Draft:Jackson G. Lu

Disaggregation of Asian Subgroups

In a 2024 study published in Current Directions in Psychological Science, Lu challenges the "model minority" stereotype suggesting Asians are uniformly successful. Drawing on an archival analysis of published research, he finds that most studies treat Asians as a monolith without specifying subgroups. He argues this methodology masks critical differences in socioeconomic status and outcomes among varying Asian populations. He advocates for researchers to avoid using the broad "Asian" category and instead categorize by subgroups.^[10]

Different Levels of Assertiveness

In a 2020 study published in the Proceedings of the National Academy of Sciences (PNAS), Lu and his co-authors demonstrate the importance of disaggregation by establishing that the Bamboo Ceiling primarily affects East Asians (e.g., ethnic Chinese, Koreans) rather than South Asians (e.g., ethnic Indians). Looking at archival analyses of chief executive officers and field surveys in companies, they found that East Asians were less likely than South Asians and whites to attain leadership positions, but South Asians were more likely than whites to attain leaderships. Examining motivation, prejudice, and assertiveness, they find that East Asians were equally motivated to become leaders and faced less prejudice than South Asians. However, East Asians consistently demonstrated lower levels of assertiveness when compared with South Asians, suggesting that the disparity is rooted in cultural fit. They argue that East Asians' lower levels of verbal assertiveness is incongruent with the communication styles traditionally expected of leaders in the United States.^[11]

Lu has also documented similar findings in educational settings. In 2022, his research revealed that East Asian students frequently underperform in U.S. law and business schools compared to their peers. He attributes the findings to lower assertiveness, revealing a Bamboo Ceiling in the classroom.^[12][13]

The Creativity Stereotype

Lu has also studied the way East Asians are often stereotyped as lacking creativity, which is an important leadership characteristic in the US. Studying MBA students, he found that East Asians were consistently perceived by their classmates as less creative than other ethnicities at the start of their program.^[14]

Academic Reception and Alternative Explanations

In a 2022 commentary published in the Proceedings of the National Academy of Sciences, Princeton sociologist Yu Xie praised Lu's work as "innovative, insightful, and provocative," noting its importance in demonstrating how culture impacts social stratification in the United States.^[15]

However, Xie also proposed several alternative methodological and cultural explanations for Lu's findings regarding East Asian students:

• Collider Bias: Xie suggested that Lu's 2022 analysis of law and business school classrooms may be subject to "collider bias". Because the study only analyzed students who had already passed highly selective admissions processes, Xie argued that strong quantitative test scores might negatively correlate with leadership potential within that specific sample, even if they do not correlate that way in the general population.

• Undergraduate Education: Xie pointed out that Lu's research did not control for where foreign-born students received their undergraduate education. He suggested that having completed an undergraduate degree outside of the U.S. might account for some of the verbal underperformance.

• The Dual Role of Culture: While agreeing that Confucianism deemphasizes assertiveness, Xie argued that the same cultural framework highly values "effort" and thorough preparation. He theorized that while low assertiveness is a liability for leadership attainment, this cultural emphasis on effort may actually provide an advantage for East Asians in settings evaluated by objective metrics.

Starting Salaries

In a 2023 study in the Journal of Applied Psychology, Lu explored the differences in starting compensation with regards to the Bamboo Ceiling. Analyzing large datasets of MBA graduates and U.S. employees, Lu finds that East and Southeast Asians start with significantly lower salaries than South Asians and Whites in non-consulting roles.^[16] He attributes this economic disparity to a lower propensity to negotiate starting salaries, which is driven by concerns of damaging the long-term relationship with a prospective employer.^[16]

Cultural Tendencies in AI

In a 2025 study published in Nature Human Behavior, Lu and his co-authors reported that AI models such as OpenAI’s GPT and Baidu’s ERNIE are not culturally neutral. They found that the models exhibited different social orientations depending on the language of the prompt. Specifically, models typically responded to an English prompt with an independent social orientation, whereas Chinese prompts generated responses characterized by an interdependent social orientation and a holistic cognitive style. Furthermore, the study indicated that culture-specific prompts can tune the model towards respective cultural tendencies. Lu’s research suggests that these models’ embedded cultural biases can influence individual users’ attitudes and behaviors without their awareness, leading the authors to advocate for the development of AI systems that account for cultural heterogeneity.^[20]

AI and Employee Creativity

Lu has also studied the impact of large language models on workplace creativity. In a field experiment conducted at a technology consulting firm and published in the Journal of Applied Psychology, Lu and co-authors found that AI assistance increases employee creativity by providing "cognitive job resources". They also hypothesized that LLM assistance is not universal, but rather effective depending on the individual user; employees with high levels of metacognitive strategies, e.g., an employee's ability to actively monitor and regulate their own thinking, are more successful at leveraging AI to enhance their creative output.^[21]

Empowerment-Entrapment Framework in Entrepreneurship

In a 2026 paper published in the Strategic Entrepreneurship Journal, Lu introduced the "Empowerment-Entrapment Framework" to formalize the way generative AI affects the entrepreneurial process. He posits that generative AI can both empower and entrap entrepreneurs. For example, while generative AI can increase creativity, it may also decrease population-level idea diversity because models often draw from the same training data.^[22]