Researchers stipulate that a mental model of a system comprises two types of sub-components: parts andstate changes. CS education researchers have noted that state changes in a program are some of the most troublesome concepts to understand. Furthermore, challenges understanding a program's dynamic state changes persist in beginning students even after instruction. Drawing from the theories of mental models, we decomposed arrays into its sub-components of parts andstate changes. Using a questionnaire, we elicited CS1 students' mental models after they received instruction on arrays and then we analyzed and compared the correctness of their mental models with a focus on this decomposition. We compared the mental model correctness of the parts and state changes components. We found that the participants' mental model correctness of parts was significantly higher (i.e., more correct responses) than the mental model correctness of state changes, regardless of teaching modality (online or in-person) or prior programming experience. Moreover, participants with prior programming experience have higher mental model correctness (both for parts and state changes) than participants with no prior programming experience. We close with a discussion of the implications of these findings for introductory courses and highlight recommendations from the literature on ways to teach dynamic aspects of programming.

This paper investigates CS1 students' misconceptions of Java arrays. Each student entering a CS1 course has an initial mental model with a system of beliefs or intuitions about the course domain. The student's initial mental model largely impacts how they integrate new information into it. Oftentimes, these initial mental models are filled with misconceptions. If these misconceptions could be identified early in the semester, educators can attempt to address them with instruction or other educational interventions. But identifying these misconceptions cannot rely on qualitative methods due to the large sections in computing. Instead, we use a quantitative approach to collect assertions that exist in a student mental model. Then we use this representation of mental models to identify misconceptions. We provide a definition of a programming misconception based on mental model theories and use a questionnaire to identify misconceptions based on mental model consistency. We define a mental model as a collection of assertions, and we define a consistently chosen wrong assertion as a misconception. We have broken down Java arrays into a set of parts and state changes. The parts include array name, index, type, and element access. The state changes include array declaration, array instantiation, assignment of literals, and the more general assignment of arrays. We placed 30 wrong assertions multiple times as distractors in our questionnaire. In this paper, we report on our data collection and discuss the 16 misconceptions of arrays found to be held by novice programmers before classroom instruction on arrays. Nine misconceptions were documented for parts components and seven for state changes. Our results show that over half of our participants (out of 93) held at least one misconception before learning arrays in classrooms. CS1 students mostly held misconceptions related to the arrays' declarations (state change), and name, and element access (parts). From our data, the top most common misconceptions were about array declarations and initialization.

Novice programmers often make haphazard mistakes due to their incomplete and inconsistent mental models. Previous studies have indicated that students have a non-viable mental model of fundamental programming concepts. From theories of mental model, Mayer [13] showed the effectiveness of explanative diagrams in shaping a novice’s mental model. My dissertation aims to evaluate the role of Mayer’s explanative diagram as a representation of a notional machine to shape novice programmers’ mental models of the array.

A mental model is a knowledge structure that reflects a learner's understanding, action, and behavior about a device in the real world. Novice programmers have been found to have inconsistent and flawed mental models of very basic programming concepts. This poster presents the development of a multiple-choice questionnaire to elicit and measure the consistency of a novice programmer's mental model of arrays in Java.

Diagrams in textbooks are an essential tool to explain concepts. Aneffective diagram should help a novice learner to build a runnable mental model, increase recall, and improve problem-solving skills. Richard Mayer and others suggest that to have an impact on the readers' mental model, diagrams must document two major features:system topology andcomponent behavior. The presence of these two features makes a diagramexplanative. In this paper, we propose a framework of what constitutesexplanative diagram for variables, arrays, and objects based on Mayer et al.'s definition. We used our framework to analyze diagrams of variables, arrays, and objects in 15 commonly used introductory Java textbooks to ascertain how these concepts are illustrated, annotated and explained. Our results show that none of the textbooks provide what we would consider as explanative diagrams. We conclude with an assessment of diagrams in introductory programming textbooks and present open questions for further study.

Drawing is an effective learning tool, incorporating engagement, increasing recall, and improving problem-solving skills. However, it is not widely practiced by computer science educators. This poster reports an analysis of 15 commonly used CS1 textbooks to investigate whether authors encourage students to draw basic programming concepts. One CS1 textbook contained textual encouragement to draw class diagrams and exercises to draw UML diagrams. One other book asked students to graphically represent a class object. We found no encouragement or instructions to draw variables or arrays.

Diagrams in textbooks are essential tools in explaining concepts. An effective diagram helps a novice learner build a runnable mental model, increase recall and improve problem-solving skills. Richard Mayer suggests that for a diagram to be effective, it must beexplanative and, thus, document two major features: \textitsystem topology andcomponent behaviour. We surveyed diagrams of variables, arrays and objects in 15 commonly used introductory Java textbooks to investigate if the diagrams are explanative. This abstract presents the analysis of arrays.

An organization with high undergraduate enrollment, the College of Computing and Informatics atUNC Charlotte is undergoing a sustainable pedagogical shift. Our Graduate Teaching Assistants (GTAs) being a crucial part to sustain this change, are also facing a shift in their climate. We aim to present measurement to gauge our GTAs' climate under this pedagogical climate shift. For this purpose, we have analyzed 184 survey responses from GTAs and developed three constructs:Self-Competence, GTA-to-Faculty Relations andCommunity Belonging. Exploratory Factor Analysis was used to identify the underlying factors, exhibiting 13 items retaining to these three constructs with a Cronbach's alpha of 0.94. This measurement shows that with engaged classroom practices, we are also fostering an engaged climate for our GTAs.

This innovative practice work-in-progress paper presents measurement and analysis of a comprehensive model of pedagogical change in the computing college of a large, urban university in the United States. Our approach to change has two major thrusts: (1) a model for systemic change in the culture and climate of computing education, and (2) pedagogical patterns that transform students who enter undergraduate computing programs from people with an interest in computing to people who identify as computing professionals. We collected a total of 940 student survey responses across four semesters. Survey items gauged students’ connection to their peers and their profession and were used to create a composite measure of peer learning and professional identity. Exploratory Factor Analysis indicates an internally consistent factor structure composed of 18 items with a Cronbach’s alpha of 0.91. In this paper, we present the methodology for measurement and analysis, describe our rationale in an engaged computing education context, and discuss how studying the constructs of peer learning and professional identity contributes to the knowledge base of the computing education community.