416 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO.

3, AUGUST 2011

Game-Themed Programming Assignment Modules:

A Pathway for Gradual Integration of Gaming
Context Into Existing Introductory
Programming Courses
Kelvin Sung, Cinnamon Hillyard, Robin Lynn Angotti, Michael W. Panitz, David S. Goldstein, and
John Nordlinger

Abstract—Despite the proven success of using computer video Index Terms—Adaptation, assessment, assignments, CS1/2,
games as a context for teaching introductory programming games.
(CS1/2) courses, barriers including the lack of adoptable mate-
rials, required background expertise (in graphics/games), and
institutional acceptance still prevent interested faculty members I. INTRODUCTION
from experimenting with this approach. Game-themed program-
ming assignment (GTA) modules are designed specifically for
these faculty members. The GTA modules are independent, and
each is a self-contained game-like programming assignment that
T EACHING computer science (CS) concepts based on pro-
gramming interactive graphical games motivates and en-
gages students while accomplishing desired student learning
challenges students on concepts pertaining to a specific curriculum outcomes [1], [2]. When properly integrated in introductory CS
topic area. A faculty member can selectively pick and choose a
subset of GTA modules to experiment with and gradually adopt courses (CS1/2), these approaches build excitement and enthu-
the materials in his or her own classes. Each GTA module also in- siasm for the discipline and attract a bright new generation of
cludes a step-by-step tutorial guide that supports and encourages students early in their academic careers (e.g., [3] and [4]). How-
interested faculty to develop their own expertise and game-themed ever, as a new approach, interested faculty need support to ex-
materials. This paper begins with a survey of previous results. plore and experiment with teaching CS1/2 courses based on in-
Based on this survey, the paper summarizes the important con-
siderations when designing materials for selective adoption. The teractive graphical games. When designing support for these
paper then describes the design, implementation, and assessment faculty members, there are two important areas of considera-
of the GTA modules. The results from ongoing GTA workshops for tion: faculty background and institutional oversight.
CS1/2 faculty members and from a yearlong project in adopting As discussed in the next section, most of the existing re-
the GTA modules in classes are then presented. In this case, sults in integrating computer gaming in CS courses involve ex-
the collected results verified that introductory programming
concepts can be examined, practiced, and learned by means of ploratory projects by faculty members with expertise in com-
GTA modules when neither the faculty nor the students involved puter graphics and gaming. With few exceptions, these projects
have backgrounds in graphics or games. More importantly, these are often student-centric, where the main goals of study are stu-
results demonstrated that it is straightforward to blend the GTA dent engagement and various student learning outcomes. Adapt-
modules into existing classes with minimum alterations. In these ability and general applicability of the resulting materials are
ways, the GTA modules are excellent catalysts enabling faculty to
begin exploring and developing their own expertise and materials usually not main concerns. For faculty members without com-
to teach with games. puter graphics or gaming backgrounds, it can be especially chal-
lenging to take advantage of these results.
When considering experimentation with CS1/2 courses, it is
Manuscript received November 02, 2009; revised May 03, 2010; accepted important to be mindful of institutional oversight procedures.
July 12, 2010. Date of publication August 30, 2010; date of current version Though becoming less controversial in recent years, many CS
August 03, 2011. This work is supported in part by Microsoft External Research educators continue to be unsure about integrating gaming in
under the Computer Gaming Curriculum in Computer Science RFP, Awards
15871 and 16531. formal educational settings (e.g., [5] and [6]). It is a challenge
K. Sung is with the Department of Computing and Software Systems, Univer- for departmental committees to arrive at a consensus for signifi-
sity of Washington Bothell, Bothell, WA 98011 USA (e-mail: [email protected]). cant modifications to CS1/2 courses, especially if the modifica-
C. Hillyard is with the Department of Interdisciplinary Arts and Sci-
ences, University of Washington Bothell, Bothell, WA 98011 USA (e-mail: tions involve computer games.
[email protected]). The CS1/2 game-themed programming assignment (GTA)
R. L. Angotti is with the Department of Education, University of Washington modules are targeted specifically for adoption in existing intro-
Bothell, Bothell, WA 98011 USA (e-mail: [email protected]).
M. W. Panitz is with the Software Programming Department, Cascadia Com- ductory programming classes.1 These assignment modules are
munity College, Bothell, WA 98011 USA (e-mail: [email protected]). self-contained so that faculty with no background in graphics/
D. S. Goldstein is with the Teaching and Learning Center, University of Wash- gaming can select a subset of the modules to combine with ex-
ington Bothell, Bothell, WA 98011 USA (e-mail: [email protected]).
J. Nordlinger is with Microsoft External Research, Redmond, WA 98052 isting assignments in current classes. The assignment modules
USA (e-mail: [email protected]).
1All modules and related materials are freely available from the GTA project
Digital Object Identifier 10.1109/TE.2010.2064315
Web site: http://depts.washington.edu/cmmr/Research/XNA_Games.

0018-9359/$26.00 © 2010 IEEE

SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 417

are limited in curriculum scope to facilitate selective experi- There are many types of “games” that are suitable for
mentation by individuals. Finally, the assignment modules in- teaching CS subjects, including many noncomputer games
clude detailed implementation tutorials to assist interested fac- [16] or games that are based on dedicated devices (e.g., LEGO
ulty in developing game-themed programming assignments. In robots [17]). The focus of this work is on interactive graphical
this way, as the GTA modules are being adopted, faculty develop computer games. As discussed by Sung [18], recent work in this
expertise and can collect and demonstrate results to assist the area can be classified into game development [19], [20] where
decision-making process of institutional oversight committees. students learn about building games, game programming [1]
This paper presents the results from the entire GTA project: where students study algorithms related to games, and game
the design and implementation of the modules [7], the assess- client [21], [22] where students learn about CS concepts via
ment of the academic merits of the materials [8], [9], and finally games. Integrating games into CS1/2 classes belongs to the
the encouraging results from the ongoing GTA workshops game client category because the objective for students in these
for CS1/2 faculty and a yearlong study of classroom adop- classes is to understand abstract programming concepts and not
tions [10]. In the next section, the paper begins with a survey of to learn about building games.
related work done in the area. It is important to note that there Existing work on presenting CS1/2 concepts in the context
is nothing magical about teaching with games. As highlighted of computer games can be broadly categorized into three
by Bayliss [11], faculty buy-in and experience are some of the approaches [18]: 1) little or no games programming [23], [24]
most important factors in realizing the potential student en- where students learn by playing custom games; 2) per-as-
gagement of a game-themed teaching approach. GTA modules signment games development [3], [7], [25] where individual
were designed to help faculty members develop expertise in the programming assignments are computer games designed
area. The objective of this pilot study was to verify that GTA around technical topics being studied; 3) extensive game devel-
modules could be adopted with minimal extra effort by faculty opment where faculty and students work with custom games
members with no background in graphics and games and with engines [26], [27], specialized programming language [28],
little change to a course syllabus. The primary goal was to environments [29], or specific curricula [4], and so on. All
verify that the GTA modules “do no harm” to student learning three approaches reported resounding success with drastically
while faculty members incrementally experiment with, and increased enrollments and student success [3], [4], [28]. Based
develop experience and expertise in, game-themed context. on these results, it is well recognized that integrating computer
The GTA modules are simple “real-time interactive graphics gaming into introductory computer science (CS1/2) courses
programs.” Strictly speaking, these programs do not qualify as is a promising strategy for recruiting and retaining potential
“games” because they have unknown entertainment value. How- students.
ever, in the current implementation, since the programs run on As discussed by Levy and Ben-Ari [30] and Ni [31], issues
both PCs and the XBox 360 gaming platform, the term “game- that faculty consider when examining new and innovative
theme” is used. In this paper, “console-based” or “console as- teaching materials for adoption include preparation time,
signments” are used to refer to conventional programming as- material contents, departmental oversight committee, and
signments that are designed around keyboard and (ASCII) char- compatibility of programming languages. Adopting/adapting
acter-driven console monitors. results from an extensive games development approach requires
a significant investment of time, which includes the need for
faculty to understand a game engine or significantly rework an
II. BACKGROUND existing curriculum. This work-intensive adoption/adaptation is
not suitable for limited-scope investigation. Projects and results
The GTA modules are designed for students to learn from the per-assignment games development approach are typ-
abstract CS concepts by programming and/or examining ically from faculty members with expertise in graphics/games
“real-world, game-like” applications. Relating abstract princi- and are “student-centric,” where the main goals of study are
ples to real-world experience has become increasingly promi- student engagement and various learning outcomes. Most
nent in mathematics, science, and technology education. For instructors of CS1/2 courses do not have the time or expertise
example, the “Calculus Reform” movement of the 1990s [12] to adapt and/or implement these projects in their courses.
included both pedagogical changes and foci on real-world The GTA modules are “student-centric” because they are as-
problems, while the Carl Wieman Science Education Initiative signments that allow students to practice CS concepts in context.
at the University of British Columbia, Vancouver, Canada, has More importantly, these modules are “faculty-centric” because
redesigned its freshmen introductory physics course to present they are the stepping stones for faculty to begin experimenting
standard introductory materials in connection with real-world with a promising new approach to teaching CS1/2 courses.
situations [13]. In the CS education arena, the Media Computa-
tion of the Georgia Institute of Technology, Atlanta, [14] is an
excellent example where foundational programming concepts III. IMPLEMENTATION CONSIDERATIONS AND DETAILS
are presented in the context of popular digital multimedia
applications. This Contextualized Computing Education [15] The above survey implies that in order to facilitate selective
is an ongoing effort, and “interactive games,” being one of the adoption and limited curriculum scope (e.g., per-assignment)
most familiar application areas for students, is a context favored experimentations by faculty members with no relevant back-
by many CS educators. ground, the GTA modules must include all relevant materials
418 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO. 3, AUGUST 2011

and be self-contained. The assignments must be simple interac-

tive graphical applications that assist students to practice rele-
vant programming concepts. At the same time, the GTA mod-
ules should be interchangeable with those from typical CS1/2
courses.

A. Choice of Technical Topic Areas

It is important to differentiate technical topic areas (e.g.,
linked lists) from individual assignments. For example, one
can design a console-based assignment to manipulate a linked
list of numbers, or one can design a game-themed assignment
Fig. 1. Simple Pong game.
where the in-game logic is based on linked lists. With careful
design, both assignments would challenge students in imple-
menting the basic linked list functionality. In this way, the
two assignments can be technically equivalent, and yet one • a sample solution for both the console-based and game-
is a traditional console-based assignment while the other is a theme versions;
game-themed assignment. • a sample grading rubric for each version;
The topics for GTA modules were chosen using a “reverse • a list of frequently asked questions;
adoption” strategy; the technical topics for the game-themed • an implementation tutorial.
assignments were adopted based on the console-based assign- The implementation tutorial is a step-by-step guide that explains
ments in existing CS1/2 courses [32]. There are several advan- the implementation of the game-themed assignment. This tuto-
tages to this approach. rial is intended to help interested faculty better understand how
1) Existing CS1/2 courses are well established with many to create their own game-themed assignment using the library
successful alumni in advanced CS courses and in industry. that was developed to support this project.
This success justifies the selected technical topic areas. For the students, each module includes:
2) Assignments with identical technical topic areas imply • a description of the assignment;
they can be interchanged. This offers a vehicle for the • a skeleton starter project for both the console-based and
subsequent phase of this project, where corresponding game-themed versions.
assignments can be replaced and the effects studied. The game-themed starter project is a game-like application
3) The console-based assignments are included as part of the where all necessary graphics and user interactions functionality
assignment modules. In this way, each assignment module are provided. Students work with the starter project to fill in
addresses a well-defined technical topic area and has two the relevant core CS concepts to complete each assignment,
versions: a console-based version and a game-themed ver- without having to know anything about computer graphics and
sion. The console-based version of the assignment is con- games.
ventional and does not necessarily include interactivity.
This version serves as an excellent and familiar reference C. Implementation Platform
for faculty members unfamiliar with game programming. There has been work done to integrate the concepts and
Seven GTA modules have been implemented. In the subsequent tools involved in building interactive graphical computer games
phase of this study, these modules replaced the corresponding into introductory programming courses, including event han-
console assignments in existing CS1/2 courses. As will be dling [33], graphical user interfaces (GUIs) [34], and graphical
detailed in next section, the current seven GTA modules cover application programming interfaces (APIs) [35]. To support
topic areas that include integer division and the modulus op- faculty without a computer graphics and games background,
erator, random number generation, single-dimensional arrays these aspects of game programming were hidden. This provided
of object references, 2-D arrays, class hierarchy/inheritance, a platform that transparently integrated all of the above tools
linked lists and queues, and binary search trees. Section IV so that faculty and students did not need to be aware of their
describes the assessment procedure that ensures that the con- existence.
sole-based and game-themed versions of the assignment are The C# programming language and the Microsoft XNA
technically equivalent. framework [36] were chosen for this platform. This choice
was governed primarily by the fact that the C#, XNA, and
B. Contents of an Assignment Module Microsoft’s Game Studio Express combination is the only
freely available solution that provides seamless integration of
Each assignment module is designed to be self-contained and
the development environment, programming language, GUI
consists of materials for both the faculty and the students.
API, and graphics API.
For the faculty, each module includes:
• a summary page describing the assignment, including pre-
D. Simple Game-Themed Example: A Pong Game
requisite knowledge, and a list of expected student learning
outcomes; This section uses a simple “Pong game” example to illustrate
• a sample pre- and post-test; game-themed application development. In this application, the
SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 419

Fig. 2. Simple Pong game with XGCS1 library.

Ball travels with a random velocity and bounces within the ap- C3 can be used as interesting examples for demonstrating dif-
plication bounds, and the players control the vertical positions ferent constructs of conditional statements. Since this applica-
of the LeftPaddle and RightPaddle to collide with and bounce tion is game-themed, it is straightforward for students to change
the Ball (see Fig. 1). Fig. 2 is the implementation for this simple the conditional constructs (e.g., replace the switch statement
application based on XGCS1, a simple 2-D library designed with an if–then–else) and then interactively examine the effects
specifically to support GTA modules. of their changes. In this simple example, a GTA module can be
Label A in Fig. 2 explains that all XGCS1 applications must defined around conditional constructs and, for example, chal-
be subclasses of the XNACS1Base class. It is this class that ab- lenge students to support the implementation of barrier blocks
stracts the Model-View-Controller architecture into the Initial- between the two paddles or determine winning conditions.
izeWorld() and UpdateWorld() two-function protocol. The Ini-
tializeWorld() function at Label B is called once at the begin-
ning of the application and, in this case, defines the Pong game IV. RESULTS: THE GTA MODULES
dimension and instantiates the ball and paddles defined at A1 The GTA modules are simple “interactive graphics applica-
and A2. The XGCS1 library automatically draws all geometries tions” where the main goal of the assignments is to reinforce
and moves the Ball object in the application window. The Up-
technical concepts rather than for students to enjoy the fun-ness
dateWorld() function at Label C is called periodically at a rate
of the game. This section describes the developed modules fo-
of 40 times a second. In this case, each update allows the user
cusing on the technical topic areas covered by each of the con-
to control the -positions of the paddles (at C1), bound the ball
sole-based and game-themed versions of the assignments. All
within the application window (at C2), and bounce the ball off
materials presented are available online at [37].
each paddle (at C3). This simple game-themed example demon-
strates the following with the XGCS1 support. Assignment One: Integer Arithmetic. This assignment is de-
• Interactive, graphical, game-themed application can be signed to be the first CS1 assignment. Fig. 3 shows the game-
simple and intuitive as in, for example, the parallel be- themed version of the assignment. In this case, the user can con-
tween the descriptive narration of the application and the trol the horizontal position of the chameleon-circle. The color
actual logic control code at label positions C1, C2, and C3. bands in the background are vertical rectangles with repeating
• The implementation is independent of details of drawing colors. Given this skeleton application, students must program
in computer graphics and complex object interactions in proper integer arithmetic to control the color of the chameleon-
game development as in, for example, the absence of pro- circle, such that as it is moved horizontally, its color always re-
gramming code to draw, manipulate the Ball movements, flects that of the rectangle underneath it. The console-based ver-
and computing spatial collisions between geometric sion of the assignment is a simple character-based flashcard quiz
objects. program.
The fundamental programming logic flow is prominent in the Assignment Two: Random Number Generation and Opera-
implementation. In Fig. 2, for example, the statements at C2 and tors. Fig. 4 shows the game-themed version, where the user
420 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO. 3, AUGUST 2011

Fig. 3. Traveling chameleon. Fig. 5. Snakes and Ladders game.

Fig. 4. Insects Garden game. Fig. 6. Othello game.

controls the hero insect catcher to net randomly generated in-

sects. Supplied with all the graphics and interaction function-
ality, students must implement random number generation and
maintain proper accumulated results, success ratio, etc. The con-
sole-based version of this assignment is on Monte Carlo integra-
tion, where students must approximate the area of a circle based
on randomly generated sample positions.
Assignment Three: Single-Dimension Arrays of Object Ref-
erences. Fig. 5 shows the game-themed version, a variation of
the classic Snakes and Ladders game. In this case, the hero can
pick up gold nuggets, and the user can dynamically create addi-
tional snakes and ladders at run-time. The game board is imple-
mented as a single-dimensional array of game cells, where each Fig. 7. Burst a Bubble game.
game cell can either be empty (i.e., null) or contain a nugget/
snake/ladder. Students’ code must properly access and allocate
new game cells for this single-dimensional array to support the game board is provided. The students’ code must work with
above functionality. In the console-based version, students com- this array to enforce game play logic where the players are
plete a program that maintains a partially filled periodic table of only allowed to place new game pieces in valid locations, and
elements. The periodic table is implemented as a single-dimen- the color of relevant game pieces on the game board must be
sional array that contains either null or a reference to an ele- flipped after each successful play. This is the only assignment
ment object. The student is responsible for filling in code that where the game-themed and console-based assignments are
creates new elements, edits existing elements, or prints out ele- basically identical. The only difference in the two is that, in the
ment objects. console-based version, ASCII characters draw the game board
Assignment Four: Two-Dimensional Arrays. Fig. 6 shows on a character-display window.
the game-themed version, the classic two-person Othello game. Assignment Five: Class Hierarchy and Inheritance. Fig. 7
In this case, an empty two-dimensional array representing the shows the game-themed version where the user launches rocks
SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 421

Fig. 8. Catch a Toy game. Fig. 9. Alphabet Hunt game.

console-based version of the assignment is a character-driven

and fireworks from the lower left corner to burst bubbles that BST implementation test program where outputs are printed
are randomly distributed in the application window. In this as- based on simple commands (e.g., add, find).
signment, students must understand the given Projectile class
and create a subclass to implement a Firework class. The con- V. ASSESSING THE GTA MODULES
sole-based version is based on the same idea, except that the
The GTA modules are designed for students to practice and
program is limited to turn-by-turn processing, and feedback to
learn fundamental concepts in programming. It is important to
the user is in the form of ASCII text. Both the console-based and
evaluate independently the academic content of the materials.
game-themed versions of this assignment expose the students to Additionally, it is important to verify the technical equivalence
the challenges of adding a new class to an existing, nontrivial between the console-based and game-themed assignments in
code base. each module.
Assignment Six: Linked Lists and Queues. Fig. 8 shows the
game-themed version in which the player can insert either A. Independent Reviewer
high-priority toys (animals) or low-priority toys (tools) into
Prof. Ruth Anderson is the independent external reviewer.
the overall queue located at the top of the window. The game She is an experienced instructor who has taught CS1/2 courses
continuously dequeues and drops the oldest toy from the pri- many times at multiple institutions and in a variety of program-
ority queue. The user moves the box to catch the dropping toy. ming languages, has won multiple teaching awards,2 and is ac-
Students must implement the overall queue class, typically by tive in CS education research (e.g., [38] and [39]). In addi-
maintaining items in one of two different (linked list-based) tion to these excellent credentials, Prof. Anderson is perfectly
queue objects, based on each toy’s priority. The game interfaces suited for evaluating the materials because she has never taught
with the overall queue class and draws all the items in a single a graphics or gaming course and has limited experience with
row so it appears that the toys are all in a single queue, regard- GUI programming. Before this project, Prof. Anderson did not
less of student implementation. The console-based version of know anyone on the project team.
this assignment is a text-based “help desk” application. The
user can enter high- or low-priority requests to be enqueued B. Procedure
to the front/back of the queue, respectively. Similar to the During the project, in-person or verbal communications
game-themed version, the retrieval of requests is a simple were avoided both to maintain impartiality and to simulate the
dequeue operation on the queues. The skeleton starter projects sort of investigations that might be made by curious faculty.
for both versions contain all necessary I/O functionality: The assessment of the assignments was conducted across the
graphics/GUI for the game-themed version, and character I/O project Web site [37], where newly released materials were
for the console-based version. In both cases, students only need downloaded, examined, and tested by Prof. Anderson. Feed-
to implement the linked list queue and the priority queue. back was provided via a custom assignment evaluation form.
Assignment Seven: Binary Search Trees (BSTs). Fig. 9 The assignment evaluation form is designed to collect both
shows the game-themed version where, in a side-scrolling formative feedback and quantitative scores [40]. Each assign-
game, the alphabet hero must leap to collect flying alphabet ment module is assessed in two areas:
targets and process the walking alphabet search requests. The 1) Quality of the assignment: assesses the merit, the technical
students’ code must implement a BST to store the collected equivalence between the console-based and game-themed
target letters and search the BST upon encountering walking assignments, and the supporting materials (e.g., pre/post-
requests. The drawing of the BST in the upper left is supported test);
via an abstract base class in the form of a dynamically linked 2ACM Faculty Award, voted best teacher by Department of Computer Science
library (DLL) where the implementation details are hidden. The students, University of Virginia, Charlottesville, 2004.
422 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO. 3, AUGUST 2011

Fig. 10. Games by faculty with no prior background.

2) Potential for adoption: assesses the factors independent of [44].3 These workshops guided faculty to develop game-themed
the quality of assignments that may prevent adoption (e.g., applications based on tutorials from the GTA modules. The
programming language used). images in Fig. 10 are screen shots of a “Worm-like” and a
“Pizza Delivery” game resulting from the second day of a mul-
C. Assessment Results tiday version of the workshops [43], [44]. Although relatively
simple, these games were designed and developed in a matter
Based on the review feedback, the assignments have been
of hours by CS1/2 faculty members with no prior background
well received overall. It is believed that because the assignments
in graphics/games.
were adopted based on existing CS1/2 classes, technical merits
The workshops have received overwhelmingly positive feed-
were never an issue. Prof. Anderson agrees that the assignments
back from the participants. For example, written feedback on
are appropriate for typical CS1/2 courses. Assignments with low
“the appropriateness of material difficulty” included: “I found
quality of assignment scores were revised and reassessed. This
the materials challenging but able to comprehend,” and “good
process continued until the formative comments were positive
balance of complexity and new materials.” The written feed-
and the numeric scores were above 4 (out of 5).
back on “the presented materials will help me develop game-
As expected, all of the low scores were caused by game-
themed applications” included: “For sure!” and “In this envi-
themed assignments. This was typically due to the following.
ronment [GTA], YES!” The results and feedback from these
1) Differences in difficulty: Initial attempts at game-themed
workshops showed that although they found the GTA materials
assignments often resulted in highly difficult, complex, or
to be nontrivial, faculty participants with no prior expertise in
intimidating programs. Based on feedback received, as-
graphics/games were able to comprehend and begin developing
signments have been adjusted accordingly.
game-like applications within a matter of hours.
2) Inappropriate use of concept: Designing assignments
around negative results from game play is a bad idea. For
example, a linked list structure might be used for tracking VII. CLASSROOM ASSESSMENT TOOLS AND PROCEDURES
when the player has been unsuccessful at a given task. As catalysts for faculty development, it is important to verify
In this case, in order to test the linked list, students must that the GTA modules “do no harm” to student learning. It is es-
deliberately be unsuccessful at playing the game. This can sential for faculty members to have the reassurance of consistent
take the fun out of the assignment. student learning outcomes when experimenting with this poten-
3) Deficiency in support: The development team often over- tially powerful approach. Instruments for understanding the ef-
looked important details. For example, in the beginning, a fectiveness of new teaching materials include analyzing quan-
specialized hardware controller was the only way to control titative exam scores, qualitative evaluation of projects, student
a game. Based on the feedback, this problem was remedied opinion polls, and success rates [45], [46]. These instruments
with a keyboard-based software object simulation. are employed to assess student learning outcomes and percep-
Because Java is the language of choice at Prof. Anderson’s insti- tions when the GTA modules are adopted in existing classes.
tution, consistently lower scores for potential for adoption were Based on the positive feedback and promising outcomes from
received. The developers are fully aware that the language issue the GTA workshops, in-classroom adoption of GTA modules
must be addressed for wide adoption of results. With experience began. Because the modules are designed as catalysts for fac-
building these assignments, and understanding the important at- ulty development, faculty members needed to have the reas-
tributes of the library, the developers are investigating possibil- surance of consistent student learning outcomes when exper-
ities for porting the results to other environments. imenting this potentially powerful approach to teaching. This
project utilized exam scores, qualitative evaluation of projects,
VI. GTA WORKSHOPS FOR FACULTY a student attitude survey, and success rates [45], [46] to assess
student learning outcomes and perceptions when the GTA mod-
Outside of classrooms, GTA-related workshops have been
ules were adopted in existing classes.
offered for interested faculty members at regional [41] and na-
tional [42] conferences and at institutions internationally [43], 3Refer to http://faculty.washington.edu/ksung for workshop lecture notes.
SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 423

TABLE I
SUCCESS RATES

For long-term effects, new teaching materials are tracked

over multiple semesters of the same course via course enroll-
ments [47] or continual student successes [3]. In the case of
GTA modules, one goal was to demonstrate the feasibility of
simple replacement of selective assignments in existing classes.
The same classes were followed over different semesters
where different GTA modules were adopted. The assess-
ment procedure was designed around existing console-based
CS1/2 classes.4 These were well-established courses producing
many successful alumni in advanced CS courses and in the
industry [32].
This study was designed around the adoption of six of Fig. 11. Average assignment results.
the GTA modules: four for CS1, and two for CS2. The
game-themed versions of these assignments were integrated
into existing CS1/2 courses over three academic quarters. Two
existing CS1/2 courses were offered without modification and
served as control groups. In the experimental classes, two of the
four existing console assignments were replaced with GTAs.
Each GTA course consisted of a mixture of two existing console
assignments and two GTA modules, in combinations that varied
from class to class. This verified that a faculty member could
select and replace some or all existing assignments with GTAs.
To minimize variation between students in the classes, the
modifications to the courses were not advertised, so students
did not know they were registering for a class that would use
the GTAs. Fig. 12. Self-reported time spent.

VIII. CLASSROOM ADOPTION RESULTS AND ANALYSIS console assignment. The scores from GTA were consistently
Throughout the experiment, the instructor avoided any extra better than the corresponding console version. From the written
effort when adopting the GTA modules. For example, no lec- feedback (detailed later), it was clear that students spent more
ture time was spent covering graphics or game aspects of the time playing with the GTA assignments, which resulted in their
GTAs. In all cases, the exact same assessment instruments were having a smaller number of errors in their final submissions.
administered under similar conditions for both GTA and con- In addition, there were interesting observations of trend in the
sole courses. score differences. At the very beginning of CS1 when the as-
1) Success Rates: The success rates of these CS1/2 classes signment was trivial (CS1-A, simple arithmetic), the difference
have fluctuated between 65% to 85% historically. With this per- between GTA and console was small. By the end of CS1, the
spective in mind, analysis began by examining the overall suc- assignments became more challenging (CS1-D, 2-D array),
cess rates of all the classes. The percentage passing in each and the difference in scores also became more prominent
class is given in Table I. The percentages of GTA classes were (89% versus 58%). As is widely recognized [48], interactive
higher than the console classes. Caution should be used when graphical feedback encourages experimentation, and this trend
examining these figures since they are well within the historic in assignment score differences reflected students’ further en-
ranges. The analysis includes all of the students in the partici- gagement with the assignment. However, this increase in score
pating classes. Because of the small number of female students, differences became less prominent by CS2 assignments (88%
to ensure strictest anonymity, the results were not disaggregated versus 81% for CS2E, and 82% versus 77% for CS2F). As
by gender. observed by Guzdial [15], CS2 students are more experienced
2) Assignment Scores: Fig. 11 plots the average scores for and often prefer assignments without elaborate setups. Results
all of the six assignments.5 The left bar above each assignment verified that as students became more proficient, the advantage
displays the results from GTA, while the right is from the of interactive graphical feedback diminished.
4The
3) Self-Reported Time Spent on Assignments: Fig. 12 shows
involved instructor has no background in graphics/games.
5Note
students’ self-reported time spent on the assignments. It was in-
that each assignment has GTA and console versions. For example, the
GTA version of CS1-A was offered during Spring, while the corresponding con- teresting to note when examining the time spent on the four CS1
sole version was offered in the control course, Winter. assignments that students reported more than twice the amount
424 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO. 3, AUGUST 2011

Fig. 13. Post-assignment surveys: (Left) CS1 and (Right) CS2.

Fig. 14. Pre- and post-course surveys: (Left) CS1 and (Right) CS2.

of time spent on CS1-B, -C, and -D console assignments. From given GTA starter project. However, in all cases, once they un-
the average scores and success rates, it was clear that students derstood the given system, students reflected that completing the
were learning comparable material. These large discrepancies assignment was not “as complicated as it first appears.” In CS2
seemed to suggest game-themed assignments were much more assignments, many students expressed frustration at not being
efficient learning tools. While it may be true that visual feed- able to improve on the assignments (e.g., “it was interesting to
back is an important learning tool, it was not found that it alone have an end result you can play with, I wish I was able to im-
could accomplish these results. From the written feedback (e.g., prove the boring game,” or “I did learn a lot about BST [Binary
“counting only the time I actually ‘worked’ on and not ‘played’ Search Tree] doing this assignment. However the ‘game’ parts
with it”), it appeared that some students discounted the time they were useless as we don’t actually get to do gaming.”).
spent playing with the sample solution and starter project. Re- The GTA modules were designed to be expertise-neutral to
searchers believe the correct numbers for GTA should be closer prevent superfluous graphics user interaction programming and
to those from console assignments. In the case of CS2, the large inappropriate gaming contents (such as violence) [5], [6]. De-
time difference between the first GTA (CS2-E) reflected the fact velopers made a conscious decision to restrict students’ access
that the assignment was more complicated and demanded time to such functionality [7]. This feedback showed that restricting
for familiarization. The similar amount of time spent for the students’ access in such a way was successful. However, the
second CS2 assignment (CS2-F) showed that students were able feedback also implied that the modules had literally taken the
to take advantage of their initial time investment. “fun” out of game programming. The delicate balance between
4) Per-Assignment Survey: After students handed in their as- allowing creativity and discouraging excessive graphics pro-
signments and before they received their grades, the exact same gramming is currently being investigated. For example, the fun
questions were asked of all students: to assess the clarity and element in the developed games may be incorporated if scoring
difficulty of the assignment and the amount they had learned and rewards are included for every right concept/code.
from the assignment, if completing the assignment made them 5) Pre- and Post-Course Survey: Course survey forms were
feel satisfied, and if the assignment was interesting. designed to elicit students’ background, self-perception, inter-
Fig. 13 plots the average of all results from GTA and console ests, and general attitudes toward the CS discipline. Students
courses for all the CS1 (left) and CS2 (right) courses. Results on completed these forms during first day of class (precourse) and
the right of Fig. 13 showed CS2 students found the console as- before the final exams (post-course). The tables of Fig. 14 show
signments easier to understand, slightly more challenging, and, the averages of all students from the GTA and Console classes
contrary to intuition, they found that working on console as- separately: The table on the left shows CS1, and the table on the
signments was more satisfying and that the console assignments right shows CS2 classes.6
were more interesting. It is interesting that replacement of some assignments in the
Fortunately, written feedback from students helped explain featured CS1/2 classes was almost transparent. Because stu-
these observations. In both CS1 and CS2 GTA courses, students dent enrollment in particular sections of the class cannot be con-
complained that they had to spend extra time understanding the 6Students had the option to not participate in these surveys.
SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 425

trolled, some differences in items across classes at the onset are for teaching individual programming concepts (e.g., linked
expected. However, to assess the impact of the curriculum, the lists or arrays) have been developed [49]. Interested faculty
change in attitudes from before and after each class section was members will be able to experiment with selected GTA and
measured and analyzed for statistically significant changes over GTI modules in their existing classes, become comfortable with
time. Except for one case, the measured change from pre- to the game-themed teaching approach, consult the tutorials pro-
post- in the items in tables of Fig. 14 was not statistically signif- vided with GTA and GTI modules, and begin to develop their
icant. This exception was the “Well prepared” question for CS1 own game-themed materials. Ultimately, the success of any
students (left-hand table of Fig. 14). In this case, when students new approach to teaching hinges on the instructor’s expertise
were asked after the class was over, “In hindsight, how well did and enthusiasm. The true potential for engaging and exciting
you think you were prepared for the class?” there was a signif- students can only be realized when the instructor becomes
icant improvement in self-confidence for GTA students. These proficient in, feels ownership of, and develops his or her own
data, in combination with the better performance results from game-themed instructional materials based on the needs and
GTA (e.g., CS1-C and -D in Fig. 11) and the little difference strengths of their students.
in attitude seen from the left of Fig. 13, implied an interesting Currently, a multilingual and API-independent platform to
and potentially important observation. CS1 students performed support GTA/GTI modules in multiple programming languages
better because the interactive graphical application supported and APIs is being designed. In addition, results of this work
experimentation and visualization. Since the applications were are being disseminated with colleagues from community col-
not really fun or flashy, they did not find the assignments es- leges and high schools, and workshops at national conferences
pecially interesting. However, after the class, they did become (e.g., [50]) and at international institutions are continuing to be
more confident about their abilities. Together, these data sug- offered (e.g., [51]).
gested that targeted “uninteresting” interactive graphical assign-
ments can be a good tool for teaching CS1 students. ACKNOWLEDGMENT
6) Feedback From Faculty: After grading each assignment, The authors would like to thank all BIT 142 and 143 stu-
the instructor filled out a feedback form detailing his efforts spe- dents from Cascadia Community College for working with the
cific to the assignments (e.g., lecture time, answer questions, chaotic schedules and draft versions of the assignments, and
grading time) and impressions of student learning. As men- L. Dirks at Microsoft Research for his ongoing support of this
tioned, the instructor did not have prior background in com- work. The reviewers’ excellent attention to details and com-
puter graphics/games and purposely avoided allocating specific ments have greatly improved the quality of this paper. All opin-
lecture time and extra help for GTAs. As a direct result, his ions, findings, conclusions, and recommendations in this work
feedback showed no significant difference in efforts between are those of the authors and do not necessarily reflect the views
the GTA and console assignments. However, he reported stu- of Microsoft.
dents’ verbal comments on GTAs to be that they were more
work and more difficult.7 In the end, because the GTAs were REFERENCES
“dropped” into the classes without any dedicated lecture time, [1] E. Sweedyk, M. deLaet, M. C. Slattery, and J. Kuffner, “Computer
the assignments had minimal effect on the class as a whole. It games and CS education: Why and how,” in Proc. SIGCSE, 2005, pp.
256–257.
is encouraging that as the instructor became more comfortable [2] U. Wolz, T. Barnes, I. Parberry, and M. Wick, “Digital gaming as a
with GTAs, he did begin experimenting with graphics/game pro- vehicle for learning,” in Proc. SIGCSE, 2006, pp. 394–395.
gramming and developed a simple card-matching game based [3] J. D. Bayliss, “The effects of games in CS1-3,” J. Game Dev., vol. 2,
no. 2, pp. 7–18, 2007.
on the provided tutorials. Currently, the instructor is experi- [4] S. Leutenegger and J. Edgington, “A games first approach to teaching
menting with incorporating game-themed instructional modules introductory programming,” in Proc. SIGCSE, 2007, pp. 115–118.
in his CS1 classes [49]. [5] H. M. Walker, “Do computer games have a role in the computing class-
room?,” SIGCSE Bull., vol. 35, no. 4, pp. 18–20, 2003.
[6] S. Haller, B. Ladd, S. Leutenegger, J. Nordlinger, J. Paul, H. Walker,
IX. CONCLUSION and C. Zander, “Games: Good/evil,” in Proc. SIGCSE, 2008, pp.
The resulting games and survey feedback from GTA work- 219–220.
[7] K. Sung, M. Panitz, S. Wallace, R. Anderson, and J. Nordlinger,
shop participants indicated that it is straightforward for faculty “Game-themed programming assignments: The faculty perspective,”
without graphics/games background to understand and begin in Proc. SIGCSE, 2008, pp. 300–304.
[8] K. Sung, R. Rosenberg, M. Panitz, and R. Anderson, “Assessing game-
working with GTA modules and to develop their own game-like themed programming assignments for cs1/2 courses,” in Proc. GDCSE,
applications. This classroom case study demonstrated that it 2008, pp. 51–55.
is possible for a CS faculty member with no background in [9] K. Sung, M. Panitz, R. Reed-Rosenberg, and R. Anderson, “CS1/2
game-themed programming assignments for faculty,” J. Game Dev.,
graphics/games to integrate GTA modules in an existing course vol. 3, pp. 27–47, Mar. 2008.
without adverse effects on student learning. These results [10] C. Hillyard, R. Angotti, M. Panitz, K. Sung, J. Nordlinger, and D. Gold-
are exciting because they mean that interested faculty can stein, “Game-themed programming assignments for faculty: A case
study,” in Proc. SIGCSE, 2010, pp. 270–274.
confidently begin limited curriculum scope experimentation [11] J. D. Bayliss, “Using games in introductory courses: Tips from the
with selected GTA modules in their own courses. To further trenches,” in Proc. SIGCSE, 2009, pp. 337–341.
[12] D. A. Smith and L. C. Moore, Calculus: Modeling and Applications,
support these faculty members, limited curriculum scope, 2nd ed. Boston, MA: Houghton Mifflin, 2007 [Online]. Available:
self-contained “game-themed instructional” (GTI) modules http://www.math.duke.edu/education/calculustext
[13] “The Carl Wieman Science Education Initiative (CWSEI),” Univ.
7The instructor did not have access to any survey information during the British Columbia, Vancouver, Canada, 2008 [Online]. Available:
classes. http://www.cwsei.ubc.ca
426 IEEE TRANSACTIONS ON EDUCATION, VOL. 54, NO. 3, AUGUST 2011

[14] M. Guzdial and B. Ericson, Introduction to Computing and Program- [42] K. Sung, “XNA game-themed applications for teaching introductory
ming With Java, A Multimedia Approach. Englewood Cliffs, NJ: programming courses,” presented at the Invited Pre-Conference Work-
Prentice-Hall, 2007. shop, 4th Int. Conf. Found. Digital Games, Orlando, FL, Apr. 2009.
[15] M. Guzdial, “Contextualized computing education,” Invited presenta- [43] K. Sung, “XNA game-themed applications for teaching introductory
tion, Microsoft Research Faculty Summit, Jul. 2008 [Online]. Avail- programming courses,” presented at the Invited 3-Day Workshop, Mi-
able: http://home.cc.gatech.edu/guzdial/169 crosoft Mexico Digital Arts Univ., Guadalajara, Mexico, Feb. 2009.
[16] P. Drake, Data Structures and Algorithms in Java. Upper Saddle [44] K. Sung, “Developing game-themed applications for teaching intro-
River, NJ: Prentice-Hall, 2006. ductory programming courses,” presented at the Invited 2-Day Work-
[17] M. McNally, M. Goldweber, B. Fagin, and F. Klassner, “Do LEGO shop, Beijing Univ. Technol., Beijing, China, Jul. 2009.
Mindstorms robots have a future in CS education?,” in Proc. SIGCSE, [45] J. Cromack and W. Savenye, “Learning about learning in com-
2006, pp. 61–62. putational science and science, technology, engineering and
[18] K. Sung, “Computer games and traditional computer science courses,” mathematics (STEM) education,” 2007 [Online]. Available:
Commun. ACM, vol. 52, no. 12, pp. 74–78, Dec. 2009. http://research.microsoft.com/ur/us/AssessmentToolkit/
[19] D. Frost, “Ucigame, a Java library for games,” in Proc. SIGCSE, 2008, [46] M. Eagle and T. Barnes, “Experimental evaluation of an educational
pp. 310–314. game for improved learning in introductory computing,” in Proc.
[20] J. Linhoff and A. Settle, “Motivating and evaluating game development SIGCSE, 2009, pp. 321–325.
capstone projects,” in Proc. FDG, 2009, pp. 121–128. [47] I. Parberry, T. Roden, and M. B. Kazemzadeh, “Experience with an
[21] K. Sung, P. Shirley, and R. Reed-Rosenberg, “Experiencing aspects of industry-driven capstone course on game programming: Extended ab-
games programming in an introductory computer graphics class,” in stract,” in Proc. SIGCSE, 2005, pp. 91–95.
Proc. SIGCSE, 2007, pp. 249–253. [48] S. Cooper, W. Dann, and R. Pausch, “Teaching objects-first in intro-
[22] S. M. Pulimood and U. Wolz, “Problem solving in community: A nec- ductory computer science,” in Proc. SIGCSE, 2003, pp. 191–195.
essary shift in cs pedagogy,” in Proc. SIGCSE, 2008, pp. 210–214. [49] R. Angotti, C. Hillyard, M. Panitz, K. Sung, and K. Marino,
[23] P. Haden, “The incredible rainbow spitting chicken: Teaching tradi- “Game-themed instructional modules: A video case study,” in Proc.
tional programming skills through games programming,” in Proc. ACE, FDG, 2010, pp. 9–16.
2006, pp. 81–89. [50] M. Panitz, K. Sung, and J. Nordlinger, “Develop game-themed exam-
[24] J. D. Bayliss and D. I. Schwartz, “Instructional design as game design,” ples for CS1/2 without background in graphics or games,” presented at
in Proc. FDG, 2009, pp. 10–17. the SIGCSE, Mar. 2010.
[25] A. Luxton-Reilly and P. Denny, “A simple framework for interactive [51] K. Sung, “Developing game-themed applications with XNA,” pre-
games in cs1,” in Proc. SIGCSE, 2009, pp. 216–220. sented at the Invited 3-Day Workshop, Serious Games Winter School
[26] K. Bierre, P. Ventura, A. Phelps, and C. Egert, “Motivating OOP by 2010, Feb. 2010.
blowing things up: An exercise in cooperation and competition in an
introductory Java programming course,” in Proc. SIGCSE, 2006, pp.
354–358.
[27] M. C. Lewis and B. Massingill, “Graphical game development in cs2:
A flexible infrastructure for a semester long project,” in Proc. SIGCSE,
2006, pp. 505–509.
[28] W. Dann, S. Cooper, and R. Pausch, Learning to Program with Alice. Kelvin Sung received the B.E.E. degree from the University of Wisconsin-
Upper Saddle River, NJ: Prentice-Hall, 2006. Madison in 1986, and the M.S. and Ph.D. degrees in computer science from
[29] M. Külling and P. Henriksen, “Game programming in introductory the University of Illinois at Urbana-Champaign in 1990 and 1992, respectively.
courses with direct state manipulation,” in Proc. ITiCSE, 2005, pp. He was an Assistant Professor with the School of Computing, National Uni-
59–63. versity of Singapore, Singapore, and a Software Architect with Alias—Wave-
[30] R. B.-B. Levy and M. Ben-Ari, “We work so hard and they don’t use front (now part of Autodesk), Toronto, ON, Canada, where he played a key role
it: Acceptance of software tools by teachers,” SIGCSE Bull., vol. 39, in designing and implementing the first version of the Maya Renderer. Cur-
no. 3, pp. 246–250, 2007. rently, he is a faculty member with the Computing and Software Systems De-
[31] L. Ni, “What makes CS teachers change?: Factors influencing CS partment, University of Washington Bothell, Bothell. His research interests in-
teachers’ adoption of curriculum innovations,” in Proc. SIGCSE, 2009, clude high-quality image synthesis, video game development, serious games,
pp. 544–548. and computer science education.
[32] “BIT142/143: Intermediate programming and data structure,” Cas- Prof. Sung is a Member of the Association for Computing Machinery (ACM).
cadia Community College, Bothell, WA, 2008 [Online]. Available:
http://faculty.cascadia.edu/mpanitz/courses/2008Fa/BIT142/ [On-
line]. Available: http://faculty.cascadia.edu/mpanitz/courses/2008Sp/ Cinnamon Hillyard received the Ph.D. degree in mathematics from Utah State
BIT143/ University, Logan, in 1999.
[33] H. B. Christensen and M. E. Caspersen, “Frameworks in CS1: A dif- She is an Assistant Professor with the Interdisciplinary Arts and Sciences Pro-
ferent way of introducing event-driven programming,” in Proc. ITiCSE, gram, University of Washington Bothell, Bothell. She also completed a post-
2002, pp. 75–79. doctorate position in mathematics with the University of Arizona, Tucson. Her
[34] V. K. Proulx, J. Raab, and R. Rasala, “Objects from the begin- scholarship focuses on undergraduate mathematics and science education, es-
ning—With GUIs,” in Proc. ITiCSE, 2002, pp. 65–69. pecially in the assessment of learning outcomes.
[35] S. Matzko and T. A. Davis, “Teaching CS1 with graphics and C,” in Prof. Hillyard currently holds the position of Past-Chair of the Special In-
Proc. ITICSE, 2006, pp. 168–172. terest Group of the Mathematical Association of America on Quantitative Lit-
[36] “XNA Game Studio,” Microsoft, Inc., 2007 [Online]. Available: http://
eracy (SIGMAA-QL) and is the Secretary/Treasurer for the National Numeracy
msdn2.microsoft.com/en-us/directx/Aa937794.aspx
Network (NNN).
[37] “Game-themed introductory programming project home page,”
Univ. Washington, Bothell, 2010 [Online]. Available: http://
depts.washington.edu/cmmr/Research/XNA\_Games
[38] R. Anderson, R. Anderson, K. M. Davis, N. Linnell, C. Prince, and V. Robin Lynn Angotti received the B.S. and M.A. degrees in mathematics from
Razmov, “Supporting active learning and example based instruction East Carolina University, Greenville, NC, in 1988 and 1990, respectively, and
with classroom technology,” SIGCSE Bull., vol. 39, no. 1, pp. 69–73, the Ph.D. degree in mathematics education from North Carolina State Univer-
2007. sity, Raleigh, in 2004.
[39] T. B. Horton, R. E. Anderson, and C. W. Milner, “Work in She taught secondary mathematics at D. H. Conley High School, Greenville,
progress—Reexamining closed laboratories in computer science,” in NC, and remedial mathematics at North Carolina State University. She was the
Proc. 34th Annu. ASEE/IEEE Frontiers Educ. Conf., Oct. 2004, vol. Assistant Director of the Center for Science, Mathematics and Technology Edu-
2, pp. F3C-15–F3C-16. cation and an Assistant Professor with East Carolina University. In 2007, she be-
[40] J. Frechtling and L. Sharp, “User-friendly handbook for mixed method came an Assistant Professor with the University of Washington Bothell, Bothell.
evaluations,” Division of Research, Evaluation and Communication, She has published in Mathematics Teacher, The NCTM Yearbook on Data and
National Science Foundation: Directorate for Education and Human Chance, The AMTE Monograph, and the Statistics Education Research Journal.
Resources, Arlington, VA, 1997. Her fields of interest are mathematics, statistics, and technology education.
[41] M. Panitz and K. Sung, “Incrementally incorporating video games Dr. Angotti is a Member of the North American Chapter of the Psychology
into instruction using XNA game-themed assignments,” in Proc. of Mathematics Education, the National Council of Teachers of Mathematics,
CCSC-NW, Oct. 2008, pp. 40–41. and the Association of Mathematics Teacher Educators.
SUNG et al.: GAME-THEMED PROGRAMMING ASSIGNMENT MODULES 427

Michael W. Panitz received the B.A. and M.Eng. degrees in computer science John Nordlinger is currently pursuing the Master’s degree in film production
from Cornell University, Ithaca, NY, in 1998 and 1999, respectively. at the University of Southern California, Los Angeles.
He has worked with Microsoft in the .Net Common Language Runtime group He joined Microsoft Research, Redmond, WA, in 2001, where he collaborated
and is currently a Senior Founding Faculty Member with Cascadia Community with academic institutions in the northeast United States and India. After con-
College, Bothell, WA. He is currently interested in using innovative technologies vincing Microsoft Research to open a research lab in Bangalore, India, he then
and techniques to teach lower division computer programming and computer focused on mitigating the decline in CS enrollments. Along with participating
science, and he has been interested in reliable distributed systems. on various panels, he coauthored two papers at SIGCSE 2008: one on teaching
Mr. Panitz is a Member of the Association for Computing Machinery (ACM). with XNA GSE, and one on teaching CS with socially relevant themes. He also
co-edited the book World of Warcraft and Philosophy (Open Court, 2009) and
contributed to the tome Ethics and Game Design: Teaching Values Through Play
(Inf. Sci. Reference, 2010), by David Gibson and Karen Schrier. His first film,
David S. Goldstein received the B.A. degree in English from the University of The Allegory of the Game, has been selected for three film festivals.
California, Riverside, in 1984; the M.A. degree in communication from Stanford
University, Stanford, CA, in 1985; the M.A. degree in American civilization
from the University of Pennsylvania, Philadelphia, in 1988; and the Ph.D. degree
in comparative culture from the University of California, Irvine, in 1997.
He edited career-guidance books for engineers with Professional Publica-
tions, Inc., Belmont, CA, and was a Lecturer with the University of California,
Irvine; Shoreline Community College, Shoreline, WA; and the University of
Washington Bothell, Bothell, where he is now a Senior Lecturer and the Di-
rector of the Teaching and Learning Center. He has presented internationally on
the scholarship of teaching and learning and co-edited, with Audrey B. Thacker,
Complicating Constructions: Race, Ethnicity, and Hybridity in American Texts
(Univ. Washington Press, 2007). He serves on the Editorial Board of Ethnic
Studies and peer reviews for Ethnic Studies Review and Multi-Ethnic Literature
of the United States.
Dr. Goldstein is a founding member of the Research Committee of the Asso-
ciation for Authentic, Experiential and Evidence-Based Learning.