February 20, 2022
Abstract
Computational thinking is a critical skillset that must be developed by engineers and other computing professionals. Computational thinking is a skillset that allows people to apply computers and technology to systematically solve problems. The ABET student outcomes and the educational outcomes section of the Taxonomy of Engineering Education acknowledge the importance of this skillset for engineers [1] [2] [3]. Moreover, many introductory engineering courses have a component of programming or computational thinking. Earlier work has shown that computational thinking is important to successfully achieving an engineering identity [4]. First year engineering students matriculate with varying levels of computational thinking. Some arrive adequately prepared for engineering computation because of early academic access to computing courses [5]. Most arrive with only pre-university mathematics and science training. In addition, inequities in course quality, self-confidence, and technology access disadvantage women, African American, and Latinx students [5]. Our multi-institutional team of researchers has been investigating how academic preparation and disparities affect computational thinking skills development, professional enculturation, and educational trajectories of first-year engineering students that are required to take introductory courses in engineering computation. With funding from the National Science Foundation, we are working to answer a broad set of research questions. In this paper, we specifically address this subset: 1. How does the integration of computing into the foundational engineering courses affect the formation of engineers? 2. In what ways do social identities (e.g. gender, ethnicity, first generation, socioeconomic status), choices (e.g. major, transfer status), and other factors impact the engineering student experience with computational thinking? a. In what ways are social identities related to perceived task time, perceived task difficulty, and perceived career confidence? In phase one of our research, we developed an Engineering Computational Thinking Diagnostic (ECTD) that can be used to measure development of computational thinking factors such as the social impact of computing, computational abstraction, data representation, task decomposition, and algorithmic thinking. Pre-validation use of the instrument showed its ability to measure skills growth [6]. We are now in the second phase of our research while the ECTD is formally validated through multi-term large population sampling and factor analysis. In this second phase, we have added a three-question instrument administered at the end of ECTD testing, as well as on submission of critical assignments during the academic term. We call these moments “positions of stress” because they represent times during the term where student career confidence may be changing. The ECTD and position-of-stress instruments were administered at the beginning of the Spring 2021 semester and after the midterm exam. From the population of position-of-stress participants, a stratified sampling technique was used to choose semi-structured interview candidates in three categories of decreasing, same, and increasing career confidence. Stratified subpopulations were identified by confidence level and race, confidence level and gender, confidence level and first-generation status, as well as subpopulations that saw the most drastic changes in confidence. During Summer 2021, semi-structured interviews will be conducted to gain qualitative insight into the computing background of participants, the entry path to engineering, how the participant feels about engineering while progressing through the course, the likelihood of enculturation as an engineer, and social identify factors that may be impacting success. Interviews will be transcribed and coded using an iterative and inductive methodology where a codebook will be developed and iteratively improved to represent the contents of the student interview data. Coding will be done by two researchers working collaboratively using NVivo. To assure reliability, interviews will be coded by individual researchers and compared until the differences between independent codings are minimal. After the differences between coders are minimal, coders will independently code the remaining interviews. Inter-rater reliability will be measured by having the coders evaluate two transcripts and comparing the percentage of the transcript with matching coding. The coded interviews will be used to analyze and explain how computational thinking requirements in engineering differentially impact students with various social identities. We expect to find differences in perceptions related to the privileges experienced by students. For example, we expect that students who have the privilege of prior experience with computing will find the integration of computational thinking into engineering less stress inducing than those without this prior exposure. The data will be represented both in counts of students whose quotes fall into the various codes and by selecting especially informative participant quotes for presentation. We expect that this paper will contribute to the literature by showing how computational thinking may exacerbate existing equity problems in engineering by allowing students with privilege to accrue additional advantages that ease the transition into engineering early in their academic programs. These accrued advantages may diminish the opportunities for success of students with less privilege, who may incorrectly view themselves as having less merit. We expect that this will play out differently in a variety of intersectional social identities, as each individual and subgroup experiences differential privileges. The variety of participants with multiple social identities in our participant pool will provide us with unique opportunities to contribute to understanding the impact of computational thinking on engineering enculturation.