ARC_total_intrude: Adjusted Ratio of Clustering (ARC) - Total Recall / Adjust...

In ggreeley/MemoryOrg: (A Few) Retrieval Organization Metrics for Human Memory Research

Description

Adapted from Roenker, Thompson, & Brown (1971). Computes ARC using the total, uncorrected recall (N) of an individual or group. Also produces several componets of ARC (E(r), N, Repetitions, and Max). DIFFERS from ARC_total() because this "adjusts" Max and E(r) by omitting intrusions (coded as blank cells - so NA) as a category. See below "Use when" for explanation.

Use when:

You want all items, correct (1) and incorrect (0), counted toward N - AND

There ARE intrusions (code as blank cells - NA) present in data - AND

You DO NOT want intrusions to count as their own category.

Adjusted Max = N - c (when N = total recall (1's and 0's) and c = number of categories recalled NOT counting intrustions NA)

Adjusted E(r) = [ (∑ni^2) / N ] − 1 (when intrusions (NA) do NOT contribute to ni as a category, while N is still total recal)

See "details" for more on the formula, notation definitions, and further explanation.

Usage

ARC_total_intrude(correct, category)

Arguments

correct	Numeric vector representing correct (1) and incorrect (0) responses.
category	Numeric or character vector representing category membership of associated item in x column.

Details

ARC = [r − E(r)] / [max − E(r)]

E(r) = [ (∑ni^2) / N ] − 1

r = number of category repetitions

N = number of recalled items (correct and incorrect)

c = number of recalled categories

max = N − c and

ni = number of items recalled in each recalled category

Adjusted Max = N - c (when N = total recall (1's and 0's) and c = number of categories recalled NOT counting intrustions NA)

Adjusted E(r) = [ (∑ni^2) / N ] − 1 (when intrusions (NA) do NOT contribute to ni as a category, and N is still total recall)

Adjusted ARC is as above, only using Adjusted Max and Adjusted E(r) in the formula

General - This will count category repetitions (NOT including back-to-back intrusions (blank cells)) and the total N is used in all computations (max and E(r)). Be sure data is properly processed (according to your protocol) before using.

Author(s)

Garrett. D Greeley

Ph.D. Student - Cognitive Science

Stony Brook University, NY

References

Roenker, D. L., Thompson, C. P., & Brown, S. C. (1971). Comparison of measures for the estimation of clustering in free recall. Psychological Bulletin, 76(1), 45–48. https://doi.org/10.1037/h0031355

See Also

ARC_total()

ARC_correct()

ARC_correct_intrude()

Examples

Data adapted from Roenker et al. (1971) - Also see ARC_3SE and other example data in the package for more examples and formatting.

recall_correct <- c(1,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1)
recall_category <- c("a","a", NA ,"a","a","b","a","c","b","b","b","b","c","b","c","c","c","c")

ARC_total_intrude(recall_correct, recall_category)

ggreeley/MemoryOrg documentation built on Aug. 8, 2020, 10:43 p.m.