This is a 4-digit numeric variable.
PERWT indicates how many persons in the population are represented by a given person in the dataset. It is necessary to apply PERWT to produce accurate statistics for the "weighted" NAPP samples, specifically in those where the person records in the sample data represent differing numbers of person in the population.
Most NAPP samples are complete-count datasets, where each person represents one person in the population. It is not necessary to use weights in those samples.
The 1901 Canada sample is a flat 1-in-20 sample in which each record represents 20 persons in the national population. The use of weights is optional.
The 1871 Canada and 1875 Norway datasets are highly stratified, weighted samples. Users must use PERWT in order to calculate accurate statistics.
Note that the sample includes no data for the counties of Flintshire (in North Wales), and Selkirkshire and Kinross-shire (both in Southern Scotland).
The U.S. 1850-1910 samples have PERWT calculated such that statistics match published population counts for counties and/or State Economic Areas.
The Mecklenburg 1819 dataset has a highly clustered sample design of portions of territories. The weights are based on Territories in Mecklenburg (TERRIMS). NOTE: Because of a highly clustered sample design, the sample weights may not ensure representative results for all types of analyses.
Users combining complete-count and sample data should use PERWT to obtain correct population totals and proportions.
It should be noted that population totals and proportions in the complete-count data have no associated standard errors.
- All persons
- Canada: 1852, 1871, 1881, 1891, 1901, 1911
- Denmark: 1787, 1801
- Great Britain: 1851a, 1851b, 1851c, 1861a, 1861b, 1871b, 1881a, 1881b, 1891a, 1891b, 1901a, 1901b, 1911
- Iceland: 1703, 1729, 1801, 1901, 1910
- Mecklenburg: 1819
- Norway: 1801, 1865, 1875, 1900, 1910
- Sweden: 1880, 1890, 1900, 1910
- United States: 1850a, 1850b, 1860, 1870, 1880a, 1880b, 1900, 1910