As Machine Learning (ML) becomes increasingly integrated into our daily lives, it is essential to teach ML to young people from an early age including also students from a low socioeconomic status (SES) background. Yet, despite emerging initiatives for ML instruction in K-12, there is limited information available on the learning of students from a low SES background. To address this gap, our study conducted an analysis of comparing the students’ performance assessment scores of ML concepts as a result of the ML4ALL! course among 266 middle and high school students from different socioeconomic backgrounds. The results demonstrated an understanding of ML concepts among students from all SES backgrounds. Although some differences were observed regarding specific parts of the ML development process, these were not substantial enough to identify SES as a determining factor affecting the performance assessment score. Also, when considering the background together with other demographic factors such as sex assigned at birth or educational stage, no significant difference of the students’ performance assessment scores was observed. These findings provide a first indication that a low SES background must not be a barrier to ML competencies and that effective and inclusive ML teaching strategies can ensure equitable access to ML education across diverse socioeconomic backgrounds.

Currently, there is little evidence about how non-English-speaking students learn computer programming. For example, there are few validated assessment instruments to measure the development of programming skills, especially for the Spanish-speaking population. Having valid assessment instruments is essential to identify the difficulties of the students, make formative decisions, and evaluate the impact of pedagogical interventions. This article presents the outcomes of the early steps toward a validation in Spanish of the SCS1 assessment, derived from the first conceptual inventory to measure programming knowledge decoupled from a particular language. The goal of this initial step is twofold: (1) identifying the questions that are too difficult in this version and might need to be adapted based on empirical evidence of how students answer them and (2) identifying students’ common difficulties with the most challenging questions. This study started with a “think-aloud” protocol to validate whether the Spanish version was similar to the original instrument and if the students understood the translation as intended. Then, the instrument was applied to a sample of 71 university students who had previously taken an introductory programming course using Python. We computed the difficulty index for each question and identified students’ difficulties in answering them. We also propose potential improvements for these items. The main contributions of this study include (1) progress toward the validation of an instrument for evaluating programming skills in Spanish, (2) noting the advantage that experienced students in Python have to complete the instrument, (3) identifying the tendency of students to translate the pseudo-code from Spanish into English, and (4) the identification of common mistakes, like misunderstanding the scope of local variables or the confusion between returning and printing.

Interdisciplinarity has been touted as a means to recruit more racially and gender-diverse students to computing. In this explanatory sequential mixed-methods study, we investigated demographic characteristics among a sample of undergraduate students pursuing interdisciplinary computing major and minor combinations at 15 institutions in the United States who completed a survey at the end of their introductory course. Descriptive analyses of responses to this survey of introductory computing students revealed that enrollment in interdisciplinary major and minor combinations was limited and did not appear to disproportionately attract women or Black/African American, Latine, Indigenous, and/or Multiracial students. We then conducted a directed content analysis of departmental websites to examine the language and policies that may have precluded or encouraged students to pursue interdisciplinary computing major and minor combinations. Findings revealed that departmental offerings of such programs were limited, and, among those that did offer such programs, communication about their goals and requirements was often lacking. Implications for research and practice, especially as they pertain to efforts to broaden participation in computing, are discussed.

Learning to respond to a computer program that is not working as intended is often characterized as finding a singular bug causing a singular problem. This framing underemphasizes the wide range of ways that students and teachers could notice discrepancies from their intention, propose causes of those discrepancies, and implement interventions. Weaving together a synthesis of the existing research literature with new multimodal interaction analyses of teacher–student conversations during coding, we propose a framework for debugging that foregrounds this open-endedness. We use the framework to structure an analysis of three naturalistic debugging situations (with US 5th–10th graders) that range from solo debugging to collaborative discourse. We argue that a broken computer program is a polysemous object through which teachers and students actively and publicly notice, reason about, and negotiate different debugging pathways. We document students and teachers improvisationally altering a debugging pathway, justifying a particular pathway, and outwardly discussing competing pathways. This paper provides a framework for structuring debugging pedagogy to be less about scaffolding a student toward a specific pathway to a fix and more about exploring multiple possible pathways and judging the (learning) value of various routes.

The current study compared American, Korean, and Indonesian middle and high school students’ CS attitudes. Concurrently, this study also examined whether the items in the CS attitudes scale exhibit country and gender measurement biases. We gathered data on CS attitudes from middle and high school students in the US, Korea, and Indonesia. The participating students took the same (translated) previously validated CS attitudes scale. We ran a unidimensional IRT, differential item functioning (DIF), a two-way ANOVA, and the Kruskal-Wallis H test. Despite the valid instrument, we found it inappropriate as is for international comparison studies because students from different countries interpreted some items differently. We then compared gender-based differences in CS attitudes across countries. The results revealed no significant differences between males and females in the Indonesian middle school data, whereas male students had significantly higher CS attitudes than female students in both American and Korean student data. Furthermore, we found the same pattern in gender differences in Korean and Indonesian high school students’ CS attitudes scores as in the middle school study. These findings underscore the importance of a country’s sociocultural context in influencing gap and diversity in secondary school students’ CS attitudes.

Motivation : Enrollments in Brazilian computing degrees are at an all-time high, but graduation numbers have not increased at the same rate. Moreover, enrollment growth has mainly attracted male students, steadily expanding the gender gap in Brazilian computing programs. Such high attrition rates have a great economic impact and may disproportionately affect women and students of color. Previous works investigated reasons for student drop-out and retention in specific courses or barriers to entrance in computing programs in narrower contexts or in a single institution. Objectives : We investigate potential actionable factors for intent to drop out in computing programs and what factors might lead students to remain in a computing program in several Brazilian institutions. We investigated how such factors may be affected by students’ race/ethnicity, gender, and socioeconomic status. Method : We analyzed Likert-style answers from an online survey with 3,193 students currently enrolled in Brazilian computing programs. Results : The results show that students value salary/job-related factors as the most important factors to potentially remain in a computing program. The excess of theoretical courses and the difficulty of programming and mathematics courses are the top-ranked factors by students to potentially abandon a computing degree. However, while there is little effect of gender, race/ethnicity, or socioeconomic status in retention factors, potential drop out factors such as a the fact that it is a male-dominated field, harassment, and the difficulty of courses were rated as more important by women. Also, costs and the difficulty of courses are relevant factors for the intent to drop out when analyzing students’ race/ethnicity and socioeconomic status. Discussion : We explore the implications of our findings for Computing programs, particularly (but not restricted to) the Brazilian context. We conjecture reasons for such students’ perspectives regarding intent to drop out and retention factors and provide recommendations of actions for instructional designers, curriculum developers, and other key stakeholders to address issues related to gender, students’ wellness, perceived authenticity of courses, and other relevant factors. Since we only observed small interactions between race/ethnicity and retention and intent to drop out factors, which may indicate a lack of sensitivity from the instrument, we present suggestions to address such limitations in future work.

What if “regular” Computer Science (CS) faculty each taught elements of inclusive design in “regular” CS courses across an undergraduate curriculum? Would it affect the CS program's climate and inclusiveness to diverse students? Would it improve retention? Would students learn less CS? Would they actually learn any inclusive design? To answer these questions, we conducted a year-long Action Research investigation, in which 13 CS faculty integrated elements of inclusive design into 44 CS/IT offerings across a 4-year curriculum. The 613 affected students’ educational work products, grades, and/or climate questionnaire responses revealed significant improvements in students’ course outcomes (higher course grades and fewer course fails/incompletes/withdrawals), especially for marginalized groups; revealed that most students did learn and apply inclusive design concepts to their CS activities; and revealed that inclusion and teamwork in the courses significantly improved. These results suggest a new pathway for significantly improving students’ retention, their knowledge and usage of inclusive design, and their experiences across CS education—for marginalized groups and for all students.

Objectives . Automatic assessment systems are widely used to provide rapid feedback for students and reduce grading time. Despite the benefits of increased efficiency and improved pedagogical outcomes, an ongoing challenge is mitigating poor student behaviors when interacting with automatic assessment systems including numerous submissions, trial-and-error, and relying on marking feedback for problem solving. These behaviors negatively affect student learning as well as have significant impact on system resources. This research quantitatively examines how utilizing submission policies such as limiting the number of submissions and applying regression penalties can reduce negative student behaviors. The hypothesis is that both submission policies will have a significant impact on student behavior and reduce both the number of submissions and regressions in student performance. The research questions evaluate the impact on student behavior, determine which submission policy is the most effective, and what submission policy is preferred by students. Participants . The study involved two course sections in two different semesters consisting of a total of 224 students at the University of British Columbia, a research-intensive university. The students were evaluated using an automated assessment system in a large third year database course. Study Methods . The two course sections used an automated assessment system for constructing database design diagrams for assignments and exams. The first section had no limits on the number of submissions for both assignments and exams. The second section had limits for the exams but no limits on assignments. On the midterm, participants were randomly assigned to have either a restriction on the total number of submissions or unlimited submissions but with regression penalties if a graded answer was lower than a previous submission. On the final exam, students were given the option of selecting their submission policy. Student academic performance and submission profiles were compared between the course sections and the different submission policies. Findings. Unrestricted use of automatic grading systems results in high occurrence of undesirable student behavior including trial-and-error guessing and reduced time between submissions without sufficient independent thought. Both submission policies of limiting maximum submissions and utilizing regression penalties significantly reduce these behaviors by up to 85%. Overall, students prefer maximum submission limits, and demonstrate improved behavior and educational outcomes. Conclusions . Automated assessment systems when used for larger problems related to design and programming have benefits when deployed with submission restrictions (maximum attempts or regression penalty) for both improved student learning behaviors and to reduce the computational costs for the system. This is especially important for summative assessment but reasonable limits for formative assessments are also valuable.

This article reports the analysis of data from five different studies to identify predictors of preservice, early childhood teachers’ views of (a) the nature of coding, (b) integration of coding into preschool classrooms, and (c) relation of coding to fields other than computer science (CS). Significant changes in views of coding were predicted by time, prior robot programming experience, and perceptions of the value of coding. Notably, prior programming knowledge and positive perceptions of mathematics predicted decreases in views of coding from pre- to post-survey.

Debugging is a core skill required by programmers, yet we know little about how to effectively teach the process of debugging. The challenges of learning debugging are compounded for novices who lack experience and are still learning the tools they need to program effectively. In this work, we report a case study in which we used a think-aloud protocol to gain insight into the behaviour of three students engaged in debugging tasks. Our qualitative analysis reveals a variety of helpful practices and barriers that limit the effectiveness of debugging. We observe that comprehension, evidence-based activities, and workflow practices all contribute to novice debugging success. Lack of sustained effort, precision, and methodical processes negatively impact debugging effectiveness. We anticipate that understanding how students engage in debugging tasks will aid future work to address ineffective behaviours and promote effective debugging activities.