Impute standard deviation from measures of dispersion.

Usage

impute_sd(point, var1, var2, n, vartype)

impute_sd_sd(point, var1, var2, n, vartype)

impute_sd_cv(point, var1, var2, n, vartype)

impute_sd_se(point, var1, var2, n, vartype)

impute_sd_ci(point, var1, var2, n, vartype)

impute_sd_iqr(point, var1, var2, n, vartype)

impute_sd_range(point, var1, var2, n, vartype)

Arguments

point: The point estimate (measurement of central tendency)
var1, var2: The first and second measures of dispersion (`var2` is not required for all imputation methods)
n: The number of observations contributing to the measurement
vartype: The type of dispersion (case insensitive; see Details)

Value

A vector of imputed standard deviations

Details

impute_sd selects the imputation method based on vartype.

impute_sd_sd() imputes the standard deviation from itself (not really imputation). var1 is the standard deviation; var2 must be NA; vartype must be "SD" (ignoring case); and n is ignored.

impute_sd_cv() imputes the standard deviation from the coefficient of variation (CV). var1 is the CV; var2 must be NA; vartype must be "CV", "%CV", or "CV%" (ignoring case); and n is ignored.

impute_sd_se() imputes the standard deviation from the standard error (SE). var1 is the standard error of the mean; var2 must be NA; vartype must be "SE" or "SEM" (ignoring case); and n is required. Imputation assumes a sample, not population, SE.

impute_sd_ci() imputes the standard deviation from the confidence interval (CI). var1 is the lower bound of the CI; var2 is the upper bound of the CI; vartype must match the regular expression "^([0-9]+)% CI$" (ignoring case); and n is required. If only an upper or lower bound of the CI is available, set the other to NA. Imputation assumes a t-distribution.

impute_sd_iqr() imputes the standard deviation from the inter-quartile range (IQR). var1 is the lower bound of the IQR; var2 is the upper bound of the IQR; vartype must be "IQR" (ignoring case); and n is required. Imputation assumes a t-distribution.

impute_sd_range() imputes the standard deviation from the range. var1 is the lower bound of the range; var2 is the upper bound of the range; vartype must be "RANGE" (ignoring case); and n is required. Imputation is performed by computing (var2 - var2)/4. This should be the last choice of imputation methods.

References

Cochrane Handbook, version 5.1 https://handbook-5-1.cochrane.org/chapter_7/7_7_3_data_extraction_for_continuous_outcomes.htm