Of the 12 remaining studies of Table 6, two (Laurent, 1963; Merson, 1961) were reported as showing improved auding with no transfer to reading, while of the 10 remaining studies incorporating a post-training auding measure into their designs, all reported a significant improvement in auding on the basis of this measure, and all found an improvement in reading ability paralleling the improvement in auding ability (Veronese, 1960; Flederjohann, 1965; Cole, 1961; Hill, 1961; McDonnell, 1962; Kohls, 1965; Prater, 1965; Skiffington, 1965; Cooper, 1966; Thorn, 1968).

The two studies which failed to support the hypothesis (Laurent, 1963; Merson, 1961) were not available for full evaluation, and hence we have no means for understanding why they failed to support the hypothesis. Nonetheless, the weight of the evidence appears to be in favor of the hypothesis that training in comprehending by auding transfers to comprehending by reading. (No studies on reading to auding. transfer were found, although this should occur.)

The strength of the support provided by the 10 studies reporting transfer is increased by their diversity along two dimensions: grade level and training time. Reddin (1968) has previously voiced caution concerning the acceptance of auding-reading transfer without regard to the grade level of the subjects. This cautionary note appears less critical in light of the 10 studies supporting the transfer hypothesis. In these studies, subjects ranged from first-grade students (Cole, 1961; McDonnell, 1962; Thorn, 1968) to college freshmen (Hill, 1961). These studies also varied with respect to the duration of auding training. The shortest training period was approximately seven hours (Flederjohann, 1965) and the longest approximately 27 hours (Cooper, 1966).

While seven hours of instruction appears somewhat brief, the transfer hypothesis has been tested on the basis of far less training. Raper (1951) based his study on some two hours of training, while Hollingsworth used approximately 2 1/2 hours of training. The failure of training in these two studies might easily have been due to this factor.

A factor which appears important for the occurrence of transfer from auding training to reading ability is the similarity or degree of correspondence (Reddin,1968) between the knowledges and skills taught in the auding program and the knowledges and skills tested in the measure of reading ability.