LICW had been teaching code at 20 WPM character speed with a high degree of Farnsworth spacing (see ANALYSES OF FARNSWORTH). We found plenty of research to support that practice. Lipmann in 1928 suggested the speed of transmission of individual characters should be held constant at about 2Ø WPM during learning, and that only the duration of intervals between characters and between words should be varied.
Experimental support for that suggestion was presented in 1932 by Biegel which held that at high speeds each individual character has a distinctive rhythmic pattern and when sending at slower speeds individual characters are no longer perceived as organized patterns but are perceived analytically as dots and dashes. Biegel thought this breakdown in organization occurred at about 16 WPM and she too sought to avoid this by sending individual characters at 25 WPM with increased spacing yielding an effective speed of 8 WPM.
In 1936 Koch conducted sending and receiving experiments in order to determine the optimum speed for learning CW. From a pure speed of learning perspective, character speed made no difference:
“Whether one transmits individual characters at high speeds or at slow speeds, during the progress of learning, makes no difference in the speed with which students learn to receive.”
Moreover, Koch determined higher character speed was less pleasing than learning the characters at a slower speed but eliminates the need to relearn the code upon reaching higher speeds:
“such a procedure tends to make the first stages of learning unduly hard and that, in many cases at least, students actually learn more rapidly when the characters are sent slowly from the beginning.””The advantage of learning individual characters initially at high speeds would not become evident until that speed of receiving had been reached where it was necessary to perceive the characters as total patterns.”
The findings were intriguing. Could student performance be increased by decreasing character speed thus making for a more pleasing learning experience?
Ultimately, Koch determined 12 WPM was optimum for initial learning. Anything less than 1Ø WPM did not allow the student to perceive the characters as total acoustic patterns. Koch maintained a tempo of 12 WPM with no enhanced character spacing until all characters were mastered, then gradually increased speed toward his ultimate goal of 20 WPM.
In 1943, Taylor carried out experiments at various character speeds. One group was sent code at 18 WPM with increased spacing yielding an effective speed of 3 to 6 WPM and the other group was sent code at 3 to 6 WPM. Contrary to what would be expected on the basis of the work of Lipmann, Biegel, and Koch, no significant difference in learning was found.
Despite LICW’s success teaching at a 20 WPM character speed with Farnsworth, two drawbacks were noted: (1) students had difficulty matching the 20 WPM character speed with their early straight key sending; and (2) 20 /5 was not representative of what students experienced when they first got on the air.
We adopted Koch’s recommended 12 WPM in the Beginners Carousel curriculum. After considerable testing and refinement, we settled on 12/8 in BC1, 12/10 in BC2, and 12/12 in Beginners Proficiency Training. This was a twofold increase in student proficiency over our previous curriculum.
Go Deeper: