Name:

Description:

In the second issue of Biometrika, W. R. Macdonell (1902) published an extensive paper, On Criminal Anthropometry and the Identification of Criminals in which he included numerous tables of physical characteristics 3000 non-habitual male criminals serving their sentences in England and Wales. His Table III (p. 216) recorded a bivariate frequency distribution of height by finger length. His main purpose was to show that Scotland Yard could have indexed their material more efficiently, and find a given profile more quickly.

W. S. Gosset (aka "Student") used these data in two classic papers in 1908, in which he derived various characteristics of the sampling distributions of the mean, standard deviation and Pearson's r. He said, "Before I had succeeded in solving my problem analytically, I had endeavoured to do so empirically." Among his experiments, he randomly shuffled the 3000 observations from Macdonell's table, and then grouped them into samples of size 4, 8, ..., calculating the sample means, standard deviations and correlations for each sample.

Variables:

Macdonell: A frequency data frame with 3000 observations on the following 3 variables giving the bivariate frequency distribution of height and finger.

height

lower class boundaries of height, in decimal ft.

finger

length of the left middle finger, in mm.

frequency

frequency of this combination of height and finger

MacdonellDF: A data frame with 3000 observations on the following 2 variables.

height

a numeric vector

finger

a numeric vector

Details

Class intervals for height in Macdonell's table were given in 1 in. ranges, from (4' 7" 9/16 - 4' 8" 9/16), to (6' 4" 9/16 - 6' 5" 9/16). The values of height are taken as the lower class boundaries.

For convenience, the data frame MacdonellDF presents the same data, in expanded form, with each combination of height and finger replicated frequency times.

Link To Google Sheets:

Rows:

Columns:

Files:

License Type:

References/Notes/Attributions:

Source

Macdonell, W. R. (1902). On Criminal Anthropometry and the Identification of Criminals. Biometrika, 1(2), 177-227. doi:10.1093/biomet/1.2.177 http://www.jstor.org/stable/2331487

The data used here were obtained from:

Hanley, J. (2008). Macdonell data used by Student. http://www.medicine.mcgill.ca/epidemiology/hanley/Student/

References

Hanley, J. and Julien, M. and Moodie, E. (2008). Student's z, t, and s: What if Gosset had R? The American Statistican, 62(1), 64-69.

Gosset, W. S. [Student] (1908). Probable error of a mean. Biometrika, 6(1), 1-25. http://www.york.ac.uk/depts/maths/histstat/student.pdf

Gosset, W. S. [Student] (1908). Probable error of a correlation coefficient. Biometrika, 6, 302-310.

R Dataset Upload:

Use the following R code to directly access this dataset in R.

R Coding Interface: