Michelle Niculescu, Ph.D. Psychology

Photo of Michelle Niculescu

Michelle Niculescu, Ph.D.
Psychology

Using H5P to help students build foundational knowledge

Dr. Niculescu explored using H5P, a web-based tool to create interactive HTML5 content. H5P is integrated into Brightspace and has a variety of tools including the single choice set used to this project.

 

 

Project Overview

Using H5P to guide students through a problem-based learning activity in PSY213 – Statistics and Research Design. Often students enrolled in PSY213 come to me disliking math/think they are bad at math/wondering how they’ll need or use math in the future. To that end, we focus equally on the “how” of statistical analysis and the “why.” From when I started teaching PSY213, I incorporated problem sets into every section so that students could see how we use statistical thinking and analysis in everyday lives (particularly as behavioral scientists). However, while I was doing a good job of getting students were seeing the big picture, they were still missing that they needed to understand the details from class to solve the higher-level problem. Luckily, we gained access to H5P which helped me create formative assessments (problem sets) that allow students to work through the problems, checking that the essential objectives are met with specific feedback from me based on their responses. Bonus – it also integrates seamlessly into BrightSpace.

Sample of H5P options

A few examples of H5P tools

 

Planning Process

Problem: When starting Tech Fellows, I knew that the problem existed that my students had access to real-world application of the content presented in my courses through assignments designated problem sets. However, in all my courses, I noted that they could big-picture problem-solve but were still struggling with how the basics fit into evidence-based reasoning. I wanted something that facilitated UDL and with a focus on diversity, equity, and inclusion.

Possible Solutions: My idea was to have conceptual checks embedded into the problem sets to make sure students were relaying on course material to problem-solve. I decided to start with PSY213: Statistics and Research Design because this is the only course where we go directly from what is in the textbook to application. My other courses have an intermediate step of what is in the current research.

With guidance, I explored many options.

  • Panopto – videos embedded with questions.
  • Brightspace – quizzes with conditions (a right answer on a previous quiz) to move forward.
  • H5P – around the same time I was struggling with Brightspace to make the quizzes we needed, we gained access to H5P. This tool allows feedback to each response (right and wrong) and allows students to determine the correct answer before proceeding forward.

Decision-making guidelines: The student learning outcomes of the course include interpreting and applying descriptive and inferential statistics, including graphing and utilizing statistical analysis software (SPSS). My goal with the problem sets was to improve student’s foundational knowledge, specifically understanding to apply that knowledge to real-world problem-solving. This includes a specific focus on the human dimension through utilizing questions that apply directly to evidence-based reasoning specific to healthcare professions, a common career goal for the majority of my students’.

I needed a tool that would allow me to redefine my previous assessment, giving me the ability to give my students specific and detailed feedback based on their responses.

Most importantly, I needed a tool that makes UDL simple and user-friendly (from the perspective of design and utilization). I wanted my student to be able get feedback immediately. Good and thoughtful feedback takes time. I wanted a tool that would allow my students to explore their understanding without consequences of missed questions. They also needed the space to identify where they struggle and need to look further.

 

Implementation

My goal was to build my students’ foundational knowledge through statistical problem-solving utilizing checkpoints. I was able to meet my objective best using H5P.

Initially, I attempted to do knowledge checks with individualized feedback in BrightSpace. This was time consuming for me and increased anxiety in students. The problem sets were designed to be low stakes learning, but when they could see each item they got wrong and were not able to correct their mistakes. Students were focusing more on the outcome than the process.

Then we got access to H5P. I now put my problem sets from past semesters into H5P. With a class of 30 students, I also did not have time to give them detailed feedback on the multiple assigned problem sets. I utilize mistakes made by previous students and then provide detailed feedback about what went wrong for each option they choose. They can retry the assessments as many times as they would like. Every student can achieve a perfect score. Students can also access the problems and solutions when studying for exams.

Here is an example question from one problem set:

Example of problem set 1

 

While initially the development of these problem sets can be time-intensive (at least two hours including looking at previous students’ responses to anticipate potential mistakes), once they are complete, they can be used from semester to semester.

For every image I include, I also include a caption. H5P requires images included have a caption for UDL purposes. Often, I will re-use the same image/table for multiple questions, and this is also done easily. Finally, this allowed my students to not only process each question at their own pace.

 

Assessment

Since this was done in a statistics course, I must do some sort of quantitative analysis!

I have always utilized problem sets, but I switched from open-ended questions to multiple choice with specific feedback. Grades on exams (Exam 2 specifically) did vary from Spring 2021 (open-ended problem sets, M = 83.29, SD = 12.37) to Spring 2022 (H5P problem sets, M = 89.24, SD = 12.54), but the difference was a non-statistically significant trend, t (51) = -1.729, p = .09. In addition, in the Spring 2022 course, that students that completed their H5P problem sets did significantly better on their exams (again, Exam 2) than those that did not, F(1,28) = 14.44, p < .001. This could be due to a variety of factors, including that those students that complete their work are better-prepared and likely more motivated to do well on exams. However, this does support that completing H5P problem sets did predict better exam grades for students.

I also utilized a qualitative assessment of how students liked the new version of problem-based learning. This was in Brightspace before I worked out all the kinks, so they were not receiving feedback until they submitted the entire problem set. Still, feedback from students was generally positive. In summary, when the problem sets were open ended, students often felt lost as to where to begin. This new way gave them a sense of empowerment to determine the foundational concepts instead of guessing on what they needed to know.

I think that this way of doing the problem set was helpful especially doing it in class in case we needed help. Also, group work is very helpful since some people get it more than others and they may have different ways of explaining things.

Doing the problem set this way helped me a lot more, to better understand the problems.

I did not mind doing the problem set this way, as I typically do the problem set in one sitting, so having the problem set in a quiz format does not really effect me.  Instead of doing my work in the assignment, I am able to do it in the TextEdit application on my laptop.

The method now being used, where we complete the problem sets online and have the ability to go back and see what we did wrong is great! Thank you for trying new things, it really helps deepen my understanding of the equations and the calculation process.

Yes i feel like this slightly helped.

I really like this way of doing things, I felt a lot more confident in my work doing this method and I really feel like I understand it better.

I liked this way a lot better. It made more sense and was easier to visualize. I definitely understood this problem set better then any other the others. The only downside is that it was really hard to type equations into the quiz.

I like it a lot better. Its definitely easier to grasp the bigger concepts. I didn’t feel like the other way made stats more real, both give that experience. There was always some confusion about how to answer certain questions but now its very strait forward. However, that might be bad for us in the end. It might be beneficial to throw us in uncomfortable situations.

I personally prefer the way we usually do problem sets. But, this did help me to grasp important concepts.

I liked it and thought it was useful. It was helpful to work through it in class with those around us and felt less overwhelming then seeing the word document with all the questions and information

I liked it a lot more than the other way of doing problem sets.

I definitely enjoyed this a lot more and felt that it allowed me to easily work through the problems with understanding. I also think that this method has kinks but once the kinks are worked out it would function perfectly.

This helped a ton, I love going step by step it helps me understand WHY I’m doing it and what Chi-Squared really means as well as making the work easier, at least for me.

I feel like I grasped the concept, however question 7 confused me slightly. I feel like the chi-square is outside of the critical region because x^2 = 9.84, which is greater than the value. But there is significant difference in the way orientations were selected. That’s the only confusion I had doing this, though.

I like the usual formats of the problem sets better as I feel like I can take more time to work on and understand each question.

I felt ok about this, getting to do more problems in class really helps

I don’t mind doing them either way!

I felt as if this problem set was fairly simple, and I would feel comfortable seeing the material on this problem set in the future. I don’t think I have a preference between the two formats of doing these problem sets (Word vs Quiz) although I think the somewhat instant feedback is definitely a help.


Reflections and Next Steps

I learned a great deal from this experience. I improved my ability to design assessments in H5P and my understanding on how to best use Brightspace. I got feedback from students throughout and I think they really appreciated seeing me struggle through this with desired outcome of their best learning.

One limitation is theoretical in that while my students are getting detailed and specific feedback, they are now limited in the space they have to generate these ideas independently and to make mistakes. I hope to figure out ways to balance this in the future.

Another limitation is technical. In H5P, the grading aspect of Brightspace integration can be frustrating. Someone needs to complete the problem set before a grade appears. You need the grade to appear before you can edit the settings, including the points. The automatic setting is also to exclude every score generated by H5P from the final grade, so that also needs to be adjusted each time.

I look forward to utilizing this in my other courses. I envision doing an escape room style learning assessment. Again, I will need to balance specific guidance with the space for them to create and stumble through the learning process independently.