Need assessment for TA training: A survey to capture particular needs at an institution

A widely-acknowledged factor inhibiting education engineering reform, interfering with learning, and reducing student retention is the inexperience and lack of pedagogical education of many instructors. This is particularly acute among graduate teaching assistants (GTAs), who often contribute the majority of contact hours within engineering courses and may gain little pedagogical content knowledge before they assume the role of primary instructor in graduate school or their first academic appointment. Often their preparation is little more than a syllabus and the textbook. An important first step in addressing this issue for engineering educators, whose responsibility is to supervise and mentor GTAs, is to identify the factors to be considered and assessed in developing an effective professional development program for GTAs. We developed a need assessment survey to measure GTAs' perceived importance of various roles and responsibilities. Our results provide strong evidence that it is a valid and reliable measure. The participants rated the importance they placed on each of 24 GTA roles and responsibilities on a 5 point Likert scale. Analysis of the data included an exploratory factor analysis and reliability tests to ascertain the construct validity and reliability of the survey. Factor analysis was conducted with oblimin rotation with the 24 items of GTA roles. The eigenvalue-greater- than-one rule was initially used in combination with a scree test to determine the number of factors that would appropriately represent the concept of GTA roles and responsibilities. The results indicated a four-factor structure, accounting for approximately 54.03% of the total variance. Based on the four-factor structure that an exploratory factor analysis indicated, we conceptualized four categories of GTA roles and responsibilities; 1) clear communication, 2) student management, 3) preparation for feedback and assessment, and 4) course management /policy knowledge. To examine concurrent construct validity, we examined how our new survey is related to teacher motivation measure, one of the empirically validated measures. Reliability coefficients with the sample of this study for the overall measure and for the four subscales of the survey ranged between 0.77 and 0.82. The correlation analysis provided concurrent validity of the measure because the four factors were positively related to GTAs' intrinsic motivation to teach. Although the survey showed favorable psychometric properties, further testing is warranted to confirm that the new measure of GTA's need assessment can be used as a reliable and valid tool across institutions.