Identifier

etd-04092013-200636

Degree

Master of Arts (MA)

Department

Communication Sciences and Disorders

Document Type

Thesis

Abstract

In this study, we analyzed the validity and reliability of the Infant-Toddler Meaningful Auditory Integration Scales (IT-MAIS; Zimmerman-Phillips, Osberger, & Robbins, 2001), an assessment designed to measure listening skills in children ages 0-3 years. The IT-MAIS is a caregiver report tool used by speech-language pathologists and audiologists to assess listening skills in children with sensorineural hearing loss (SNHL) pre- and post-cochlear implant (CI). The IT-MAIS is widely used; however, it has not undergone thorough psychometric analysis. Using longitudinal data collected by the University of Iowa Children’s Cochlear Implant Program, we analyzed the psychometric properties of the IT-MAIS via Rasch analysis, a 1-parameter (1-PL) model of Item Response Theory (IRT; Lord & Novick, 1968). Pre- and post- CI assessments from 23 CI users aged 10 to 36 months were evaluated. IRT is a form of psychometric analysis that is emerging in the behavioral sciences as a viable alternative to Classical Test Theory for test development and analysis. IRT results are similar, but not identical to classically derived concepts of validity and reliability. Specifically, we analyzed the content and construct validity of the IT-MAIS. We found that 2 out of 10 items exceeded misfit criteria, meaning participants did not respond predictably to these 2 items. We also found that the item-difficulty range did not capture the full range of participant ability, especially the higher range of participant ability. Therefore, the IT-MAIS may not be assessing higher-level listening skills, particularly in children post-CI. Rasch analysis also revealed that 1 of the 5 rating scale categories was not used predictably, indicating that the rating scale was not used as the test developers (Zimmerman-Phillips, Osberger, & Robbins) intended. To analyze item order relative to sequential development of listening skills, we established an a priori item rank order and compared it to item difficulty order established by Rasch analysis. Overall, our results indicated the IT-MAIS did not demonstrate ideal item-level psychometric properties according to Rasch analysis and item order did not reflect sequential development of listening skills. We concluded that the IT-MAIS should not be used to assess listening development from pre- to post-CI.

Date

2013

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Barker, Brittan

Share

COinS