stonegold546 / cohens_d_calculators Goto Github PK

View Code? Open in Web Editor NEW

11.0 1.0 1.0 297 KB

Effect size calculator within a number of contexts

Home Page: https://effect-size-calculator.herokuapp.com/

License: GNU General Public License v3.0

Ruby 27.01% JavaScript 38.06% CSS 0.06% Slim 34.83% Procfile 0.04%

confidence-intervals effect-sizes calculator cohens-d hedges-g cohen equation icc hlm mlm

cohens_d_calculators's People

Contributors

Stargazers

Watchers

Forkers

arete0

cohens_d_calculators's Issues

There is a corrected formula for paired tests

There is a corrected formula for paired tests, dividing the d obtained by sqrt(1-r). See Cohen J, 1988, p.49, at this link
http://www.utstat.toronto.edu/~brunner/oldclass/378f16/readings/CohenPower.pdf#page=66

Add generalized eta squared

Use python package pyvttbl and use bootstrapping to generate CI.

Be sure to round all results to 7 decimal places

.round(7)

Add effect sizes in ANOVA context

Partial eta-squared
Omega-squared
90% confidence intervals

Let user choose whether S&B or N&S

Snijders and Bosker require ML
- Return tau's, sigma^2's, ICC's
Nakagawa and Schielzeth ideally require REML
- Add var_fixed, only residual ICC
Let user choose which R^2 they want or both fit using ML

Add RR, AR, NNT

relative risk
absolute risk
number needed to treat

Add effect sizes in HLM context

With confidence intervals

Usability

Add clear buttons for input and results
Add go-to's for positions on page

Add cohen's dz for paired samples.

Dear stonegold546,

thanks for providing this great tool. I just wanted to recommend that you could add the standard d-value (d_z) often used (even if not recommended by Lakens) for paired measurements as stated in the beginning in the d section of Lakens article:
https://www.frontiersin.org/articles/10.3389/fpsyg.2013.00863/full