A Comparison of the High Relief Finger Maze and the Stylus Maze

The methodology of experimental studies of human learning and retention. I. The functions of a methodology and the available criteria for evaluating different experimental methods.

“…In the case of human learning, measures of this type have been used by Langdon (62) for evaluating the reliability of a simple motor learning test, by Nyswander (87) for comparing the reliability of stylus and finger mazes, and by Husband (S3) for comparing the reliability of maze measures obtained with rats and human subjects. The significance of the method has been severely criticized by Leeper (64) and Spence (110), but the special case of the correlation between the total errors or time in the first and second halves of the learning period has been considered as one of the few valid methods by Tolman and Nyswander (123) and Nyswander (87).…”

“…In spite of this possibility that the scores obtained during the trials in the first and last halves of the learning period may represent different abilities, Tolman and Nyswander (123) and Nyswander (87) have maintained that the method is adequate whenever " piecemeal" learning problems are employed. Likewise, Spence (110) has suggested that the exaggeration of the r due to the correlation between errors of measurement and the attenuation attributable to the measurement of different abilities may balance each other and give a fairly accurate estimate of the reliability of the maze that is not of the " piecemeal" type.…”

“…There is another danger in using this r as an index of reliability. Without an independent criterion of the type of learning represented, the use of the r between test-halves in selecting materials and methods for experimentation, as Nyswander has done (87), implies that the investigator should use only those mazes, lists of verbal material, etc., that are learned in piecemeal fashion. This seems to be an illustration of Willoughby's point (139, p. 161) that the application of statistical criteria, which assume additive identical units, to problems of behavior sometimes ends in " the baby .…”

“…The method seems to have been introduced by Heron (41) and has received explicit approval for use in learning studies by Tolman and Nyswander (123), Nyswander (87), Leeper (64), Anastasi (4), and Tryon (130). Tryon (130, p. 154) not only accepts the method, but has argued that " for such experimentally determined material as scores on successive trials in a maze, Spearman devised the method of collecting odd elements in one set, and even elements in the other set (the elements may be trials or blinds), correlating these two independent sets, and then applying the correction which eventually became known as 'Brown's formula' (sic).…”

Factors Affecting the Reliability of the Maze: A Comparison of the High-Relief Finger Maze and an Improved form of the Stylus Maze

“…Nyswander (6) has made a statistical comparison between the high-relief finger maze and the old-form stylus maze, a comparison which is quite unfavorable to the stylus maze. She finds the finger maze much less di5cult.…”