Rodrigo Duran

I am interested in defining and evaluating the cognitive complexity of computer programs. I have developed (with Juha Sorva and Sofia Leite) a model to extract meaningful metrics of cognitive complexity from computer programs (our work presented at ICER 2018, here).

Based on this model, in our recent ACM TOCE paper, we were able to predict the program's difficulty (corroborated by students' subjective ratings) and present some evidence that plan-composition strategies have an impact on the difficulty of comprehending programs.

In our ITiCSE 2017 WG5, we suggested that readability is one of the most important factors students, instructors, and professional developers value regarding code quality. You can see the result of this work here.

I am interested in how students read and comprehend code and how merged vs sequenced plan-composition strategies impact code comprehension and what kind of patterns or strategies students use to read these types of programs.

Philipp Kather, I, and Jan Vahrenhold conducted a study using eye-trackers to investigate, at the plan level, how expert students read code with different complexities and different plan-compositions strategies. Our results align with learning theories such as CTL in showing that code-composition matters more when the program is not too easy or too difficult, and students form a mental model faster when using sequenced code than with merged code. You can read our paper, openly available, at ACM Transactions on Computing Education here.

In 2019, our ITiCSE Working Group investigated how to promote program comprehension activities and how to break down program comprehension activities into subtasks that explore several aspects of students' comprehension of programs. You can read our paper here.

The Brazilian Computer Society is deeply concerned with current trends in gender balance and dropouts in computing programs in Brazil. To understand the reasons why students drop out of computing programs, we conducted a nationwide survey in 2020 with more than 3000 enrolled students to probe potential factors of dropout. Our results published in our ACM TOCE paper suggest that students rank courses as being too theoretical (in their perspective), the difficulty of Math and Programming courses, and the necessity to work full time as the main causes of dropout, in general. However, women rank harassment and a predominantly male environment as the main reasons for dropout.

I have been exploring some learning theories from educational psychology (e.g., CLT, and MHC) and how they could be applied to the programming context. Given my interest in Cognitive Load Theory (CLT), I, Abby Zavgorodniaia, and Juha Sorva reviewed how CLT has been used in CER, particularly how CER has kept up with recent CER advances in terms of methods and conceptualizations of their theory. Our upcoming ACM Transactions on Computing Education paper has a pre-print here. We (and colleagues) also investigated if current instruments that evaluate subjectively cognitive load in computing would be valid using the new 2 CL formulation of CLT. Our results can be found here.

My institution allows me to teach undergraduate and vocational high school students. It is a peculiar context and very little explored:

- Very often I teach the same topic to HS and undergrads. On the same day.

- The programming classes are mandatory for all high-school students! And they learn not only basic programming concepts, but databases, networks, web development, operational systems, and software engineering as well.

- We have gender balance!

- Most HS students don't want to go to a CS undergraduate program later!

I want to explore this context and investigate how the differences in these two contexts impact the instructional design of programming classes, how the HS students learn to program, and how it compares to the Undergrad students. Interested in this research setting? Want to explore it too? Contact me!

Well, a very interesting approach to allow students to express themselves through projects in our HS CS courses is to let them develop any kind of projects, from the most basic to the most complex ones. However, most of these projects are not research-oriented, and traditional venues such as conferences were not suited to this kind of work.

The science fairs proved to be a very engaging context for students. Not only they are able to show their projects to everyone, but also develop their communication, scientific writing, and networking skills. We have been very successful and our students were awarded regional, national, and even being able to show their projects in the biggest stage, the Intel ISEF science fair.

I am interested in how the science fair impacts student engagement, what kind of CS projects students present at these science fairs, how impacts their choice regarding STEM careers, and what kind of difficulties they face when developing this kind of complex project.