To the Editor. —
Krieg et al1 advocate the use of a computerized tablet to extract data from published dot plot graphs when tables of data are not available. The implication is that this technique has sufficient accuracy to reconstruct data tables or at least to enumerate data points within intervals of the data distribution.Although the authors were not attempting to reproduce exact data values, they seem to have had difficulty defining the boundaries of their chosen intervals and reproducing the number of dots in each interval. In reviewing one of the original research articles,2 it is apparent that the number of counts by their computerized system is inaccurate. In particular, clinical state A should have 30 patients in the first interval and two in the second, whereas only 29 patients in the first interval and two in the second are shown. This error is repeated in
Simpson RJ, Amara IA. To Dot Plot Or Not. JAMA. 1989;261(22):3245. doi:10.1001/jama.1989.03420220059022
Customize your JAMA Network experience by selecting one or more topics from the list below.