A simulated dataset for evaluating diagnostic tests in the absence of a perfect gold standard. It includes results from multiple imperfect tests and, for simulation purposes, a 'true' disease status.
Usage
data(nogold_standard)Format
A data frame with 200 rows and 6 variables:
- caseID
Integer. Unique identifier for each case or patient.
- disease
Character. The true underlying disease status (e.g., "Present", "Absent"). In real "no gold standard" scenarios, this would be unknown and estimated.
- test1
Character. Result of the first diagnostic test (e.g., "pos", "neg", or "Positive", "Negative").
- test2
Character. Result of the second diagnostic test.
- test3
Character. Result of the third diagnostic test.
- test4
Character. Result of the fourth diagnostic test.
Examples
data(nogold_standard)
str(nogold_standard)
#> 'data.frame': 200 obs. of 6 variables:
#> $ caseID : int 1 2 3 4 5 6 7 8 9 10 ...
#> $ disease: chr "healthy" "disease" "healthy" "disease" ...
#> $ test1 : chr "pos" "neg" "pos" "pos" ...
#> $ test2 : chr "neg" "pos" "neg" "pos" ...
#> $ test3 : chr "pos" "pos" "neg" "neg" ...
#> $ test4 : chr "neg" "pos" "neg" "pos" ...
head(nogold_standard)
#> caseID disease test1 test2 test3 test4
#> 1 1 healthy pos neg pos neg
#> 2 2 disease neg pos pos pos
#> 3 3 healthy pos neg neg neg
#> 4 4 disease pos pos neg pos
#> 5 5 disease pos pos neg pos
#> 6 6 healthy neg pos pos neg
table(nogold_standard$test1, nogold_standard$test2)
#>
#> neg pos
#> neg 118 25
#> pos 26 31