Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


BGSU_Dissertations_0232_Curran.pdf (7.63 MB)

ETD Abstract Container

Abstract Header

The Use of Multiple Statistical Re-Analysis Procedures in Stuttering Research

Curran, Michael F.
http://rave.ohiolink.edu/etdc/view?acc_num=bgsu156629770206167

Abstract Details

1974, Doctor of Philosophy (Ph.D.), Bowling Green State University, Communication Disorders.
The present study was designed to validate the assumption that multiple data re-analysis procedures yield meaningful information and maximize the interpretability of stuttering data. A previously acquired body of stuttering data was used for this re-analysis. The original study was designed to investigate the effects of communicative stress on audible, inaudible and avoidance-escape components of stuttering. The original results were replicated through the use of the BMD02V univariate analysis of variance computer program and the Newman-Keuls procedure. Several other univariate and multivariate statistical procedures were employed for the re-analysis. Those procedures are as follows: Transformation of raw data to standard score data, Univariate trend analysis, Multivariate analysis of variance, Multiple discriminant analysis, Factor analysis, Multiple regression analysis, Canonical correlation and Power analysis. Results of the experiment replication and multiple re-analysis supported the original findings that imposed communicative stress affects stuttering behavior. However, additional information was obtained from the re-analysis procedure. Transformation of raw data to standard scores illustrated that data transformation may be used without altering interpretation of the test results. Trend analysis, both with raw data and standard score data, revealed that significant linear and cubic trend equations existed for the disfluency ratios M/S, TDF-T/M and SDF-T/S. Multivariate analysis of variance and multiple discriminant analysis revealed that the dependent variable vector, consisting of the eight disfluency ratios was significantly differentiated across the four conditions of communicative stress. Factor analysis (principal factors method) showed a single factor to exist after orthogonal rotation. This factor was labelled: A Disfluency Profile. Multiple regression and canonical correlation analyses indicated that the eight disfluency ratios were highly correlated. Power analysis revealed a statistical power of .43 to exist for a medium effect size. Research implications are discussed in terms of the theory of indications.
Stephen B. Hood (Advisor)
Raymond K. Tucker (Advisor)
Speech Therapy

Recommended Citations

Citations

  • Curran, M. F. (1974). The Use of Multiple Statistical Re-Analysis Procedures in Stuttering Research [Doctoral dissertation, Bowling Green State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=bgsu156629770206167

    APA Style (7th edition)

  • Curran, Michael. The Use of Multiple Statistical Re-Analysis Procedures in Stuttering Research. 1974. Bowling Green State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=bgsu156629770206167.

    MLA Style (8th edition)

  • Curran, Michael. "The Use of Multiple Statistical Re-Analysis Procedures in Stuttering Research." Doctoral dissertation, Bowling Green State University, 1974. http://rave.ohiolink.edu/etdc/view?acc_num=bgsu156629770206167

    Chicago Manual of Style (17th edition)

bgsu156629770206167
186
This document is provided by Bowling Green State University for research and educational purposes, and may be under copyright by the author or the author's heirs. Please contact the ETD Center administrator at Bowling Green State University etd@bgsu.edu with any questions or comments. In your email, be sure to include the URL and title of the specific items you are inquiring about.
This open access ETD is published by Bowling Green State University and OhioLINK.