ACCLAB / dabestr

Data Analysis with Bootstrap Estimation in R
https://acclab.github.io/dabestr
Apache License 2.0
215 stars 34 forks source link

Estimation plot for generalized linear mixed model with binomial outcome #131

Open TommasoCurro opened 2 years ago

TommasoCurro commented 2 years ago

Hello,

I am using dabestr package in R to create an estimation plot on correctResponses (1 correct, 0 incorrect). When I run the the model,

  dabest(condition, correctResponse
         ,idx = c("h", "s", "l")
         ,paired = TRUE
         ,id.col = participant
  )

model_ACC_exp1_meanDiff = mean_diff(model_ACC_exp1, reps = 5000)

I get this error:

"Error in if (const(t, min(1e-08, mean(t, na.rm = TRUE)/1e+06))) { : missing value where TRUE/FALSE is requested"

I tried adding more reps but the error is the same.

How do I get past this? Can dabestr create estimation plots for categorical outcomes?

Thank you!

the dataset for the model is as follow: Dataset: The first five rows are: Num_part condition correctResponse sequenceTrials 1 1 hands 1 GNG 2 1 hands 1 NGNG 3 1 hands 1 GNG 4 1 hands 1 GNG 5 1 hands 1 GNG

X Variable : condition Y Variable : correctResponse

Paired effect size(s) will be computed for:

  1. s minus h
  2. l minus h