Research Roundup: July 2023

Teaching Shift Benefits Disadvantaged Students

Last month’s Research Roundup featured a study showing that first-year undergraduates who completed an exercise related to belonging performed better during their college careers. A second group of researchers from the University of Texas at Austin (UTA) has tested whether another intervention might improve learning outcomes, especially for disadvantaged students. This approach, however, focused on encouraging teachers, not students, to embrace growth mindsets.

The researchers shared their results in a study published in PNAS. The co-authors included Cameron Hecht, a postdoctoral scholar in psychology at UTA’s Population Research Center (PRC); Christopher Bryan, an assistant professor in business, government, and society at the McCombs School of Business; and David Yeager, an assistant professor in psychology at the PRC.

The researchers asked 155 teachers, who together taught 5,393 students, to complete a 45-minute online module. The module introduced the growth mindset concept and described how teachers could adopt growth mindset practices in their classrooms. Another group of 164 teachers, who taught 6,167 students, completed a control module on an unrelated subject.

Before and after the intervention, teachers in both groups were asked about their views about growth mindsets and their plans to employ growth mindset teaching practices with their students. Researchers then assessed student performance six to seven months after the invention.

The team found that the intervention improved students’ pass rates by 11 percentage points and course grades by 0.36 grade points over those in the control group. The improvement was greatest in classrooms with a high percentage of students from socioeconomically disadvantaged backgrounds. The effect was much lower in classrooms where most students came from privileged backgrounds.

In classrooms where teachers employ growth mindset practices, students’ pass rates improve by 11 percentage points and course grades by 0.36 grade points.

Many interventions of this kind focus on encouraging students to change their behaviors, says Hecht. This study shows that encouraging teachers to adjust their practices “can help to change educational contexts for the better while relieving students of the burden of change.”

That said, Hecht admits that many attempts to ask teachers to change their classroom strategies fail because teachers already are juggling large workloads—they simply do not have the capacity to adjust their priorities. The success of this intervention, Hecht explains, might be because it took a values alignment approach, in which desired behavioral changes were associated with teachers’ core values.

This research shows that “values alignment can help to overcome this behavior-change challenge, even among teachers who are often among the most overworked and overwhelmed professionals.”

Training Employees to MYB (‘Manage Your Boss’)

Employees need a wide range of soft skills to succeed in the workplace, but three U.S. researchers have added a new one to the mix: They call it MYB, or “Manage Your Boss.” In a paper published in Personnel Psychology, the researchers suggest that employees who are adept at MYB have better working relationships with their supervisors, are better able to foster trust in the workplace, and receive higher performance ratings from their managers.

The paper’s co-authors include Ravi Gajendran, professor of global leadership and management at the Florida International University College of Business in Miami; Sal Mistry, assistant professor of management at the University of Delaware’s Lerner College of Business and Economics in Newark; and Subrahmaniam Tangirala, Dean’s Professor of Management at the University of Maryland’s Smith School of Business in College Park.

MYB, they explain, involves effectively anticipating managers’ goals, plans, priorities, and weaknesses. The three tested their theory involving MYB’s importance by conducting four surveys involving 1,313 respondents. They used this feedback to establish an MYB scale.

Then, using a matched sample of 200 employees and managers, the team asked the employees to rate the importance of workplace behaviors such as MYB, the standardization of job roles, and a failure of managers to structure tasks clearly (“laissez-faire leadership”). The managers, on the other hand, rated the employees’ job performance. The researchers found that employees’ higher ratings of MYB correlated with higher job performance ratings from their managers.

Employees who make an effort to understand their managers’ priorities and style can better align their actions with their managers’ expectations, thereby improving job performance.

The concept of MYB “recognizes that employees can feel empowered to take action to establish a good relationship and to make the relationship operate smoothly,” says Gajendran. That means that “employees can be proactive by making an effort to understand their managers’ priorities and style” so that they can better align their actions with their managers’ expectations.

The researchers emphasize that proactive followership behavior such as MYB is especially important for employees working in creative fields or in unstructured work environments, where their jobs are not standardized and they do not receive direction or feedback from managers. In these cases, companies that invest in coaching and training that helps employees learn MYB are likely to see better performance.

How Students View AI in the Classroom

How do students view the use of artificial intelligence tools such as ChatGPT in education? A*Help, a company that provides reviews of a range of academic services, recently surveyed 600 students to determine their perspectives on and use of AI for assignments. Of this group, approximately 82 percent were evenly split between high school and college students, with the rest coming from middle school, vocational, and postgraduate education. 

When asked which AI tools they regularly use for academic writing, students were most likely to use grammar checkers (nearly 63 percent), followed by citation generators (nearly 51 percent) and plagiarism checkers (34 percent). About 20 percent used AI text generation tools, while less than 15 percent regularly used text paraphrasing, essay writing, or academic assistance tools. 

While nearly 30 percent of students said they were not aware of weaknesses inherent to AI tools, many other students were still wary of using AI in their assignments: 

• 58 percent said they preferred doing assignments on their own.  
• 42 percent said that they did not use AI for their coursework because policies at their schools have limited or banned its use in the classroom. 
• Nearly 37 percent expressed distrust of AI’s accuracy. 
• Nearly 30 percent were concerned about the lack of citations for the information AI generates. 
• Just over 22 percent said they did not know how to use AI tools for their homework effectively. 

This last finding, the report notes, reveals opportunities “for educational institutions and AI tool providers to raise awareness and offer training in this area.”

Overall, students appreciated AI’s capacity to help them save time and generate ideas in tasks ranging from writing to coding, but they approached its use cautiously. AI is “a promising trend ... with room for growth and improvement,” the report concludes. The challenge for students and educators is to use the technology “ethically and effectively in education.” 

Some Humans Suffer from ‘Algorithm Aversion’

Researchers also are looking closely at how human workers are managing the increasingly pervasive presence of AI in the workplace. A new study shows that because many still mistrust AI’s accuracy, they are often overriding its outputs—and their decision-making suffers as a result.

The study’s co-authors Francis de Véricourt and Huseyin Gurkan, both professors of management science at ESMT Berlin, call this mistrust “algorithm aversion.” In their research, they explain that human decision-makers who systematically override AI algorithms might do so either because of algorithm aversion or because their working environments prevent them from checking a machine’s accuracy.

To test this phenomenon, de Véricourt and Gurkan created an analytical model in which human decision-makers were asked to oversee a machine tasked with making important decisions, such as whether doctors should perform a biopsy on a patient. Each time, the human decision-makers had to make the best choice based on the information the machine generated. Some who heeded the machine’s recommendations followed up on the decision, to see if its recommendations were proven correct. Others did not follow up, so they did not receive feedback on the machine’s accuracy.

Humans need more opportunities to collaborate with machines and gather experience so they can know when it’s better to trust AI and when it’s better to supercede its decisions with their own.

Levels of trust in AI’s accuracy increased among the human decision-makers in the first group, an indication that some might have developed a “verification bias” that would lead them to put too much trust in AI, the researchers explain. Among the decision-makers in the second group, levels of trust stayed the same, which means they might overrule AI even when its recommendations are accurate.

Many believe that humans who tend to override AI algorithms do so because of an intrinsic mistrust of AI, but that’s not always the case, says de Véricourt. “It may also be the case that we are simply not learning how to effectively use machines correctly, when our learning is based solely on the correctness of the machine’s predictions.”

This indicates that humans need more opportunities to collaborate with machines and gather experience so they can know when it’s better to trust AI and when it’s better to supercede its decisions with their own. There currently is a “lack of opportunities for human decision-makers to learn from a machine’s intelligence unless they account for its advice continually,” Gurkan says. “We need to adopt ways of complete learning with the machines constantly, not just selectively.”

Research News

■ New prize launched in sustainable research. The Haas School of Business at the University of California, Berkeley, announced the creation of its Berkeley Haas Sustainable Business Research Prize, which will be administered by the Berkeley Haas Center for Responsible Business. Each year, the prize, worth 20,000 USD, will recognize the research paper that holds the greatest potential to move businesses to immediately adopt sustainable practices that address urgent global environmental crises.

In recognition of the urgent need to address the climate change crisis, the school’s 2023 prize will look specifically at papers that study economic levers that motivate individuals, organizations, and markets to adopt initiatives that mitigate climate change and preserve resources. The initiative is supported by Haas School alumnus Allan Spivack, former president and CEO of RGI Home.

■ The EUA issues recommendations for Ukraine. The Ukraine Task Force of the European University Association has published a set of recommendations for ways that the higher education community can shift from offering emergency relief to Ukraine during Russia’s war against the country to providing long-term support. To ensure the continued strength and development of the Ukrainian higher education and research sector, the task force recommends that the global higher education community support the country’s universities in five areas: inter-institutional partnerships; virtual exchange and cooperation, particularly involving digital infrastructure; placements for Ukrainian academics and students; participation in reconstruction and rebuilding plans; and information sharing, cooperation, and coordination.

With these recommendations, the EUA also has launched an online questionnaire to gather information on university-based initiatives launched in response to the war.

“University collaboration is one of the critical lifelines connecting Ukraine with Europe and the world,” says Ivanka Popović, vice-president of the EUA and chair of the task force. “Beyond strengthening resilience, it is also a strategic investment into the country’s rebuilding.”

■ Digital transformation drives research partnership. In a specific collaboration between a European and Ukrainian institution, the University of Liverpool in the United Kingdom and Sumy State University (SumDU) in the Ukraine have entered into a five-month joint research project focused on the areas of digital economics, digital transformation and infrastructure, and renewable energy. The partners have received 197,000 GBP (approximately 250,500 USD) from Universities UK International as part of that organization’s “twinning” grant program that pairs institutions in the UK with those in the Ukraine.

Send press releases, links to studies, PDFs, or other relevant information regarding new and forthcoming research, grants, initiatives, and projects underway to AACSB Insights at [email protected].