Beyond the traditional notions of intelligence and academic achievement, the successful development of noncognitive or "soft" skills (e.g., personality factors, motivation, creativity) among school-aged youth represents an important objective for any educational system. Such skills have been shown to be significantly predictive of successful outcomes across both academic and labor market settings, and are prized in applicants by institutions of higher education and employers alike. These phenomena have rendered their positive expression a valuable asset among students preparing to enter college or the workforce. Concomitant with the identification of a need to encourage the development of students' noncognitive skills is the need for multifaceted assessment tools demonstrated to accurately measure such skills. Such an assessment, if shown to exhibit psychometric integrity (e.g., reliability, convergent/divergent validity, measurement invariance) across multiple noncognitive dimensions, would provide a valuable device for future research as well as schools interested in optimally preparing their students for both future academic pursuits and the workforce. The current study makes substantial progress toward the development of such an instrument using data from a large sample of middle school students (Grades 6-8) attending eighteen schools across the U.S. Students completed the assessment online, and were followed longitudinally at two time points over a six-month period. Multiple informants and item types were used to gather data from students on six noncognitive constructs of interest (Intrinsic Motivation, Resilience, Teamwork, Time Management, Creativity, and Ethics). This report presents the results of structural analyses using techniques appropriate for item response data represented as ordered categories. Exploratory and confirmatory factor analyses were conducted using independent subsamples, with further tests to evaluate the degree to which each revealed construct exhibited longitudinal measurement invariance. Longitudinal multitrait-multimethod (MTMM) models were planned to investigate construct validity, but their estimation encountered problems. Therefore, MTMM models were examined separately at each assessment wave. Results revealed that the assessment offers a wide array of constructs demonstrating reliability and longitudinal measurement invariance, with further evidence for discriminant and modest convergent validity. Implications for future research and practical applications are discussed. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]