Compute Gender-Based Sample Weights Using Baseline Wave Proportions

Compute sample weights for gender adjustment based on the baseline wave proportions.

Compute sample weights for gender adjustment based on the baseline wave proportions.

Usage

margot_compute_gender_weights_by_wave(
  data,
  male_col = "male",
  wave_col = "wave",
  target_wave,
  target_male_prop = 0.5
)

margot_compute_gender_weights_by_wave(
  data,
  male_col = "male",
  wave_col = "wave",
  target_wave,
  target_male_prop = 0.5
)

Arguments

data

A data frame containing gender and wave information.

male_col

A character string specifying the column that indicates male gender (1 for male, 0 for female). Default is "male".

wave_col

A character string specifying the column indicating the wave. Default is "wave".

target_wave

The value in wave_col that identifies the baseline wave.

target_male_prop

A numeric value between 0 and 1 representing the target proportion of males. Default is 0.5.

Value

A numeric vector of sample weights for all rows.

A numeric vector of sample weights for all rows.

Details

The function calculates the gender proportions in the baseline wave and computes weights so that the overall sample aligns with the target gender distribution. The same weights are then applied to all rows.

The function computes the sample proportions in the baseline wave and calculates weights by comparing these proportions with the target proportions. It upweights the underrepresented gender and downweights the overrepresented gender. The resulting weights are applied to the full dataset.

The function calculates the gender proportions in the baseline wave and computes weights so that the overall sample aligns with the target gender distribution. The same weights are then applied to all rows.

The function computes the sample proportions in the baseline wave and calculates weights by comparing these proportions with the target proportions. It upweights the underrepresented gender and downweights the overrepresented gender. The resulting weights are applied to the full dataset.

Examples

dat <- data.frame(
  id = 1:100,
  male = sample(c(0, 1), 100, replace = TRUE, prob = c(0.7, 0.3)),
  wave = rep(1:2, each = 50)
)
weights <- margot_compute_gender_weights_by_wave(dat, male_col = "male",
                                                 wave_col = "wave",
                                                 target_wave = 1,
                                                 target_male_prop = 0.52)
head(weights)
#> [1] 0.6486486 0.6486486 0.6486486 0.6486486 0.6486486 2.0000000

dat <- data.frame(
  id = 1:100,
  male = sample(c(0, 1), 100, replace = TRUE, prob = c(0.7, 0.3)),
  wave = rep(1:2, each = 50)
)
weights <- margot_compute_gender_weights_by_wave(dat, male_col = "male",
                                                 wave_col = "wave",
                                                 target_wave = 1,
                                                 target_male_prop = 0.52)
head(weights)
#> [1] 0.750000 1.444444 0.750000 0.750000 1.444444 1.444444