose a survey must estimate 70% vaccination coverage with a 5 percentage point margin of error. Previous surveys suggest a design effect of 1.5 and an 85% response rate.

library(svyplan)

coverage <- n_prop(
  p = 0.70,
  moe = 0.05,
  deff = 1.5,
  resp_rate = 0.85
)
coverage
#> Sample size for proportion (wald)
#> n = 570 gross (net: 485) (p = 0.70, moe = 0.050, deff = 1.50, resp_rate = 0.85)
#> expected cases = 338.8

The reported sample is the number of eligible people to approach. The response-rate adjustment increases that number so the expected responding sample meets the precision target. It does not correct nonresponse bias.

Pass the result to the paired precision function to check the same plan:

prec_prop(coverage)
#> Sampling precision for proportion (wald)
#> n = 570 (net: 485)
#> se = 0.0255, moe = 0.0500, cv = 0.0364, rmoe = 0.0714
#> expected cases = 338.8
confint(coverage)
#>  2.5 % 97.5 %
#>   0.65   0.75

The interval is an expected interval under the planning values. Survey data and realized design features determine the interval after collection.

Uncertain assumptions should be varied rather than treated as exact:

predict(
  coverage,
  data.frame(resp_rate = c(0.70, 0.80, 0.90, 1.00))
)
#>   resp_rate        n         se  moe         cv       rmoe
#> 1       0.7 691.4626 0.02551067 0.05 0.03644382 0.07142857
#> 2       0.8 605.0298 0.02551067 0.05 0.03644382 0.07142857
#> 3       0.9 537.8042 0.02551067 0.05 0.03644382 0.07142857
#> 4       1.0 484.0238 0.02551067 0.05 0.03644382 0.07142857

Find the right guide

Planning task	Functions	Guide
Start with sample size and precision	n_prop(), n_mean(), prec_prop(), prec_mean()	Getting started
Choose a precision target or ratio estimand	n_prop(), n_mean(), n_ratio()	Precision targets
Combine indicators or domain requirements	n_multi(), prec_multi()	Multiple indicators and domains
Construct strata and allocate a sample	strata_bound(), n_alloc(), prec_alloc()	Stratification and allocation
Meet several allocation constraints	n_alloc()	Joint allocation
Plan a clustered survey	varcomp(), n_cluster(), design_effect(), effective_n(), design_df()	Multistage surveys
Plan screening or nonresponse follow-up	n_twophase(), prec_twophase()	Two-phase surveys
Plan changes, averages, panels, and rotations	n_change(), n_pooled(), n_panel(), design_rotation(), design_schedule()	Repeated surveys
Plan a detectable comparison	power_prop(), power_mean(), power_did()	Power

Function reference pages document every argument and returned component. The package website collects the reference and the task-oriented guides. Report problems through the issue tracker.