whipple: Whipple Index (Original and Modified)

In heaping: Correction of Heaping on Individual Level

Whipple Index (Original and Modified)

Description

The Whipple index is a demographic measure used to detect and quantify age heaping (digit preference) in population data. This function implements both the original (standard) and modified versions of the index.

Usage

whipple(x, method = "standard", weight = NULL)

Arguments

x	numeric vector holding the ages of persons.
method	character string specifying which index to calculate: "standard" Original Whipple index (default). Ranges 0-500, with 100 indicating no heaping. "modified" Modified Whipple index. Ranges 0-1, with 0 indicating no heaping.
weight	optional numeric vector holding the sampling weights of each person. Must be the same length as x. If NULL (default), unweighted counts are used.

Details

Calculate the original or modified Whipple index to evaluate age heaping.

The original Whipple index is obtained by summing the number of persons in the age range between 23 and 62, and calculating the ratio of reported ages ending in 0 or 5 to one-fifth of the total sample. A linear decrease in the number of persons of each age within the age range is assumed. Therefore, low ages (0-22 years) and high ages (63 years and above) are excluded from analysis since this assumption is not plausible.

The original Whipple index ranges from:

• 0: when digits 0 and 5 are never reported

• 100: no preference for 0 or 5 (perfect data)

• 500: when only digits 0 and 5 are reported (maximum heaping)

For the modified Whipple index, age heaping is calculated for all ten digits (0-9). For each digit, the degree of preference or avoidance is determined, and the modified Whipple index is given by the absolute sum of these (indices - 1), scaled between 0 and 1:

• 0: ages are distributed perfectly equally across all digits

• 1: all age values end with the same digit

Value

A single numeric value representing the Whipple index.

Author(s)

Matthias Templ

References

Shryock, H. S. and Siegel, J. S. (1976). The Methods and Materials of Demography. New York: Academic Press.

Spoorenberg, T. and Dutreuilh, C. (2007). Quality of age reporting: extension and application of the modified Whipple's index. Population, 62(4), 729-741.

See Also

sprague for disaggregating 5-year age groups.

Other heaping indices: bachi(), coale_li(), heaping_indices(), jdanov(), kannisto(), myers(), noumbissi(), spoorenberg()

Examples

# Equally distributed ages (no heaping)
set.seed(42)
age_uniform <- sample(1:100, 5000, replace = TRUE)
whipple(age_uniform)                    # Should be close to 100
whipple(age_uniform, method = "modified")  # Should be close to 0

# Strong heaping on 5 and 10 (ages ending in 0 or 5 only)
age_5year <- sample(seq(0, 100, by = 5), 5000, replace = TRUE)
whipple(age_5year)                     # Should be 500
whipple(age_5year, method = "modified")   # Should be close to 0.8

# Extreme heaping on 10 only (ages ending in 0 only)
age_10year <- sample(seq(0, 100, by = 10), 5000, replace = TRUE)
whipple(age_10year)                    # Should be 500
whipple(age_10year, method = "modified")  # Should be close to 1

# Using weights
weights <- runif(5000)
whipple(age_uniform, weight = weights)

heaping documentation built on Feb. 10, 2026, 1:08 a.m.