In recent years, pre-college age students have expressed increasingly less interest in careers related to science, technology, engineering and math (STEM) fields. In 2013 a survey of 1,025 teens conducted by Junior Achievement USA and the ING U.S. Foundation, revealed a 15 percent decline in interest in STEM and medical fields from the previous year. Meanwhile, according to the National Math and Science Initiative, only 44 percent of the 2013 U.S. high school graduates were deemed ready for college-level math. For college-level science, the number dropped to 36 percent.
A team of researchers from SIUE and the University of Southern California (USC) are honing in on middle-school age children and their self-confidence in science, knowing that “self-efficacy,” or one’s belief in his or her ability to achieve goals, can be an important factor in career-related choices. After school programs, such as robot-building competitions, have been shown to help boost kids’ self-efficacy in science-related fields like computer science, mechanical engineering, and math. Yet, one key element in many after-school programs related to science has not yet been well studied—the role mentors can play in supporting the development of a child’s self-efficacy.
Mentorship plays a frequent and vital role in today’s science curriculum. From peer tutors, to teachers, to professional role models, mentorship has long been recognized as a key factor in helping to raise students’ self-efficacy in science. However, very little is known about which mentorship practices are most effective in boosting students’ attitudes. The SIUE-led team of researchers proposed to contribute a deeper understanding of good mentorship practices in STEM education and how students respond to that mentorship.
The study, funded by the National Science Foundation’s ITEST (Innovative Technology Experiences for Students and Teachers) grant program, is headed by Dr. Gary Mayer, assistant professor of computer science in the SIUE School of Engineering, but it represents the collaborative efforts of a multidisciplinary team from two institutions. Expertise was recruited from Mayer, Dr. Jerry Weinberg, associate provost for research and professor of computer science at SIUE; Dr. Stephen Marlette, associate professor of curriculum and instruction in the SIUE School of Education, Health and Human Behavior; Dr. Sharon Locke, associate professor and director of the SIUE Center for STEM Research, Education and Outreach; Dr. Susan Thomas, former psychology professor at SIUE; researchers from the Illinois Education Research Council; as well as Dr. Maja Mataric and Ross Mead, a doctoral candidate, at USC. The team represented not only a cross-section of researchers who had the right combination of robotics expertise and interest in science education, but they also hail from institutions surrounded by local schools with diverse and underserved student populations.
The ITEST idea emerged from past research by Weinberg and Thomas that explored whether or not robotics might help raise girls’ confidence in STEM fields. Their study found that robotics competitions could raise girls’ self-efficacy, and results suggested that mentoring could play a role in enhancing the students’ self-perceptions in science. However, the researchers knew little about the types of mentorship most effective in raising STEM self-efficacy. Did it matter if the mentoring employed self-efficacy enhancing techniques, or was it enough to use general best practices? Might there be other factors, such as sex, race or ethnicity, that moderate the impact of mentorship? Understanding each aspect of mentoring and how it contributes to self-efficacy, the research team argued, is vital in making the best use of it in STEM-related activities.
With robotics as a core element, the study culminated in spring 2013, when Mayer oversaw 435 Latino, African-American and Caucasian seventh- and eighth-grade students from underserved schools in the St. Louis and Los Angeles areas as they competed in Botball®, a robotics competition organized by the KISS Institute for Practical Robotics (KIPR). Combining the excitement of competition with an opportunity for personal expression through robot design, the Botball competition annually holds 13 regional U.S. tournaments as well as international competitions. Over the course of a season, students involved in Botball work together to design, build and program a robot that will meet a particular challenge. Teachers and mentors work with students to prepare for the competition and lead them into the final tournament. Unfortunately, according to Locke, although teachers play a critical role in the process as team mentors, there is currently only limited mentor training available for Botball addressing how to successfully guide a team and be a good mentor.
Thus arose the opportunity to perceive how various types of training could affect mentorship practice and, in turn, examine how mentorship strategies influence students’ self-efficacy. A key focal point of the project was 45 K-12 teacher mentors who each worked with groups of eight to ten students. The ITEST research team split the teachers into four groups: those who received no mentor training; those who received training in general best practices in mentorship; those who received training in mentor strategies specific to self-efficacy; and those who received a combination of general best practices and techniques specific to self-efficacy. The team also analyzed student Botball participants’ attitudes about pursuing science courses and careers and their expectations for success in STEM areas, using questionnaires to measure the students’ self-efficacy throughout the process.
By examining the impact of various mentorship groups, the team not only sought to track the rising confidence of students as they trained for and competed in Botball, but they also attempted to delineate the effects of various mentorship practices. While overall mentor group differences were not found, the research group did determine that some teachers had positive mentor practices to which the students responded. For instance, students of those mentors who engaged in positive mentor practices (such as: encouraging students during difficult tasks, giving students helpful feedback, helping students believe they could accomplish the tasks, and helping students to appreciate all team members’ contributions) made more gains by the end of the study in student self-efficacy, STEM achievement-related choices, and STEM expectations for success than those students whose mentors did not engage in positive practices. Furthermore, the team determined that the relationship of prior STEM self-efficacy and post STEM self-efficacy was stronger for Caucasian students than for minority students (i.e, African American and Latino/Latina students) which indicates that there may be greater opportunity to affect STEM self-efficacy in minority students compared to Caucasian students.
Based on the results of their explorations, the team is developing a mentor-training package that could be used in all robotics programs and be generalized to other STEM activities.
According to Locke, the study could be applied across STEM disciplines. “The materials will fill a gap in resources on mentoring, and because they are research-based and not specific to robotics, they will be broadly useful for other mentoring situations such as academic Olympiads or math competitions,” she said. Through developing models of mentoring for STEM activities and constructing supporting materials, the ITEST team hopes to provide a variety of techniques that allow STEM educators to make the best use of mentors—a powerful tool in providing students with the confidence they need to advance to STEM careers.