The snappier block editor and read-write interface

NORTHWESTERN UNIVERSITY

Modality Matters: Understanding the Effects of Programming Language Representation in High

for the degree

DOCTOR OF PHILOSOPHY

September 2016

2 Abstract
Computation is changing our world. From how we work, to how we communicate and how we relax - few parts of our world have been left unaffected by computation and the technologies that it enables. The field of computer science and the ideas of the discipline are driving these changes, yet relatively little of it is present in contemporary K-12 education. Numerous local and national initiatives are underway to bring the powerful ideas of computing into classrooms around the world. An increasingly popular strategy being employed in this effort is the use of graphical, blocks-based programming environments like Scratch, Snap! and Alice.

At the outset of any academic pursuit, you are challenged by the adage: stand on the shoulders of the giants3. With that in mind, it is important for me to acknowledge the visionaries and great thinkers upon whose shoulders I peer out on the world from. Two in particular stand out: Seymour Papert, whose ideas and visions we are still chasing, and Andrea diSessa, for writing the book Changing Minds, which set me on the course that led me to the Learning Sciences and everything that followed. For a full list of giants, please refer to page 350.

It is also important to me to acknowledge two other groups of people who made this dissertation possible. First are the talented teachers and brilliant students who participated in my studies, without whom, there would be no dissertation. Secondly, are the talented, resourceful

Finally, this dissertation would never have been possible without my family. To my parents, who always encouraged me to ask questions and supported me in finding answers. To my sisters, for their support and never-ending older-sibling wisdom. To my children, Maya and Jonah, whose arrival precipitated my finishing this document, and quickly replaced it as the central focus of my daily life. And finally, and most importantly, to my wife, who has been by my side offering help, support, and love on every day of this journey. Thank you for helping me become the person I am.

6 Dedication
This dissertation is for Jonah and Maya. It is for you and your future friends, classmates, and peers that I undertook this challenge and sought to answer the questions I did. My only hope for this work is that it will make some small difference in your lives and the lives of learners of your generation.

Table of Contents .............................................................................................................. 7

List of Figures .................................................................................................................. 11

4.Design ...................................................................................................................... 100 Modality ............................................................................................................................. 100 Year One – Snappier! ....................................................................................................... 102

8 The Three Versions of Snappier! ................................................................................................... 103 Findings from Year One with Snappier! ..................................................................................... 107 Limitations of Snappier! .................................................................................................................... 117 Year Two – Pencil.cc ........................................................................................................ 119 The Three Versions of Pencil.cc ..................................................................................................... 123 Limitations of Pencil.cc ...................................................................................................................... 126

9.Discussion and Conclusion .................................................................................... 310 Review of the Program of Research ................................................................................ 310 Summary of Findings ....................................................................................................... 313 Comparing Blocks and Text Modalities ...................................................................................... 313 The Case of the Hybrid Modality ................................................................................................... 319 Implications ....................................................................................................................... 324 Implications of Modality on the Learner ................................................................................... 324 Implications of Modality on the Teacher ................................................................................... 327 Implications of Modality Choice on Schools and Administrators ................................... 329 On Modality, Learning, and Design ................................................................................ 330 Modality Matters .................................................................................................................................. 331 Modality is Malleable .......................................................................................................................... 333 Modality and Structurations ........................................................................................................... 336 Modality and Milieu ............................................................................................................................ 341 Limitations and Future Work ......................................................................................... 343 Conclusion ......................................................................................................................... 348

10 10.References ............................................................................................................... 350

14.Appendix D – Interview Protocols ....................................................................... 413 Pre Interview Protocol ..................................................................................................... 413 Mid Interview Protocol .................................................................................................... 415 Post Interview Protocol .................................................................................................... 415

15.Appendix E – Coding Manuals ............................................................................. 417

13 List of Tables

Table 3.1. The 5-week introductory curriculum ............................................................... 71 Table 3.2. The 13 assignments of the introductory curriculum ........................................ 71 Table 3.3. The concepts covered in the Commutative Assessment .................................. 78 Table 3.4. The 35 student interviews ................................................................................ 82 Table 3.5. Log events captured in Pencil.cc ..................................................................... 85 Table 3.6. Log data capatured in Pencil.cc ....................................................................... 85 Table 3.7. Data captured by JavaSeer ............................................................................... 87 Table 3.8. Student responses to why they decided to enroll in the class. ......................... 97 Table 4.1. Mappings between Snap! and JavaScript ...................................................... 104 Table 4.2. Student responses comparing Snappier! and Java (quantitative) .................. 108 Table 4.3. Student responses comparing Snappier! and Java (qualitative) .................... 113 Table 6.1. Sample responses of function metaphors used by students ........................... 180 Table 7.1. Frequency of data collection events in Pencil.cc ........................................... 241 Table 7.2. Run events collected for each assignment ..................................................... 242 Table 7.3. Frequency of consecutive runs in under five seconds ................................... 247 Table 8.1. Pencil.cc concepts perceived as useful for Java ............................................ 275 Table 8.2. Successful compilations by Levenshtein distance ......................................... 293 Table 8.3. High-level descriptive patterns of failing compilations and errors. ............... 298

While recent attention has focused on computer science and programming, that is just one of the many ways the computers interact with learning. Since the emergence and recognition of computers as ‘Protean ‘devices with widespread applications major scholars have seen great potential in their use for learning and education (A Kay & Goldberg, 1977; Papert, 1980; Perlis, 1962; Suppes, 1966). The role computers were to play amongst these and other early advocates of computers as tools for education differed. Some were excited by discipline of computer science itself, while others saw it having much wider potential, spreading across disciplinary boundaries. Some focused on the power of programming as a learning activity, including the Constructionist learning approach, while others focused not on the practice of programming but the tools and environments that could be built through programming, such as cognitive tutors and

16 other forms of computer-aided instruction. Others viewed computers as supports for
standardized content-based instruction while others viewed computers as a medium for
supporting personal expression and fostering creativity, art, and the creation of personally meaningful artifacts. Others still argued for learning computer science ideas is broadly applicable both when working on a computer, as well as for being more efficient and success at non-computer-based activities. Collectively, these various perspectives on the role of computation in education and their work, along with others in the decades since, have produced a plethora of technologies, curricula, pedagogies, and learning experiences designed to leverage the power of computers. Having recognizing the larger picture of the intersection of computers and
computational technologies and learning, we bring the discussion back to programming and computer science, the focus of this dissertation. A longer discussion of this broader computer and learning landscape will be discussed in the next chapter, along with situating the work being presented within this larger orientation.

Given this co-occurrence, it is not surprising that many of the new curricula that are being designed and widely adopted rely on recently designed programming tools that have little empirical evidence with respect to their effectiveness for high school aged learners in formal settings. As such, there is great need and opportunity for investigating the effectiveness of design features of the current generation of introductory programming tools in formal settings. This dissertation will answer three sets of interrelated research questions all of which address different facets of the guiding question: How best can we design high school computer science learning environments to educate the next generation of computationally literate citizens?

The first set of questions investigates the relationship between the modality students use while learning to program and the resulting attitudinal and conceptual outcomes. By modality we mean to capture the representational infrastructure used to depict the program, as well as the various forms of composition supported by the representation and the affordances it provides for each. Understanding the impact of modality is important because, as previously mentioned, new environments for teaching programming are emerging and becoming increasingly used in formal educational settings, but we lack a clear understanding of the relationship between these new tools and the resulting conceptual gains, attitudinal outcomes, and programming practices they promote. Prominent among the features of these new environments is the introduction of graphical, blocks-based interfaces (Figure 1.1) that allow learners to use only a mouse to drag-and-drop commands together to form functioning programming. Led by the popularity of environments such as Scratch (Resnick et al., 2009), Alice (Cooper, Dann, & Pausch, 2000), and Pencil Code (Bau, Bau, Dawson, & Pickens, 2015), a growing number of formal curricula are

Figure 1.1. Comparable blocks-based (A) and text-based (B) programs

The second set of research questions looks at the suitability of these new introductory programming approaches for preparing learners for future computer science learning
opportunities. Research is emerging that suggests that blocks-based programming environments, while successful in changing attitudes and engaging learners, do not adequately prepare them to transition to more conventional programming languages, thus imposing an artificial ceiling on