CONVERT_ES {NO.PING.PONG}

Converts between r, d, g, and OR effect sizes

Description

Converts effect sizes, including (a) r to d, g, or OR, (b) d or g to r, d, g or OR, and (c) OR to r, d, or g.

Usage

CONVERT_ES(donnes, ES, ES_type_IN='r', ES_var = NULL, 
           totalN = NULL, grp1_n = NULL, grp2_n = NULL,
           gvar_type_OUT = 'd', 
           CI_level_out = 95,
           CI_level_in = 95, CI_in_lb = NULL, CI_in_ub = NULL, 
           verbose = TRUE)

Arguments

donnes

A dataframe or matrix with effect sizes and corresponding information, such as the effect size variances, sample sizes, confidence intervals.

ES

The name of the column in donnes with the effect sizes.

ES_type_IN

The type of effect sizes in ES. The options are 'r' (the default), 'z' for Fishers z transformation of r, 'd', 'g', and 'OR'.

ES_var

(optional) The name of the column in donnes with the variances of the effect sizes.

totalN

(optional) The name of the column in donnes with the total N for each study.

grp1_n

(optional) The name of the column in donnes with the Ns for group 1.

grp2_n

(optional) The name of the column in donnes with the Ns for group 2.

gvar_type_OUT

(optional) The kind of SMD variance. The options are 'd' or 'g'.

CI_level_out

(optional) The confidence interval for the output (in whole numbers). The default is 95 (for 95 percent).

CI_level_in

(optional) The confidence interval for the input (in whole numbers). The default is 95 (for 95 percent).

CI_in_lb

(optional) The name of the column in donnes with the lower bound confidence intervals for input, if provided (in whole numbers).

CI_in_ub

(optional) The name of the column in donnes with the upper bound confidence intervals for the input, if provided (in whole numbers).

verbose

(optional) Should detailed results be displayed in console? TRUE (default) or FALSE.

Details

This function converts r, z (Fishers z transformation of r), d, g, and OR effect sizes to r, z, d, g, and OR effect sizes using conventional formulas (Borenstein & Hedges, 2019; Borenstein, Hedges, Higgins, & Rothstein, 2009). The effect size variances and confidence intervals are also computed if sufficient data are provided as input.

When the input effect sizes are d or g values, it is helpful if the group Ns are also provided whenever possible.

The gvar_type_OUT argument provides a choice between d or g effect size variances whenever ES_type_OUT is set to g. The reason for this option is that authors of published meta-analyses sometimes report d variances when their analyses were conducted on g effect sizes. This is presumably not a wise practice, but it also does not make much difference in the computed values.

Value

A matrix containing the following statistics, if they can be computed using the input informaiton: the r, z, d, g, and OR effect sizes and their corresponding variances, confidence intervals, and the totalNs used in the analyses.

Author(s)

Brian P. O'Connor

References

Borenstein, M., & Hedges, L. V. (2019). Effect sizes for meta-analysis. In H. Cooper, L. V., Hedges, & J. C. Valentine (Eds). The handbook of research synthesis and meta-analysis (pp. 207-244). (3rd. edition). New York, NY: Russell Sage Foundation.

Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Converting among effect sizes. In, Introduction to meta-analysis pp. 45-49. Chichester, UK: John Wiley & Sons.

Valentine, J. C. & Cooper, H. (2003). Effect size substantive interpretation guidelines: Issues in the interpretation of effect sizes. Washington, DC: What Works Clearinghouse.

Examples

# convert d effect sizes to r	
head(data_NPP$Omega3_Depression)
CONVERT_ES(donnes = data_NPP$Omega3_Depression, ES = 'SMD', ES_type_IN='d', ES_var = NULL, 
           grp1_n = 'CN', grp2_n = 'EN')

 
# convert r effect sizes to g	
head(data_NPP$Math_Performance)
CONVERT_ES(donnes = data_NPP$Math_Performance, ES = 'r', ES_type_IN='r', ES_var = NULL,
           totalN = 'N')