percent.r: percent change to r

In phoebehlam/michaela: convert statistical metrics for meta-analytic purpose

percent change to r

Description

percent change derived from transformed exponentiated regression coefficients and confidence interval to correlation coefficient
percent change and ci -> expnentiated b and ci -> t -> r
unadjusted t-statistics estimate bivariate correlation coefficient (enter k = 1)
adjusted t-statistics (e.g., from a regression) estimate partial correlation coefficient (aloe & thompson, 2013).

Usage

percent.r(percent, cil, ciu, n, k, result = c("cil", "ciu", "avg"))

Arguments

percent	the percent change calculated from exponentiated logistic regression coefficient specifically computed such that percent change = 100*(exp(b) - 1); where b is the regression coefficient if percent change expressed in decimals, please *100 first or if percent change not computed with above equation, please compute back to exponentiated b and then use expb.r()
cil	the lower bound of the confidence interval around the percent change
ciu	the upper bound of the confidence interval around the percent change
n	the sample size
k	the total number of predictors, enter k = 1 if bivariate
result	one of the following: "ciu" uses the upper bound of ci to compute se (defaults if missing input) "cil" uses the lower bound of ci to compute se "avg" uses both upper bound and lower bound and then take the average of the resulting se's

Examples

percent.r(7, 3, 11, 3134, 6)
dat %>% mutate (r_from_perc = percent.r(percent, percent_lowerci, percent_upperci, percent_n, percent_predictors)) -> dat

phoebehlam/michaela documentation built on Oct. 23, 2024, 4:10 p.m.