Session T3A THE PERFORMANCE OF ENGINEERING STUDENTS ON THE GROUP EMBEDDED FIGURES TEST
Sheri Clark, Elaine Seat, and Fred Weber1
Abstract - The Group Embedded Figures Test (GEFT) measures the ability of a person to disembed or pull out specified objects from a given background. The ability to disembed has been shown to be a necessary skill in problem solving and is consequently, thought of as a necessary trait for individuals interested in engineering. Therefore, the GEFT has the potential to be used as a predictor of student success in an engineering program. In order to test this hypothesis, the GEFT was given to undergraduate engineering students (with both high and low academic success) and to a control group of liberal arts students at The University of Tennessee in Knoxville. This paper presents the results of the comparisons and the implications for improving engineering student performance. required for problem setup for beginning engineering statics and dynamics problems. Since 1994, the College of Engineering has collected data to determine a potential indicator of success in engineering school. This predictor has been termed the Success Performance Index, or SPI. The SPI is calculated by multiplying the high school GPA (on a four-point scale) by ten and adding the mACT score. Six years of research in the College of Engineering suggests that students with a SPI under 50 have a less than 1% chance of survival in the engineering program, and students with a SPI between 49 and 57 have a less than 15% chance of survival. It is hypothesized that the SPI is a valid indicator of student success because the mACT indicates the student’s math ability, while the High School GPA suggests the student’s discipline and ability to persevere. For instance, a student with average math ability (ACT of 24), but a good work ethic (GPA of 3.5) would have a SPI of 59. However, a student with a GPA of 3.8, perhaps from a high school with a marginal mathematics program, might perform poorly on the math portion of the ACT with a score of 20, indicating a deficiency in the necessary math skills, although this individual is most likely a good student. This paper addresses relationships that may aid in understanding the difficulty that some students have with engineering problem solving. First, the relationship between satisfactory performance in the freshman engineering program, as represented by the math and science component of the grade, and the ability to disembed, as determined by the Group Embedded Figures Test (GEFT), is explored. In addition the relationship between an entering student’s SPI and GEFT is examined. Finally, a comparison is made between the GEFT performances of engineering students and a control group of liberal arts students in order to investigate the use of the GEFT as a predictor of success in engineering.
Good students exhibit many qualities such as a strong work ethic, time management and study skills, self-discipline, and an intrinsic motivation for high performance. While all of these qualities are desirable, they are not always sufficient for success in engineering. In addition to these traits and skills, students who succeed in engineering programs must also have a strong background in math and science, as well as superior problem solving and analytical abilities. The College of Engineering at the University of Tennessee is very interested in understanding the success of their students. As a result, they have developed a new integrated freshmen program that has improved the second year retention rate from 45% to 60% . This program has improved the performance of freshman engineering students by presenting the first-year curriculum in a variety of formats to address different learning styles . However, evaluation of student performance in this program suggests that some students do not possess or exhibit adequate problem solving skills. Evaluation of students in the freshman program is based on their math and science performance, team project scores, and computer tools skills. It has been shown that there is a population of students who perform well in the computer tools and team project areas of the course but do poorly in the math and science component. While these students often have satisfactory overall high school GPAs, they often lack a strong math background, as evidenced by low math ACT (mACT) scores. These students also experience difficulty with the basic problem solving skills
Cognitive style is the particular approach a person takes to address a problem within a given situation . Witkin, et al. defined cognitive style as field-dependent/field-independent. Field-dependent individuals approach a given situation by paying attention to and primarily focusing on the surrounding field. They tend to observe the individual components and the surrounding structure as though they are fused together. In other words, field dependent individuals
The University of Tennessee, Knoxville, Tennessee 37996
0-7803-6424-4/00/$10.00 ? 2000 IEEE October 18 - 21, 2000 Kansas City, MO 30 th ASEE/IEEE Frontiers in Education Conference T3A-1
become distracted by the background and are not able to identify key features within it. Conversely, fieldindependent individuals tend to view components separately from the “organized ground” [3, p. 4]. They are able to overcome the organization of the surroundings in order to focus on the individual components. This ability allows field-independent people to be superior problem-solvers . The classifications of field-dependence/ field-independence are not discrete. Rather, they are extremes of a continuum. The ability to disembed, or “pull out” specified objects from a given background, has been shown to be a necessary skill in problem solving. Consequently, field independence is thought to be a necessary trait for individuals interested in engineering. Barrett and Thornton  wrote, “One of the major findings, in over 50 studies, was that fieldindependent people tended to be analytical, logical and well articulated in being able to extract subtle aspects from problems for analysis” (p. 789). The Group Embedded Figures Test (GEFT) is a frequently utilized instrument to measure an individual’s degree of field-dependency. Participants who take the GEFT are asked to identify a series of simple figures within more complex forms as shown below in Figure 1. WISC. Individuals showing higher field-independence traits scored higher than field-dependent participants did on the analytic components of the intelligence tests. It is important to note that they did not score higher on the two factors that did not require disembedding (i.e., Verbal-Comprehension and Attention-Concentration). Therefore, no correlation exists between field-independence/field-dependence and overall intelligence.
The Group Embedded Figures Test (GEFT) measures perceptual and problem-solving ability by requiring participants to identify simple geometric figures within more complicated drawings. The ability to disembed has been shown to be a necessary skill in problem solving. Consequently, field-independence is thought of as a necessary trait for individuals interested in engineering. Therefore, there is a possibility that the GEFT has the potential to predict success of a student in an engineering program. We hypothesized that: 1) higher performing (higher grade point averages in freshmen math and science courses) engineering students will score significantly higher on the GEFT than the lower performing students, 2) that the students with higher SPI scores will have significantly higher GEFT scores than the students with lower SPI scores, and 3) engineering students will score significantly higher on the GEFT than the liberal arts students.
Find the “X” in this figure.
Participants Figure 1 Example of Simple and Complex GEFT Figures The instrument is scored from 0 to 18 with higher scores indicating higher degrees of field-independence. Benbasat and Dexter  stated that “Mean scores falling in the range 13-14 have often been used to distinguish between the two dimensions (of field-dependency); persons achieving scores below the mean are thus classified as field dependent.” The Manual for Embedded Figures Test, Children’s Embedded Figures Test and Group Embedded Figures Test , addressed the intellectual component of fielddependency. Two studies performed by Goodenough and Karp  and Karp  expounded on this connection.. The studies related performance on the individual form of the Embedded Figures Test (EFT) to the analytic factors of both the Wechsler Adult Intelligence (WAIS) and the Wechsler Intelligence Scale for Children (WISC). The three subtests associated with the analytic factors of the intelligence tests are Block Design, Object Assembly and Picture Completion. These subtests require separating a particular item from the overall organized context. Both studies found a correlation between the EFT and the analytic portion of the WAIS and The participants were all students at the University of Tennessee in Knoxville. All students volunteered and received extra credit in exchange for their participation in the experiment. Since the norms for the GEFT were obtained using liberal arts students, they were judged to be the best control for the present study. There were 53 liberal arts students and 157 second semester freshmen engineering students involved in the experiment. Of the engineering students 25% were female and 75% were male. Eightyseven percent were Caucasian, 8% African American, 2.5% Asian and 1.5% Hispanic. Instrument The GEFT was utilized to assess the students’ level of fielddependency. The instrument is presented in booklet form. The participants are asked to examine each of the complex figures and trace the simple form indicated. The simple forms are printed on the back cover of the booklet for quick and easy reference. There are three sections to the instrument. Section 1 is a practice section. It contains seven relatively easy forms and the participants have two minutes to answer them. Both Sections 2 and 3 contain nine items
0-7803-6424-4/00/$10.00 ? 2000 IEEE October 18 - 21, 2000 Kansas City, MO 30 th ASEE/IEEE Frontiers in Education Conference T3A-2
and participants are given five minutes to complete each of them. Only Sections 2 and 3 are scored. An item is considered “correct” if the simple item is correctly outlined within the more complex figure. A score of 18 is considered a perfect score and a score of zero indicates that the student did not correctly identify any of the embedded figures. Higher scores indicate higher degrees of field-independence. Benbasat and Dexter  classified individuals scoring below 13 as field-dependent and those scoring above 13 as fieldindependent. Procedure high GEFT score is the ability to identify or “pull out” pertinent information. Liberal arts students correctly identified an average of 11.7 simple items from the more complex figures. The scores had a standard deviation (SD). The engineering students correctly identified an average of 14.6 items and the SD of scores was 3.73. A Mann-Whitney nonparametric test comparing the liberal arts students’ GEFT scores to those obtained by the engineering students indicated that the engineering students significantly identified more items than did the liberal arts students (p <.001).
Instructors in both colleges informed their students about the study. They explained that the experiment would take about thirty minutes to administer and that they would give extra credit on an assignment for participation in the experiment. Students were tested in groups. First, they read and signed an informed consent form, supplied demographic information and asked any questions they may have had. Each GEFT booklets contained a non-identifying participation number to ensure anonymity and confidentiality. The students provided information about their college, major and date of birth in the space provided. The engineering students’ informed consent form also contained information about the process of obtaining their engineering GPAs. They were assured of the anonymity and the confidentiality of their responses. The students received both verbal and written instructions for the GEFT. A short practice session preceded the two experimental portions of the test. In each experimental section, the students were given five minutes to locate and identify the simple objects embedded within the complex object. Following the completion of the booklet, students were debriefed and dismissed. One of the major purposes of the present study was to determine if the Group Embedded Figures Test could be used as a predictor for student success in an engineering program. This study was important in order to begin to examine the complex relationship of student ability and student motivation. The GEFT identifies individuals who have the ability to “pull out” important information. However, It seems as though the situation has not yet been fully addressed. Both ability and motivation seem to be integral parts of student success. A student who has the ability to do complex problem-solving but little or no motivation to do the work will most likely be no more successful than a student who has significant motivation with less problem-solving skills. The significant positive correlation between the GEFT and student grades supported the theory that students with superior problem-solving skills earn higher grades in the math and science portion of the engineering program. However, apparently there is some uncontrolled factor that is interfering with this relationship (yielding a moderate instead of a strong correlation). IN a review of the data points for the engineering students, there were a few individuals who scored high on the GEFT and received very poor grades. Perhaps this apparent discrepancy is due to low motivation to succeed in the program. There is a possibility that these students have all of the skills necessary to complete the work in the program, they merely lack the inner drive or discipline to persevere in times of stress. There is also the possibility that the low-scoring students no longer want to become engineers. In order to assess these hypotheses, researchers could give the students an occupational or interest assessment (i.e., the Strong Interest Inventory). Additionally, perhaps the GEFT could be used in high schools to give the students an idea if they have the disembedding skills necessary to succeed in an engineering program.
All hypotheses were supported by the data. Freshmen engineering students who scored higher in the math and science portion of the freshmen program also scored significantly higher on the Group Embedded Figures Test (GEFT) as measured by a Spearman’s Rho correlation (p< .001). This was a positive moderately strong correlation (r=.315). Another Spearman’s Rho correlation indicated a statistically significant relationship between GEFT scores and SPI scores (p=.003). This means that students who obtain high SPI scores are also more field-independent. This relationship was predicted. The investigators theorized that students who obtained high GPAs in high school and high scores on the math section of the ACT (resulting in a high SPI score) would most likely have the disembedding skills necessary to achieve a high score on the GEFT. This demonstrates that a common element for a high SPI and a
0-7803-6424-4/00/$10.00 ? 2000 IEEE October 18 - 21, 2000 Kansas City, MO 30 th ASEE/IEEE Frontiers in Education Conference T3A-3
. Gilliam, Fred T., Klukken,P. Gary, Pionke, Christopher, D.,
Scott, Tom H., Seat, J. Elaine, Symonds, Fred, Weber, Fred E., & Yoder, Daniel C. “The Engage Program: Evaluating the First Year Experience at the University of Tennessee,” Proceedings, 2000 Frontiers in Education Conference, ASEE/IEEE, 2000. . Gilliam, Fred T., Klukken,P. Gary, Pionke, Christopher, D., Scott, Tom H., Seat, J. Elaine, Symonds, Fred, Weber, Fred E., & Yoder, Daniel C. “The Engage Program: Renovating the First Year Experience at the University of Tennessee,” Proceedings, 1998 Frontiers in Education Conference, ASEE/IEEE, 1998. . Witkin, H., Oltman, P., Raskin, E., & Karp, S. Manual: Embedded Figures Test; Children’s Embedded Figures Test; and Group Embedded Figures Test. Palo Alto, CA: Consulting Psychologists Press, Inc., 1971. . Ducker, in Witkin, et al., 1971.
. Barrett, G., & Thorton, C. “Cognitive Style Differences
Between Engineers and College Students,” Perceptual and Motor Skills, 25, 1967, pp. 789-793. . Benbasat & Dexter, 1982, pp. 952, cited in Mykytyn, P. “Group Embedded Figures Test (GEFT): Individual Differences, Performance and Learning Effect,” Educational and Psychological Measurement, 49, 1989, pp. 951-959. . Goodenough, D., & Karp, S. “Field dependence and intellectual functioning,” Journal of Abnormal and Social Psychology, 63, 1961, pp. 241-246. . Karp, S. “Field Dependence and Overcoming Embeddedness,” Journal of Consulting Psychology, 27, 1963, pp. 294-302.
The implementation of the Engage program is supported by the National Science Foundation through grant number EEC-9972944.
0-7803-6424-4/00/$10.00 ? 2000 IEEE October 18 - 21, 2000 Kansas City, MO 30 th ASEE/IEEE Frontiers in Education Conference T3A-4