Parsing Banner Downloaded Data into a Flat-File Format for Analysis

Setting The Stage

Assessments and measuring student performance in mastering American Dental Education Association competencies became a mandatory requirement of all dental colleges after 1997. The intent was to explain competency in terms of curricular development processes which were supposed to be dynamic and constantly evolving. This required the use of quantitative modeling with regression outputs of partial coefficients that reflected achievements in mastering competencies and thus performance on national boards.

The national board is a two part event, with Part I being taken at the end of the Sophomore year and Part II at the end of the first semester of the Senior year. Part I measures competence in the basic sciences core courses, such as biochemistry, anatomy, physiology and microbiology. Part II includes dental specialty areas in content, such as radiology, pathology, oral surgery and periodontology. Both are critical, high stakes standards that reflect how well curricular, teaching and measurement components were being used within a dental academic institution.

In this case, what LSCD needed was the means to determine the effect of didactic, clinical and laboratory instruction on achieving competencies and thus the relationship between mastery of competencies and national board outcomes. The process required that end-of-course behavioral outcomes listed on syllabi come from our stated competencies. Additionally, each question on the mid-term and final examinations had to be linked to a competency. Competencies were also linked to subject areas on the national board exams. Thus, individual competencies could be measured and courses that were discipline specific could be used as predictors of corresponding areas on the national board. All of this was possible if the relational data in Banner could be downloaded in an efficient manner and converted to a flat file that listed each student’s performance as a row in the table. Parsing data from a table involves the writing of scripts that explore the data in the table and extracts specific variables or fields of data according to the directions encoded in the scripts. This normally requires advanced programming skills.