The Transition Readiness Assessment Questionnaire (TRAQ): Its Factor Structure, Reliability, and Validity

• Published in: Academic pediatrics 2014-07, Vol.14 (4), p.415-422
• Main Authors: Wood, David L., MD, MPH, Sawicki, Gregory S., MD, MPH, Miller, M. David, PhD, Smotherman, Carmen, MS, Lukens-Bull, Katryne, MPH, Livingood, William C., PhD, Ferris, Maria, MD, MPH, PhD, Kraemer, Dale F., PhD
• Format: Article
• Language: English
• Subjects: Activities of Daily Living; Adolescent; Adult; Boston; Cystic Fibrosis - psychology; Cystic Fibrosis - therapy; Disabled Persons - psychology; Disabled Persons - statistics & numerical data; Factor Analysis, Statistical; Female; health care transition readiness; Health Services Needs and Demand; Humans; Male; Massachusetts; Neonatal and Perinatal Medicine; North Carolina; Pediatrics; Reproducibility of Results; self-assessment questionnaire; Surveys and Questionnaires - standards; Transitional Care - statistics & numerical data; Young Adult; youth with special health care needs
• ISSN: 1876-2859; 1876-2867
• DOI: 10.1016/j.acap.2014.03.008

Abstract Objective National consensus statements recommend that providers regularly assess the transition readiness skills of adolescent and young adults (AYA). In 2010 we developed a 29-item version of Transition Readiness Assessment Questionnaire (TRAQ). We reevaluated item performance and factor structure, and reassessed the TRAQ's reliability and validity. Methods We surveyed youth from 3 academic clinics in Jacksonville, Florida; Chapel Hill, North Carolina; and Boston, Massachusetts. Participants were AYA with special health care needs aged 14 to 21 years. From a convenience sample of 306 patients, we conducted item reduction strategies and exploratory factor analysis (EFA). On a second convenience sample of 221 patients, we conducted confirmatory factor analysis (CFA). Internal reliability was assessed by Cronbach's alpha and criterion validity. Analyses were conducted by the Wilcoxon rank sum test and mixed linear models. Results The item reduction and EFA resulted in a 20-item scale with 5 identified subscales. The CFA conducted on a second sample provided a good fit to the data. The overall scale has high reliability overall (Cronbach's alpha = .94) and good reliability for 4 of the 5 subscales (Cronbach's alpha ranging from .90 to .77 in the pooled sample). Each of the 5 subscale scores were significantly higher for adolescents aged 18 years and older versus those younger than 18 ( P