Open djmeter opened 2 years ago
Where do you see discrepancies?
esc::esc_2x2(
grp1yes = 205,
grp1no = 98,
grp2yes = 512,
grp2no = 178,
es.type = "or"
)
#>
#> Effect Size Calculation for Meta Analysis
#>
#> Conversion: 2x2 table (OR) coefficient to effect size odds ratio
#> Effect Size: 0.7272
#> Standard Error: 0.1505
#> Variance: 0.0227
#> Lower CI: 0.5415
#> Upper CI: 0.9768
#> Weight: 44.1438
Created on 2022-02-24 by the reprex package (v2.0.1)
And for logits:
esc::esc_2x2(
grp1yes = 205,
grp1no = 98,
grp2yes = 512,
grp2no = 178
)
#>
#> Effect Size Calculation for Meta Analysis
#>
#> Conversion: 2x2 table (OR) to effect size logits
#> Effect Size: -0.3185
#> Standard Error: 0.1505
#> Variance: 0.0227
#> Lower CI: -0.6135
#> Upper CI: -0.0235
#> Weight: 44.1438
Created on 2022-02-24 by the reprex package (v2.0.1)
The discrepancy was in the r calculation, not OR. Thanks!
The calculator returns RR, not r.
Ok, the esc package documentation says differently, but it is helpful to know that. Thank you. Diana
Ok, the esc package documentation says differently
Not sure what you mean here? The function returns the effect size that is given in the es.type
argument, which is by default logits
. The web-based 'Practical Meta-Analysis Effect Size Calculator' returns OR, which you can also return from esc_2x2()
, and the results are the same.
I also encountered this problem. esc_2x2(23,29,10,3, es.type="r")gave r = -0.3019; but the website showed different results.
Furthermore, esc_2x2(23,29,10,3, es.type="r")give different result from esc_2x2(23,10,29,3, es.type="r").
I ran the following code and got a different result than with the web-based 'Practical Meta-Analysis Effect Size Calculator' from David B. Wilson (http://www.campbellcollaboration.org/escalc/html/EffectSizeCalculator-Home.php)
esc_2x2( grp1yes = 205, grp1no = 98,
grp2yes = 512, grp2no = 178,
es.type = c("r") )
Could you please let me know why the formula is different? Thank you! Diana