AC 2008-1995: NONCOGNITIVE CHARACTERISTICS OF INCOMING ENGINEERING STUDENTS COMPARED TO INCOMING ENGINEERING TECHNOLOGY STUDENTS: A PRELIMINARY EXAMINATION

Studies have shown promise in predicting success for students in engineering based on noncognitive or affective characteristics. However, little if any literature exists on similar studies in the related discipline of engineering technology. Data has been collected from incoming engineering students in a large, Midwestern university using an instrument assessing students’ self-reported noncognitive characteristics over a four year period. The instrument has been shown to be stable and repeatable over this four year period. Cluster analysis has shown that students entering engineering cluster into three distinct groups. Five additional constructs have been added to the survey for the current cohort of students. This paper will examine results of an analysis of students in a pilot study in engineering technology using the same instrument already in use for incoming engineering students. Differences found between engineering and technology cohorts will be presented to the extent that the relatively small sample of technology students will allow. Finally, suggestions for future work in this area, the suggested applicability of individual items in the instrument and potential findings within the data will be presented. An effective instrument will allow for the development of a model to predict success, currently operationalized as retention into the second year. Such a model should prove to be valuable to address long standing issues of gender and ethnic disparities within engineering technology and should lead to improved advising and improved overall student success. Introduction: Engineering and engineering technology are comparatively lucrative fields of employment with continued strong demand. These fields attract academically strong students; however, graduation rates in engineering programs have been shown to be at or around 50% nationwide. Astin 1 showed that only 44% of students entering engineering showed engineering as their major four years later. The National Academies report “Rising Above the Gathering Storm” 2 describes technical programs as having some of the lowest retention rates among academic disciplines. Studies have shown that students who leave engineering tend to show differences in their noncognitive characteristics rather than cognitive ability or academic performance 3,4 . An instrument consisting of 161 items designed to assess noncognitive characteristics of incoming students was originally designed to assess characteristics of incoming engineering students. The instrument was used to assess nine noncognitive constructs, specifically focusing on those characteristics for which an institution may be able to develop intervention strategies / programs rather than an attempt to simply collect a wide range of data. Data has been collected for a four year period for incoming engineering students and were found to be psychometrically P ge 13933.2 stable and robust 5 . Additional characteristics which showed promise as being predictive of retention were found in a further review of the literature. Five additional constructs were added to the survey for students in the 2007 cohort for a total of fifteen constructs measured in 239 items. The overall intent is to find characteristics which may predict a student’s propensity to remain in an engineering program into the second year then incorporate this information in the development of a predictive model. It is theorized that a complete model to predict student matriculation into the second year should incorporate noncognitive characteristics in addition to cognitive variables as inputs; data that can be collected prior to the beginning of the first-year is expected to be especially useful as it would allow for more effective structuring of a student’s first-year experience. Similar instruments The Cooperative Institutional Research Program (CIRP) Freshman Survey covers a wide variety of attributes. These include attributes of the student’s background (for example, financial state of the family), self reported noncognitive attributes from the student (attitudes towards school) and cognitive information (high school performance). This large database holds information for over 190,000 students at over 300 institutions. High school GPA, SAT math scores and SAT verbal scores were shown to be cognitive predictors of future performance (defined as collegiate GPA). Self-ratings of ability in mathematics, computers and overall academic ability were shown to be noncognitive predictors 6,7 . The Pittsburgh Freshman Engineering Attitudes Survey (PFEAS) measures 13 characteristics using self-reported answers to 50 items. Students who left engineering were found to have a significant difference in “general impression of engineering”, “perception of the work engineers do for the engineering profession” and “engineering comparing positively to other fields of study”. Students who left also had a lower self-reported confidence in their engineering skills 8 . Zheng 9 reported on the relation between retention and six cognitive and noncognitive variables in a study using the Southeastern University and College Coalition for Engineering Education (SUCCEED) longitudinal database. Correlation between high school GPA and SAT math scores and retention was shown, although results varied by campus. Reasons cited for students leaving engineering included an inability to handle stress, a mismatch between personal expectations and college reality and lack of personal commitment to a college education. The Persistence in Engineering (PIE) instrument was developed as part of the overall Academic Pathway Study (APS) 10 . This survey covers a wide variety of characteristics from environmental (financial), post enrollment (attitudes about the first year experience) and self reported noncognitive characteristics (motivation). Noncognitive predictors of retention into the fourth year of engineering included motivation due to family member influences, confidence in math and science and level of engagement in the classroom 11 . These instruments collect data during the first year of study and/or data which, while potentially predictive, isn’t always likely to be affected by first year programs or intervention strategies. P ge 13933.3 Seymour and Hewitt report that students who left and students who remain in engineering were very similar in their academic abilities 12 . Students who left primarily cited reasons dealing with the culture of the institution and aspects of engineering as a career rather than academic factors. Taken as a whole, the literature suggests that differences in noncognitive characteristics may play a more important role in retention in engineering than differences in cognitive characteristics. This would suggest that interventions assisting in noncognitive needs of students prior to and during the first year of study would benefit more students than strictly academic assistance. As stated by Pascarella, 13 “A significant amount of student attrition may be prevented through timely and carefully planned institutional interventions. Such interventions will be most effective if those students with a high probability of dropping out can be accurately identified.” Constructs in the instrument: The initial instrument consisted of nine constructs divided into subconstructs as specified in their original design or discovered though factor analysis. Motivation: Motivation was evaluated using the Academic Intrinsic Motivation Scale (AIMS) 14 , a scale consisting of 25 items with four subfactors: Control, Challenge, Curiosity and Career. Metacognition: The Metacognition scale consists of planning, self-monitoring, cognitive strategy and awareness subfactors. This describes a student’s perception of their strategies for monitoring and modifying their cognition 15,16 . Deep learning and Surface learning: Items for these scales were adapted from the Study Process Questionnaire (SPQ) 17 . Deep Learning consists of subfactors motive (or intrinsic interest) and strategy (or maximizing meaning). The surface learning construct consists of memorization and studying. Self-efficacy: Many studies indicate the importance of self-efficacy and it has been shown to be predictive of retention 18,19 . No subfactors were found in this ten-item scale. Expectancy-Value: The expectation of the value placed in each of four subfactors, including employment opportunities, persistence, academic resources and community involvement is assessed in these 32 items. Major Indecision: The scale items in Major Indecision were developed based on models of career indecision as described by Osipow 20 , who discusses a number of different career indecision inventories. The Major Indecision scale consists of 21 items with the following subfactors: urgency, personal issues (related to the student’s choice in major), certainty of decision and difficulty in decision (self-assessment of the student’s tendency to be indecisive in general). One item was unaccounted for in these subfactors and is treated separately (independence). Page 13933.4 Leadership: The student’s self assessment of their leadership is based on four subfactors including self-assessment (primarily self assessment of their organizational and leadership skills), motivation, planning and (interaction with) teammates. Characteristics of leadership are theorized to have a positive effect on student retention 21 . Team vs. Individual Orientation: A student’s ability to contribute to a team environment is increasingly valuable in industry. This construct is comprised of two subfactors: individual and team dynamic and consists of 10 items. A bifactor model structure was found, allowing each item to load to a construct and exactly one subfactor within that construct. This model structure implies that each subfactor is orthogonal or independent of other subfactors. Internal consistency, a common measure of reliability, was assessed within each construct and