PhD student Nikhita Joshi (MMath ’20) has been selected to attend the prestigious Rising Star in Electrical Engineering and Computer Science (EECS) workshop.
In 2012, the Massachusetts Institute of Technology (MIT)’s EECS department launched the workshop to encourage early-career women and gender minorities in these fields to explore life in academia. Since its inception, the Rising Stars in EECS workshop has been hosted by various universities across America. This year, it will be held at MIT from October 24 to 25.
“I’m honoured to have been named a Rising Star in EECS by MIT! I’m thrilled to see top schools like MIT invest in the career development of young women and other historically marginalized groups. I know many past attendees have since had prolific and impactful careers in EECS,” says Nikhita. “I'm interested in becoming a computer science professor, so the workshops, panels, and poster sessions will prove to be invaluable. I also can’t wait to meet all of the other Rising Stars! Special thanks to my lab mate, Ryan Yen, for telling me about the program and to my advisor, Daniel Vogel, for the support.”
Nikhita Joshi has been invited to the prestigious Rising Star in Electrical Engineering and Computer Science (EECS) workshop. This career-building workshop encourages early-career researchers of historically marginalized or underrepresented genders to explore life in academia.
The Rising Stars in EECS workshop is open to graduate students and postdoctoral researchers of historically marginalized or underrepresented genders such as cisgender women and non-binary people. However, it is extremely competitive. Only 70 researchers were chosen from an application pool of 370.
“Nikhita has an incredible research profile, where she is pushing the boundaries of human-computer interaction (HCI) research. She blends various theories in psychology to create innovative tools and devices,” says her supervisor, Professor Daniel Vogel. “On top of that, she is an outstanding communicator, leader, and teacher. She’s a strong role model for other young women considering computer science as a career.”
The workshop invites distinguished academics including Nobel laureates and Association for Computing Machinery (ACM) Fellows, to help guide and mentor the next generation. For example, participants can attend panel discussions and presentations on how to prepare for the academic search process and succeed as a junior faculty member. There are also several social events to encourage collaboration and networking.
Another event is the poster session, where attendees can showcase their latest research projects. Nikhita will present her current work on “bounded interfaces.”
In HCI, user interface (UI) constraints are mainly used to prevent mistakes. Instead, Nikhita is using constraints to help users while reading and writing, evolving into three trailblazing projects.
Nikhita explored whether technology that restricts the number of words a user could highlight can affect their reading comprehension. She led an experiment where participants read a short story and then completed a reading comprehension test. They were divided into three groups: no highlights, limited highlights of 150 words, and unlimited highlights.
The group with the limited highlights performed the best, with scores almost 10 to 20 per cent higher than their counterparts. Participants noted that the constraints encouraged them to think more critically, such as focusing on keywords.
This novel technology could solve a long-standing problem in education: excessive highlighting. Although highlighting can be an effective studying tool, many people do it mindlessly, making it redundant. Many proposed solutions, like in-person self-regulation training, can be time-consuming and strenuous. Instead, this research can be a faster and more efficient method. It was also the first to prove that restricted highlighting can boost reading comprehension and explore user interface constraints for text marking.
This research won the Best Paper Award at the 2024 Conference on Human Factors in Computing Systems (CHI)— the premier international HCI conference. Multiple news sources, including Phys.org, and MSN News, reported about it.
Similarly, people tend to scroll ahead while reading, which hinders their comprehension. As a result, Nikhita is creating interaction techniques to discourage “mindless scrolling,” such as implementing a timeout between scroll events, adjusting the control-display gain to make scrolling slower, or making scrolling more effortful by requiring multiple fingers or pressure. In the future, she will host an experiment where participants will use these techniques while reading a short story, and then complete a comprehension test.
Recently, Nikhita investigated whether writing with AI can lower psychological ownership, which are feelings of something belonging to someone. In her first experiment, the participants created a 150 to 200 words short story, either independently or by writing a prompt for generative AI. The prompts had different length requirements like three words, 50 to 100 words or 150 to 200 words. She found that longer content, in this case writing independently or using the 150-to-200-words prompt, correlated with high levels of psychological ownership. Namely, the participants would include more details about the plot in their prompts, which produced more personalized stories. She investigated this relationship by running the same experiment but with longer prompts, ranging from 150 to 250 words. However, this effect plateaued at 150 to 200 words.
Nikhita also realized that imposing a strict word count may be unrealistic. Instead, she investigated whether a time delay could encourage users to write more. In her final experiment, participants had to write a prompt for a generative AI, and then hold the submission button for either zero, 20 or 60 seconds. The shorter the prompt, the longer the wait, which could encourage participants to write more. Although the participants tried to submit shorter prompts during the first trial, they eventually wrote longer prompts.
These findings could enhance human-AI collaborations. For example, imposing a long word limit could accelerate a tech support chatbot, as it forces users to explain their problems in detail.
Overall, Nikhita’s dissertation on “bounded interfaces” is transforming how UI constraints are traditionally used in HCI. It has revealed key insights into human behaviour, leading to novel and beneficial technology. Through the EECS workshop, she can share her work with other upcoming and distinguished researchers while learning how to build a successful career as a professor.