Chauvin, Karen M. (1988, ASQC) U.S. Department of Education, Washington, DC
This paper will describe the procedures the U.S. Department of Education (ED) used to measure deviations from standards and calculate program-wide quality in five Federal student financial aid programs. These procedures were pilot tested in earlier studies on individual programs, and subsequently refined. This quality measurement was the first time all five inter-related programs had been measured simultaneously. Because a student can receive awards from all five programs, each with different rules, unique procedures were needed to capture the consequences of error on the interactions among the programs.
No national data base of student aid recipients exists. This, coupled with each program's (grants, loans, work-study) differing rules and requirements, presented challenges for sampling and error definition. This paper will explain the sampling plan, the several error definitions developed, and the error calculations conducted to measure the quality and accuracy of Federal student aid awards to individual students.
A methodology called "best" value selection for award recalculation will be described. The "best" value is the one that is documented and comes from the most reliable source. Values for both student and institutional data elements are ranked by the reliability of the supporting documentation. These "best" values are substituted for original values and individual awards are recalculated to determine if standards have been met.
The paper will describe the analysis conducted to determine probable causes of error and to test for associations between student error and student characteristics, and institutional error and institutional characteristics. Some simulations of proposed corrective actions will also be described.
Despite differing program regulations, intent, student, populations, institutional characteristics, and a host of other unique program features, the quality and accuracy of awards of Federal student aid can be measured and quantified. The probable causes of error can be analyzed and correction actions simulated.
Education,Financial aid,Quality control (QC)