Even worse, there is clear evidence that the results were altered through a series of inappropriate data manipulations. For example, the initial typing scores for the QWERTY typists were measured differently from the initial scores of the Dvorak typists so as to greatly disadvantage the QWERTY results. The report states that, because three typists in the QWERTY group had initial net scores of zero words per minute (!), the beginning and ending speeds were calculated as the average of the first four typing tests and the average of the last four typing tests. This has the effect of raising the measured initial typing speed, and lowering the measured ending speed. In contrast, the initial experiment using Dvorak simply used the first and last test scores. Using numbers reported in the footnotes of the report, we were able to calculate that this truncation of the reported values at the beginning of the test reduced the measured increase in typing speed on the QWERTY keyboard by almost half. The effect of the truncation at the end of the measuring period also decreases the reported gains for the QWERTY typists, though the size of this distortion cannot be determined from the report. The important thing, however, is that the numbers appear to be cooked in favor of Dvorak.