How Can AI and Project-Based Learning Interact to Enhance Student Learning

AI and PBL Interaction

In an authentic project-based learning environment, students are required to focus on solving a problem. Those problems are usually multidisciplinary and require students to identify its different disciplines, which are then assigned to different members in the team. Each team member conducts research and investigates their discipline, while making sure to collaborate with the other team members in order to coordinate their efforts towards their mutual goals. The entire endeavor is an exercise in good investigative, collaborative and management skills, that no AI tool can orchestrate for students. In project-based learning students actually need to take a leading role on the entire project and do the work themselves. However, by using AI tools to help their project, they can come up with a final product that is superior to what they would otherwise have without it. Let’s explore how to do this in a STEM project. Here are several examples of using Chatbots in different parts of the project.

Project Parts

Identifying the project parts that are absolutely necessary for solving or responding to the driving question is not easy. It requires research and it is an iterative process because as students work on the project, they discover other necessary components for reaching a wholesome solution. AI can make this process much easier for students from which they can learn a great deal. Students can use ChatGPT or any other ChatBot to provide them with an outline of the things that are necessary for  investigating their driving question. Here is an example of a query I provided ChatGPT “Can you provide an outline of the components that I need to investigate in order to solve this problem “How does extreme heat affect our city and what can be done to improve the situation?”” The outline that I got provided me with much more than I wanted to cover in this project, but it gave me the main project parts that I need to investigate in this project:

Effects of Extreme Heat on Cities:

  1. Physical Effects
    1. Infrastructure Damage
      1. Roads and Bridges
      2. Buildings and Homes
      3. Power and Utility Systems
    2. Environmental Impact
      1. Increased evaporation rates in water bodies
      2. Drying out of green spaces
      3. Changes in urban biodiversity
  2. Human Health Effects
    1. Direct health impacts
      1. Heat strokes
      2. Dehydration
      3. Burns
      4. Exacerbation of chronic illnesses

ChatGPT provided a much more comprehensive outline, but this is probably enough for students to understand the effects of extreme heat and to come up with ideas on how to counteract the extreme heat impact. Using a ChatBot can make this process much easier for students because it helps narrow down the problem at hand and navigate them in the right direction. 

Some of the more complicated issues that students have difficulty with in project-based learning are: generating a driving question, coming up with the topics they need to learn in order to answer their driving question and finding reliable resources for investigating their topic. To help out students with this process, we created a GPT using OpenAI’s GPT development tool. Instead of students having to figure out the driving question of a topic of interest, the topics they need investigate to answer the question and the appropriate resources, the Project Topic Advisor tool can help them do all these things. Check it out on our Project Topic Advisor page.

How to Conduct Research

After being assigned their part of the project, each team member has to focus on researching their part of the project. Research contains the ability to search correctly, to be able to navigate ambiguity, triangulate knowledge, and question the material that was found. This blog entry can provide you and your students with valuable information on  why research is so important and How to teach it to students. This blog also directs the user to two excellent resources for teaching research skills. However, these resources are not helping students determine which sources would be better for their investigation than others or provide them with a short summary of what each resource provides. 

There are AI-powered tools that can explore the internet for you, organizing your data into digestible visualizations, and even formatting references correctly. One such AI research app is Genei, that will search the web for relevant content, offer suggestions to add new sources, crunches the content, offers a summary, extracts a list of common keywords along with all the images inside—and, to top it off, rounds up all the references to other work.

Another research tool that students will be able to use is provided by Google. Google’s AI-powered Search Generative Experience (SGE) will be able to summarize articles that students are reading on the web. SGE can already summarize search results so that students don’t have to scroll forever to find what they are looking for, and this new feature is designed to take that further by helping them out after they have actually clicked a link.

This can be a tremendous help for students because this is the part of project-based learning that most students fail on. Having access to a helper app that can find relevant resources, summarize them for the students and format the references correctly can free up students to focus on analyzing which materials will fit which parts of the project and how to analyze and synthesize everything they learned. All skills that require their executive faculty. 

Scheduling Tasks – One of the most important things in a team project is the ability to manage tasks, prioritize them and follow up on task completion. In the past, well managed teams used the calendar to schedule the tasks of the entire project, but many times got confused how to prioritize tasks and synchronize tasks for all team members. AI builds the entire schedule for the team by synchronizing tasks based on task priorities and it makes following up on task completion much easier. One such AI powered application is Motion, which focuses on project management, helping students keep track of all the tasks they still have to complete. By tweaking each task priority, the student will be telling the AI engine when it should land on the calendar and how to place the other tasks around it.  

Summarizing Meetings – Up until the advent of AI, student meeting conversations were usually lost. AI changed all that and there are AI transcription apps that can turn voice into text, letting students browse it later. This can help students be more present in their meetings and, at the same time, be able to thoroughly analyze the transcription later. One such transcription app is Fireflies, which can transcribe all student meetings, tracking the conversation topics along the way. It has its own bot called Fred that can handle summarizing the meeting’s contents, generating text, and searching through the history to meet students’ queries.

Presenting the Project – Once students have their final essay, they can further use AI to create a presentation based on their essay. Many AI tools can create a presentation based only on a topic. Of course, that will not require student work at all. That is not what I am recommending here. Students can use to input their text, choose the total number of slides, and let do the heavy lifting of transforming the text into visually appealing slides. Once students have a professionally crafted presentation, they can then use a different AI tool to coach them to present their presentation in front of the class. PowerPoint Speaker Coach feature is specifically designed to evaluate various aspects of a presentation, including pacing, pitch, use of filler words, and other common speaking habits. It allows students to practice in private, and the feedback provided can help identify areas for improvement.

Project Assessment

Project-based learning assessment is not easy because the project environment does not lend itself to quantitative assessment. Project assessment is mostly qualitative. In addition, the teacher is not always present while students work and there is a lot that students can tell about each other’s work in the team.  A series of research studies and surveys of students provided the following recommendations about project-based learning assessment: 

  1. Tailoring the assessment to the learning environment and using a diversity of types of assessment
  2. Students’ preferred assessment type was the reflective journal, which they felt gave an insight into group dynamics, facilitated feedback on the project and enabled students to explain their performance.
  3. Formative assessment of journals was preferred but students were also happy for part of this to be summative.
  4. Students felt it essential that the reflective journal should be kept confidential from other students and entries not too frequent.
  5. Students were wary of self-assessment and peer assessment as being too subjective but supported co-assessment (by peers and staff) as this provided 
  6. an element of peer assessment with the perceived “safety net” of staff evaluation.

If the implementation of an assessment process seems too time consuming and complex, AI can be helpful for both teachers and students. Gradescope is an AI tool that enables students to assess each other while providing feedback, which saves time and energy. Gradescope also enables the teacher to effortlessly manage and evaluate all assessments, whether they are conducted online or in a physical classroom. By using it, teachers can streamline their grading process and gain a comprehensive understanding of their students’ progress.