Thank you for the opportunity to respond to a letter submitted by M. C. Hodge and T. M. Bach relative to our 1997 publication. The authors raise several points to which we would like to briefly respond.
As is suggested in the title of the paper (i.e., “A Preliminary Investigation”), this work was meant to stimulate interest in this area and hopefully encourage researchers to replicate, explore, and expand on the ideas we put forth. It was not intended to be a definitive source of information. The article was submitted following the normal peer-review process set forth by the journal, and the process of scientific review was followed.
Hodge and Bach raise the point of their inability to replicate statistical results reported using the mean data presented in the original manuscript, Table C. Our reevaluation using the data in Table C and a repeated-measures analysis of variance are: VO2, p = .05; HR, p = .39 (we reported no differences); RPE, p = .02 (S1, S2, S4, S5, and S6). As noted in the text, subject no. 1 was dropped from the physiological analysis because the VO2 value was below instrument accuracy detection level. Subject no. 3 was also dropped from the physiological analysis because that person could not tolerate walking without the use of the orthotic device. We did use RPE data from subject no. 1 because the data obtained represented the subject’s perceived exertion during walking with and without the orthotic device. We did not use subject no. 3’s RPE data because that person was unable to complete the treadmill walking protocol (could only tolerate cycling without the orthotic device). Regarding the RPE analysis, we agree the text could have been more succinct in explaining which subjects were included/excluded from the analyses.
The authors raise another statistical issue, that being their opinion of the inappropriateness of the statistical test (repeated-measures analysis of variance) for a small sample size. We agree conceptually with this point and, as stated in the original manuscript, addressed this issue with use of a single-subject (SS) analysis technique (Model Statistic)1 rather than a nonparametric test. Results from the SS analysis are discussed, by subject, in the original manuscript.
Relative to correlation results, the authors correctly point out an error in the reported correlation between VO2 and heart rate (r = 0.848 vs. the reported value of 0.884). This was clearly a typographical error as is evidenced in the remainder of the paragraph, which reports the explained variance associated with this correlation correctly (for r = 0.848) of 72%. In our original work, we did not examine the level of significance of the correlations, because we were more interested in the magnitude of the shared relationship (explained variance) between parameters examined. The correlation between VO2 and RPE was interpreted correctly in the manuscript; however, as Hodge and Bach state, it should have carried a negative sign.
Finally, in discussing the results obtained from our work, we suggested an explanation for our observations of improved locomotion ability while wearing the orthotic device—out-of-plane motion. This conjecture can (and should) be tested empirically using a research plan such as the one outlined by Hodge and Bach. Such a study would contribute greatly to the knowledge base in the design and application of lower extremity orthotic appliances.
Janet S. Dufek, PhD, FACSM
Human Performance and Wellness, Inc.
1.Bates BT. Single-subject methodology: An alternative approach. Med Sci Sports Exerc