Scores the Madden, Petry, & Johnson (2009) PDQ: three 10-item blocks
(block 1: $20 for sure vs. a chance of $80; block 2: $40 vs. $100;
block 3: $40 vs. $60), each an ascending ladder of h values at
indifference under the hyperbolic odds model V = A / (1 + h * theta),
theta = (1 - p) / p (Rachlin, Raineri, & Cross, 1991). Each block is
scored independently by the same consistency-maximization algorithm as
the MCQ scorers; this implementation reproduces the Gray et al. (2016)
scoring syntax lookup tables for every possible response pattern.
Usage
score_pdq(
dat = dat,
impute_method = "none",
round = 6,
random = FALSE,
trans = "none",
return_data = FALSE,
verbose = FALSE
)Arguments
- dat
Dataframe (longform) with subjectid, questionid (1-30), and response (0 for the smaller guaranteed reward, 1 for the larger risky reward)
- impute_method
One of: "none", "ggm", "GGM", "inn", "INN"
- round
Numeric specifying number of decimal places (passed to
base::round())- random
Boolean whether to insert a random draw (0 or 1) for NAs. Default is FALSE
- trans
Transformation to apply to h values: "none", "log", or "ln". Default is "none"
- return_data
Boolean whether to return the original data and new imputed responses. Default is FALSE.
- verbose
Boolean whether to print subject and question ids pertaining to missing data. Default is FALSE.
Value
If return_data = FALSE (default), a summary data frame with one
row per subject: pooled overall_h (see Details), per-block h
(block1_h, block2_h, block3_h), their arithmetic mean (mean_h;
Gray et al.'s recommended composite) and geometric mean (geomean_h),
pooled and per-block consistency plus their mean
(composite_consistency), and pooled plus per-block proportions of
risky choices (block*_proportion is Gray et al.'s risky choice ratio).
If return_data = TRUE, a list with results and data (the input plus
a newresponse column reflecting any imputation).
Details
Each subject's data must satisfy a strict contract: exactly one
row per canonical question id (30 of them; no duplicates, no unknown
ids, none missing) and responses coded 0, 1, or NA (numeric, logical,
or character/factor values that coerce to 0/1). Malformed input errors
rather than silently mis-scoring. Contrast with pdq_to_choice()'s
lenient, ragged contract.
The published scoring (Madden et al., 2009; Gray et al., 2016) has no
overall 30-item ladder; blocks are scored separately, and mean_h is
Gray et al.'s recommended composite. overall_h and
overall_consistency are a beezdiscounting extension: all 30 items are
pooled into a single ascending ladder (exact-rational h order, ties
broken by question id; four item pairs tie exactly) and scored by the
same consistency-maximization algorithm with the repeat-last edge.
The tied pairs are questions 1 and 21 (h = 1/3), 4 and 24 (h = 0.75),
5 and 15 (h = 1), and 16 and 26 (h = 1.5). When a subject answers the two
items of a tied pair differently, the question-id tie-break decides which
response sits lower on the pooled ladder, so overall_consistency (and
possibly overall_h) can depend on that order. The per-block scores have
no ties and are unaffected.
INN imputation groups items sharing an h rank (one item per block).
Ladders with remaining NA responses score NA; Gray et al. recommend
excluding subjects below 80% consistency on any block.
References
Madden, G. J., Petry, N. M., & Johnson, P. S. (2009). Pathological gamblers discount probabilistic rewards less steeply than matched controls. Experimental and Clinical Psychopharmacology, 17(5), 283-290. doi:10.1037/a0016806
Gray, J. C., Amlung, M. T., Palmer, A. A., & MacKillop, J. (2016). Syntax for calculation of discounting indices from the monetary choice questionnaire and probability discounting questionnaire. Journal of the Experimental Analysis of Behavior, 106(2), 156-163. doi:10.1002/jeab.221
Examples
score_pdq(pdq)
#> subjectid overall_h block1_h block2_h block3_h mean_h geomean_h
#> 1 1 1.224745 1.215571 1.224745 1.233988 1.224768 1.224745
#> 2 2 0.329268 0.333333 0.329268 0.333333 0.331978 0.331973
#> overall_consistency block1_consistency block2_consistency block3_consistency
#> 1 1 1 1 1
#> 2 1 1 1 1
#> composite_consistency overall_proportion block1_proportion block2_proportion
#> 1 1 0.5 0.5 0.5
#> 2 1 1.0 1.0 1.0
#> block3_proportion impute_method
#> 1 0.5 none
#> 2 1.0 none