Configure single-choice scoring

Create a scoring specification for singleChoice() or a single-choice Rmd item. multiple_choice_scoring() documents the separate checkbox models. The singleChoice() item default remains scoring_scheme = "standard". This helper defaults to method = "custom" for convenient calls such as single_choice_scoring(incorrect = -0.2). With no settings it produces correct/incorrect/omitted coefficients of 1/0/0, matching standard scoring.

Usage

single_choice_scoring(method = "custom", ...)

Arguments

method: A single-choice method. The default "custom" accepts explicit credits. "standard" awards full points for the correct answer and zero otherwise; "penalty" deducts points/(k-1) for an incorrect answer, where k is the number of options, and awards zero for an omission. "kanzow_01" through "kanzow_21" use the coefficients below. Aliases "number_right", "fair_penalty", "correction_for_guessing", and "rights_minus_wrongs" select methods 1, 3, 6, and 13. "exams" uses the R/exams single-choice penalty setting.
...: Settings for the selected method, expressed as fractions of the item's points. For custom: correct (default 1), incorrect (default 0), and omitted (default 0). All must be finite; correct must be positive and at least as large as the other credits. For exams: negative (default FALSE), a logical or finite numeric penalty magnitude, following exams::exams_eval(). negative = FALSE means no penalty, TRUE deducts the item's full points for an incorrect answer, and a numeric value d deducts abs(d) * points. Correct answers receive full points; omissions receive zero. Thus single_choice_scoring("exams", negative = 0.25) is equivalent to single_choice_scoring("custom", incorrect = -0.25). This is an R/exams parameter convention, not another scoring formula. Unlike "penalty", it does not automatically calculate 1/(k-1). The partial and rule settings belong to multiple-choice scoring and are not accepted here. For kanzow_21: the required logical statement_truth, explained below.

Value

A named list for the scoring_scheme argument of singleChoice(). It is checked against the item type; passing it to multipleChoice() is an error. Plain YAML mappings are interpreted in the item's own context.

Rmd/YAML

The same settings can be supplied as a YAML mapping:

scoring_scheme:
  method: custom
  correct: 1
  incorrect: -0.2
  omitted: 0

With points: 2, this awards 2, -0.4, and 0 points. A preset can also be supplied as a string, for example scoring_scheme: kanzow_05. The number of options is inferred from the item, not passed to this helper.

Restrictions and score bounds

Methods 10, 12, 14, 15, 17, 18, and 20 require exactly two options. Method 21 requires a single true/false statement with two answer options. Invalid option counts raise errors when an item is constructed or an Rmd file converted. Method 21 also requires statement_truth immediately. points scales the published coefficients. Thus methods 4, 7, and 8 can have a maximum different from points; getPoints() reports the actual maximum. Negative single-choice credits remain negative, without a zero floor. Supporting a preset does not recommend it for every assessment.

What statement_truth means

Only kanzow_21 needs this setting. statement_truth = TRUE means that the assertion posed to the student is actually true, for example "2 + 2 = 4". FALSE means it is actually false, for example "2 + 2 = 5". It does not describe the student's response or the position of the correct option. For a one-point item, an incorrect response scores -62/38 for a true assertion and -38/62 for a false assertion. Correct responses score 1; omissions score 0. Other single-choice methods do not need this setting. The correct option is still identified by singleChoice(solution = ...). rqti does not infer truth from option labels, which may be translated or contain HTML. The author must keep solution and statement_truth consistent.

Published coefficients

Published coefficients before multiplication by points:

Method	Correct	Incorrect	Omitted
`kanzow_01`	1	0	0
`kanzow_02`	1	0	1/k
`kanzow_03`	1	1 - 1/k	0
`kanzow_04`	1/(k-1)	0	0
`kanzow_05`	1	-1/(2(k-1))	0
`kanzow_06`	1	-1/(k-1)	0
`kanzow_07`	1/(k-1)	-1/(k-1)	0
`kanzow_08`	(k-1)/k	-1/k	0
`kanzow_09`	1	-1/3	0
`kanzow_10`	1	-0.48	0
`kanzow_11`	1	-0.5	0
`kanzow_12`	1	-0.6	0
`kanzow_13`	1	-1	0
`kanzow_14`	1	-1	0.7
`kanzow_15`	1	-1.1	0.7
`kanzow_16`	1	-k/(k-1)	0
`kanzow_17`	1	-1.5	0
`kanzow_18`	1	-1.8	0
`kanzow_19`	1	-2/(k-1)	0
`kanzow_20`	1	-3	0
`kanzow_21`	1	-62/38 if true; -38/62 if false	0

References

Kanzow AF, Schmidt D, Kanzow P (2023). Scoring Single-Response Multiple-Choice Items: Scoping Review and Comparison of Different Scoring Methods. JMIR Medical Education, 9, e44084. doi:10.2196/44084.

Examples

singleChoice(choices = c("A", "B", "C"), points = 2,
  scoring_scheme = single_choice_scoring(incorrect = -0.2))
#> An object of class "SingleChoice"
#> Slot "solution":
#> [1] 1
#> 
#> Slot "scoring_scheme":
#> $method
#> [1] "custom"
#> 
#> $item_type
#> [1] "single"
#> 
#> $incorrect
#> [1] -0.2
#> 
#> 
#> Slot "choices":
#> [1] "A" "B" "C"
#> 
#> Slot "choice_identifiers":
#> [1] "ChoiceA" "ChoiceB" "ChoiceC"
#> 
#> Slot "response_identifier":
#> [1] "RESPONSE"
#> 
#> Slot "shuffle":
#> [1] TRUE
#> 
#> Slot "orientation":
#> [1] "vertical"
#> 
#> Slot "identifier":
#> [1] "id_task_7748"
#> 
#> Slot "title":
#> [1] "id_task_7748"
#> 
#> Slot "content":
#> list()
#> 
#> Slot "prompt":
#> [1] ""
#> 
#> Slot "points":
#> [1] 2
#> 
#> Slot "feedback":
#> list()
#> 
#> Slot "files":
#> [1] NA
#> 
#> Slot "stylesheet_path":
#> character(0)
#> 
#> Slot "css":
#> character(0)
#> 
#> Slot "calculator":
#> [1] NA
#> 
#> Slot "metadata":
#> An object of class "QtiMetadata"
#> Slot "contributor":
#> [[1]]
#> An object of class "QtiContributor"
#> Slot "name":
#> [1] ""
#> 
#> Slot "role":
#> [1] "author"
#> 
#> Slot "contribution_date":
#> Date of length 0
#> 
#> 
#> 
#> Slot "description":
#> [1] ""
#> 
#> Slot "rights":
#> [1] ""
#> 
#> Slot "version":
#> character(0)
#> 
#> 
singleChoice(prompt = "2 + 2 = 4", choices = c("False", "True"), solution = 2,
  scoring_scheme = single_choice_scoring("kanzow_21", statement_truth = TRUE))
#> An object of class "SingleChoice"
#> Slot "solution":
#> [1] 2
#> 
#> Slot "scoring_scheme":
#> $method
#> [1] "kanzow_21"
#> 
#> $item_type
#> [1] "single"
#> 
#> $statement_truth
#> [1] TRUE
#> 
#> 
#> Slot "choices":
#> [1] "False" "True" 
#> 
#> Slot "choice_identifiers":
#> [1] "ChoiceA" "ChoiceB"
#> 
#> Slot "response_identifier":
#> [1] "RESPONSE"
#> 
#> Slot "shuffle":
#> [1] TRUE
#> 
#> Slot "orientation":
#> [1] "vertical"
#> 
#> Slot "identifier":
#> [1] "id_task_0387"
#> 
#> Slot "title":
#> [1] "id_task_0387"
#> 
#> Slot "content":
#> list()
#> 
#> Slot "prompt":
#> [1] "2 + 2 = 4"
#> 
#> Slot "points":
#> [1] 1
#> 
#> Slot "feedback":
#> list()
#> 
#> Slot "files":
#> [1] NA
#> 
#> Slot "stylesheet_path":
#> character(0)
#> 
#> Slot "css":
#> character(0)
#> 
#> Slot "calculator":
#> [1] NA
#> 
#> Slot "metadata":
#> An object of class "QtiMetadata"
#> Slot "contributor":
#> [[1]]
#> An object of class "QtiContributor"
#> Slot "name":
#> [1] ""
#> 
#> Slot "role":
#> [1] "author"
#> 
#> Slot "contribution_date":
#> Date of length 0
#> 
#> 
#> 
#> Slot "description":
#> [1] ""
#> 
#> Slot "rights":
#> [1] ""
#> 
#> Slot "version":
#> character(0)
#> 
#>