ttestBF() from the BayesFactor package throws the following error when trying to analyse data with missing values:
> ttestBF(formula = Level2 ~ Completers, r = sqrt(2)/2, data = tib)
Error in checkFormula(formula, data, analysis = "indept") :
Dependent variable must not contain missing or infinite values.
Despite being mentioned in its documentation, adding argument is.na to the ttestBF() command does not change anything (but perhaps it's my fault because I don't know how to apply it):
> ttestBF(formula = Level2 ~ Completers, r = sqrt(2)/2, is.na = TRUE, data = tib)
Error in checkFormula(formula, data, analysis = "indept") :
Dependent variable must not contain missing or infinite values.
In addition: Warning message:
data coerced from tibble to data frame
My questions:
ttestBF() from the BayesFactor package?Reproducible data:
nn <- 100 # total sample size (both Completers and Non Completers)
# create tibble with 2 level samples (completers & non completers), 50 each
tib <- tibble::tibble(Completers = rep(c(1, 0), each = nn/2),
Level = c(sample(1:9, 50, replace = TRUE,
prob = c(57/1043*100, 275/1043*100,
398/1043*100, 199/1043*100,
72/1043*100, 27/1043*100,
12/1043*100, 2/1043*100,
1/1043*100)),
sample(1:9, 50, replace = TRUE,
prob = c(57/1043*100, 275/1043*100,
398/1043*100, 199/1043*100,
72/1043*100, 27/1043*100,
12/1043*100, 2/1043*100,
1/1043*100))))
# Value labels
expss::val_lab(tib$Completers) = expss::num_lab("
1 Completers
0 Non Completers")
expss::val_lab(tib$Level) = expss::num_lab("1 A
2 B
3 C
4 D
5 E
6 F
7 G
8 H
9 I")
# Copy 2nd column into 3rd and call it Level2
tib$Level2 <- tib$Level
# Insert single missing value (NA)
tib[3, 2] = NA
Output with no missing values:
> BayesFactor::ttestBF(formula = Level ~ Completers, r = sqrt(2)/2, data = tib)
Bayes factor analysis
--------------
[1] Alt., r=0.707 : 0.3507102 ±0.02%
Against denominator:
Null, mu1-mu2 = 0
---
Bayes factor type: BFindepSample, JZS
Warning message:
data coerced from tibble to data frame
Output with missing values:
> BayesFactor::ttestBF(formula = Level2 ~ Completers, r = sqrt(2)/2, data = tib)
Error in checkFormula(formula, data, analysis = "indept") :
Dependent variable must not contain missing or infinite values.
In addition: Warning message:
data coerced from tibble to data frame
Output with missing values and argument is.na = TRUE:
> BayesFactor::ttestBF(formula = Level2 ~ Completers, r = sqrt(2)/2, is.na = TRUE, data = tib)
Error in checkFormula(formula, data, analysis = "indept") :
Dependent variable must not contain missing or infinite values.
In addition: Warning message:
data coerced from tibble to data frame
Sources:
sessioninfo::session_info() extract:
setting value
version R version 4.2.1 (2022-06-23)
os macOS Monterey 12.5.1
rstudio 2022.07.1+554 Spotted Wakerobin (desktop)
package * version date (UTC) lib source
BayesFactor * 0.9.12-4.4 2022-07-05 [1] CRAN (R 4.2.0)
expss * 0.11.1 2022-01-07 [1] CRAN (R 4.2.0)
tibble * 3.1.8 2022-07-22 [1] CRAN (R 4.2.0)
You can remove NA from your data:
BayesFactor::ttestBF(formula = Level2 ~ Completers, r = sqrt(2)/2, data = filter(tib, !is.na(Level2) & !is.na(Completers))