Show simple item record

dc.contributor.authorKoh, Seong A
dc.creatorKoh, Seong Aen_US
dc.date.accessioned2011-12-05T21:59:17Z
dc.date.available2011-12-05T21:59:17Z
dc.date.issued2010en_US
dc.identifier.urihttp://hdl.handle.net/10150/193705
dc.description.abstractThe purpose of this study was to examine the similarity of outcomes from three different treatment integrity (TI) methods, and to identify the method which best corresponded to the assessment of a child’s behavior. Six raters were recruited through individual contact via snowball sampling. A modified intervention component list and 19 video clips were derived from Stahr’s (2005) study, “An Intervention for Children with Autism Spectrum Disorders (ASD) who Have Food Selectivity.” The raters, randomly and evenly assigned to three dyads. Each dyad received an average of six hours training and reached 85% interobserver agreement (IOA) with a 0.60 kappa score. After training, each dyad watched 5 video clips per day and measured both the child’s behavior and TI. The percentages of IOA, kappas, and indices of dependability for assessment of the child’s behavior and TI were analyzed. The data revealed that all raters reached over 80% IOA and the whole interval (WI) and yes/no (Y/N) dyads reached .60 kappa, but the two raters in the Likert-type (LIK) dyad could not reach .60 kappa. The indices of dependability indicated that the six raters consistently observed and rated both the child’s behavior and TI, but there was a discrepancy in scores (i.e., percentages of TI) between the two raters in the two indirect measure dyads (i.e., Y/N and LIK). An analysis of the percentages of total variance showed that the two indirect TI methods may affect the discrepancy between the two raters’ rating scores. A comparison of the three different TI methods and correlation between the child’s behavior and TI were examined using the PASW Statistics 18 software program. There was no significant difference between the WI and the Y/N dyads, while the assessments from the LIK dyad indicated a significant difference from the other two dyads. Both the WI and the Y/N dyads showed correlations between the degree of the child’s behavior and the degree of TI, but there was no significant difference between the two correlation coefficients. Questions about reliability with the indirect TI measures suggest one should be careful in considering these results.
dc.language.isoENen_US
dc.publisherThe University of Arizona.en_US
dc.rightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.en_US
dc.subjectBehavioral interventionen_US
dc.subjectComparisonen_US
dc.subjectTreatment Integrity methoden_US
dc.titleA Comparison of Treatment Integrity Assessment Methods for Behavioral Interventionen_US
dc.typetexten_US
dc.typeElectronic Dissertationen_US
dc.contributor.chairUmbreit, Johnen_US
dc.identifier.oclc659753788en_US
thesis.degree.grantorUniversity of Arizonaen_US
thesis.degree.leveldoctoralen_US
dc.contributor.committeememberLiaupsin, Carlen_US
dc.contributor.committeememberFerro, Joleneaen_US
dc.contributor.committeememberMacFarland, Stephanie Z. C.en_US
dc.identifier.proquest10872en_US
thesis.degree.disciplineSpecial Educationen_US
thesis.degree.disciplineGraduate Collegeen_US
thesis.degree.namePh.D.en_US
refterms.dateFOA2018-08-15T14:05:01Z
html.description.abstractThe purpose of this study was to examine the similarity of outcomes from three different treatment integrity (TI) methods, and to identify the method which best corresponded to the assessment of a child’s behavior. Six raters were recruited through individual contact via snowball sampling. A modified intervention component list and 19 video clips were derived from Stahr’s (2005) study, “An Intervention for Children with Autism Spectrum Disorders (ASD) who Have Food Selectivity.” The raters, randomly and evenly assigned to three dyads. Each dyad received an average of six hours training and reached 85% interobserver agreement (IOA) with a 0.60 kappa score. After training, each dyad watched 5 video clips per day and measured both the child’s behavior and TI. The percentages of IOA, kappas, and indices of dependability for assessment of the child’s behavior and TI were analyzed. The data revealed that all raters reached over 80% IOA and the whole interval (WI) and yes/no (Y/N) dyads reached .60 kappa, but the two raters in the Likert-type (LIK) dyad could not reach .60 kappa. The indices of dependability indicated that the six raters consistently observed and rated both the child’s behavior and TI, but there was a discrepancy in scores (i.e., percentages of TI) between the two raters in the two indirect measure dyads (i.e., Y/N and LIK). An analysis of the percentages of total variance showed that the two indirect TI methods may affect the discrepancy between the two raters’ rating scores. A comparison of the three different TI methods and correlation between the child’s behavior and TI were examined using the PASW Statistics 18 software program. There was no significant difference between the WI and the Y/N dyads, while the assessments from the LIK dyad indicated a significant difference from the other two dyads. Both the WI and the Y/N dyads showed correlations between the degree of the child’s behavior and the degree of TI, but there was no significant difference between the two correlation coefficients. Questions about reliability with the indirect TI measures suggest one should be careful in considering these results.


Files in this item

Thumbnail
Name:
azu_etd_10872_sip1_m.pdf
Size:
592.2Kb
Format:
PDF
Description:
azu_etd_10872_sip1_m.pdf

This item appears in the following Collection(s)

Show simple item record