psychmeta (version 0.2.4)

correct_r_dich: Correct correlations for artifical dichotomization of one or both variables

Description

Correct correlations for artifical dichotomization of one or both variables

Usage

correct_r_dich(r, p_x = NA, p_y = NA, n = NULL)

Arguments

r

Vector of correlations attenuated by artificial dichomization.

p_x

Vector of proportions of the distribution on either side of the split applied to X (set as NA if X is continuous).

p_y

Vector of proportions of the distribution on either side of the split applied to Y (set as NA if Y is continuous).

n

Optional vector of sample sizes.

Value

Vector of correlations corrected for artificial dichomization (if n is supplied, corrected error variance and adjusted sample size is also reported).

Details

$$r_{c}=\frac{r_{obs}}{\left[\frac{\phi\left(p_{X}\right)}{p_{X}\left(1-p_{X}\right)}\right]\left[\frac{\phi\left(p_{Y}\right)}{p_{Y}\left(1-p_{Y}\right)}\right]}$$

References

Schmidt, F. L., & Hunter, J. E. (2015). Methods of meta-analysis: Correcting error and bias in research findings (3rd ed.). Thousand Oaks, CA: SAGE. https://doi.org/10/b6mg. pp. 43<U+2013>44.

Examples

# NOT RUN {
correct_r_dich(r = 0.32, p_x = .5, p_y = .5, n = 100)
# }