irt (version 0.2.9)

biserial: Calculate biserial correlation

Description

Calculate biserial correlation

Usage

biserial(score, criterion, method = "default")

Value

Biserial correlation value

Arguments

score

Item scores of each examinee for which biserial correlation will be calculated

criterion

Total score of each examinee

method

Type of the biserial correlation calculation method.

"default"

The most common way to calculate biserial correlation.

"point-biserial"

Calculate point-biserial correlation.

"clemans-lord"

Modified biserial correlation value based on Clemans (1958) and Lord (1962).

"brogden"

Modified biserial correlation value based on: Brogden (1949)

"rank"

Rank biserial correlation value based on Cureton (1968).

Author

Emre Gonulates

References

Brogden, H. E. (1949). A new coefficient: Application to biserial correlation and to estimation of selective efficiency. Psychometrika, 14, 169-182.

Clemans, W. V. (1958) An index of item-criterion relationship. Educational and Psychological Measurement, 18, 167-172.

Cureton, E. E. (1968). Rank biserial correlation when ties are present. Educational and Psychological Measurement, 28, 77-79.

Kraemer, H. C. (1981). Modified biserial correlation coefficients. Psychometrika, 46(3), 275-282.

Lord, F. M. (1963). Biserial estimates of correlation. Psychometrika, 28, 81–85.

Examples

# The example is from Salkind, Rasmussen (2007) Encyclopedia of measurement
# and statistics, pages 94-97
score <- c(rep(0, 16), rep(1, 22))
total_score <- c(87, 90, 94, 94, 97, 103, 103, 104, 106, 108, 109, 109, 109,
                 112, 119, 132, 100, 103, 103, 106, 112, 113, 114, 114, 118,
                 119, 120, 120, 124, 133, 135, 135, 136, 141, 155, 157, 159,
                 162)
# Calculate biserial correlation
biserial(score, total_score)
# Calculate point-biserial correlation
biserial(score, total_score, method = "point-biserial")
# Calculate modified biserial correlation (based on Brogden (1949))
biserial(score, total_score, method = "brogden")
# Calculate modified biserial correlation (Clemans-Lord)
biserial(score, total_score, method = "clemans-lord")