Tangibles vs. Mouse in Educational
Programming Games: Influences on
Enjoyment and Self-Beliefs
Computer Science (CS) and related skills such as
programming and Computational Thinking (CT) have
recently become topics of global interest, with a large
number of programming games created to engage and
educate players. However, there has been relatively
limited work to assess 1) the efficacy of such games
with respect to critical educational factors such as
enjoyment and programming self-beliefs; and 2)
whether there are advantages to alternative, physically
embodied design approaches (e.g., tangibles as input).
To better explore the benefits of a tangible approach,
we built and tested two versions of an educational
programming game that were identical in design except
for the form of interaction (tangible programming
blocks vs. mouse input). After testing 34 participants,
results showed that while both game versions were
successful at improving programming self-beliefs, the
tangible version corresponded to higher self-reports of
player enjoyment. Overall, this paper presents a
comparison between the efficacy of tangible and mouse
design approaches for improving key learning factors in
educational programming games.
Physical Embodiment; Embodied Interaction;
Tangibles; Educational Programming Game;
Permission to make digital or hard copies of part or all of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. Copyrights
for third-party components of this work must be honored. For all other
uses, contact the Owner/Author.
Copyright is held by the owner/author(s).
CHI'17 Extended Abstracts, May 06-11, 2017, Denver, CO, USA
Edward F. Melcer
New York University
Brooklyn, NY 11201, USA
University of California, Santa Cruz
Santa Cruz, CA 95064, USA
University of California, Santa Cruz
Santa Cruz, CA 95064, USA
ACM Classification Keywords
H.5.m. Information interfaces and presentation (e.g.,
The application of technology to virtually every field has
made Computer Science (CS) and related skills such as
programming and Computational Thinking (CT) crucial
21st century skills to develop. This has recently become
a topic of global interest with notable initiatives and
calls from the National Science Foundation ,
president of the United States of America , and UK's
Year of Code . To address this need, industry and
academia have created numerous animation
programming environments (e.g., Logo , Scratch
, and Blockly ) as well as educational
programming games (e.g., Mazzy , BOTS [14, 15],
and Machineers ).
However, development of games to teach CS has
ultimately lagged behind the design of educational
games in general , and little is known about
whether CS-focused games actually improve important
educational factors for STEM learning (such as
engagement, enjoyment, and programming self-beliefs
[1, 38]) or if they simply function as chocolate-covered
broccoli. Even the breadth of design choices explored in
existing educational programming games has been
shown to be fairly narrow. For instance, a recent survey
 found that most educational CT/programming
games have similar playable characters (robot), genre
(puzzle), number of players (single player), and form of
interaction (touchpad/mouse and keyboard). All of this
becomes problematic when trying to determine the
validity of such designs for CS education and whether
there are more effective alternative designs.
In contrast to the ubiquity of mouse/keyboard input,
recent research suggests that body-based, physically
embodied designs provide affordances that aid in the
meaning-making process and offer greater learning
benefits than traditional keyboard and mouse input [26,
30, 32, 46]. One common physically embodied design
approach utilized in HCI and Learning Science
communities is the use of tangibles/physical
manipulatives, which have demonstrated many
potential benefits over traditional desktop applications.
The primary advantages of tangibles is that they allow
for learning concepts to be embedded directly into the
physical material and design of an object, as well as
through the embodied interactions learners have by
manipulating these objects [30, 33]. Tangibles have
been found to result in benefits for a wide range of
factors such as engagement , interest , and
collaboration  in non-programming domains.
To compare the efficacy of tangible and mouse design
approaches, we created two versions of an educational
programming game called Bots & (Main)Frames (see
Figure 1) that differed only in the form of interaction
(tangible programming blocks vs. mouse input). This
was done to isolate the impact of tangible and mouse
designs in an educational game when all other
mechanics, aesthetics, etc. are identical. The game
itself was designed to incorporate common design
characteristics of educational programming games (i.e.,
players program a virtual robot to solve puzzles), but
also provides a comparison point between the two
versions for different forms of interaction. The mouse
version of Bots & (Main)Frames mirrors the prototypical
design where players use a mouse to program by
clicking buttons representative of programming
commands (see Figure 2). In lieu of mouse controls,
Figure 1: The Bots & (Main)Frames
game. The UI is identical for both
the mouse version and tangibles
the tangibles version instead utilizes tangible
programming blocks that players physically hook
together to create a program (see Figure 3).
We had a total of 34 novice programmers (i.e., having
less than 6 months of programming exposure) play
both versions of the game and compared outcomes for
crucial educational factors such as enjoyment and
programming self-belief (comprised of programming
self-concept, interest, and aptitude).
Tangibles and Computational Concepts
There has been some work in the tangible and
embodied interaction community on the creation of
tangibles to teach computing concepts such as roBlocks
[36, 37] and Electronic Blocks [44, 45] where learners
connect physical sensor, actuator, and logic blocks to
explore programming and physical computing concepts;
Note Code  where learners program, play, and
connect buttons on noteboxes to play musical
sequences; Thingy Oriented Programming  where
users program wirelessly connected objects by
recording sequences of actions and reactions using
tangible objects; and TanProRobot 2.0  where
children use physical blocks to program a toy car.
Notably, concepts covered by these tools focus more on
physical computing, electronics, and music than actual
programming or games. In a closer vein, Strawbies
 utilizes wooden tiles to program a game character
to solve mazes, but has only been evaluated through
qualitative descriptions of play sessions at local schools.
Physical Embodiment in Educational Games
In a meta-review dissecting embodiment strategies,
Melcer and Isbister  identified five forms of physical
embodiment commonly used in educational games:
Direct Embodied, Enacted, Manipulated, Surrogate, and
Augmented. Notably, they also found that the
Manipulated form of embodiment frequently uses
tangibles to teach computational concepts . We
similarly employed the affordances of Manipulated
embodiment (e.g., applying embodied metaphors and
interactions to physical objects , and utilizing
physical form to better represent learning concepts [18,
31]) in the design of the blocks for our tangibles
version of Bots & (Main)Frames.
Self-beliefs play an important role in academic
development and learning outcomes . Programming
in particular presents many challenges to learners' self-
beliefs as the radical novelty  of the concepts and
material can evoke strong negative feelings [17, 21,
35], creating barriers to learning . To better assess
and understand these complex affective issues, Scott
and Ghinea  used the Control-Value Theory of
Achievement Emotions [28, 29] to develop and validate
an instrument that assesses programming self-beliefs.
Underlying the self-beliefs model of their instrument
are three core constructs/subscales: 1) programming
self-concept—a composite of self-perceptions that are
formed through experience with and interpretations of
one’s environment ; 2) programming interest—the
extent to which an individual enjoys engaging with a
set of tasks ; and 3) programming aptitude—based
on Dweck’s  notion of mindsets where students
have either a growth mindset (i.e., belief that their
capacities can be improved through practice) or a fixed
mindset (i.e., belief that their capacities are inherent).
Students with a growth programming aptitude tend to
maintain practice when they encounter difficulty while
Figure 2: Buttons used to program
in the mouse version of Bots &
Figure 3: Tangible programming
blocks used to program in the
tangible version of Bots &
those with a fixed programming aptitude do not .
For the programming aptitude subscale, lower scores
indicate a greater growth mindset.
Bots & (Main)Frames Game Design
Bots & (Main)Frames (see Figure 1) is designed to
reflect typical aspects of educational programming
games. The game objective is to program a virtual
robot to reach all of the red tiles in a maze from a given
starting point, using a limited number of commands in
each level. Players are able to program the robot to
move forward, rotate 90 degrees left or right, use a
loop (repeat one command a specified number of
times), or use a function. Each level is designed to
increase in complexity and programming commands
needed to solve it. In the mouse version of Bots &
(Main)Frames, players click UI buttons to program (see
Figure 2). In the tangibles version, players instead
physically connect wooden blocks to program
commands (see Figure 3). The UI and gameplay is
identical to the mouse version in Figure 1, except that
buttons are disabled and programming commands that
appear on screen are instead created by configurations
and connections of the tangible programming blocks.
The tangible programming blocks utilize fiducial
tracking from the ReacTIVision framework  in order
to detect which blocks are connected and program the
virtual robot in game. Loop and Function/Use Function
blocks are also designed to utilize physical form to
represent corresponding programming concepts
(similar to [16, 22, 43, 44]). Loops have an additional
slot for players to slide in the command that will be
repeated (Figure 4), and Function/Use Function blocks
are connected by a chain to better illustrate the flow of
code execution from one function to another (Figure 5).
Based on research illustrating the effectiveness of
tangibles for enhancing engagement and motivation [8,
11, 42], we hypothesized that: 1) both versions of Bots
& (Main)Frames would improve programming self-
beliefs for novice programmers, but that the tangibles
version would show relatively greater improvements;
and 2) that the tangibles version would be considered
more enjoyable to play than the mouse version.
Experimental Design and Procedure
To test the first hypothesis (H1), we used a between-
group study design where participants were randomly
assigned to a mouse or tangibles condition. In the
pretest survey, participants completed a demographic
questionnaire (collecting age, gender, academic major,
years of prior video game experience, and years of
prior programming experience) and validated subscales
from the programming self-beliefs questionnaire (i.e.,
programming self-concept, interest, and aptitude) .
After submitting the pre-test scales, participants were
randomized into either the mouse or tangibles
condition, played 10 levels of the corresponding game
version of Bots & Main(Frames), and again submitted
responses to the programming self-beliefs
questionnaire (experimental phase).
To test the second hypothesis (H2) regarding perceived
enjoyment, we used a within-subjects design where
participants also played 10 levels of the other version
of Bots & (Main)Frames (comparison phase). The levels
were modified to be of similar difficulty to levels in the
experimental phase but still require novel solutions.
Lastly, participants filled out a questionnaire that rated
gameplay enjoyment on a 7-point Likert scale ranging
from 1 (very unenjoyable) to 7 (very enjoyable) for
Figure 4: A Loop programming
block where players insert a
command that will be repeated a
specified number of times.
Figure 5: Function and
corresponding Use Function
programming blocks connected by
Figure 6: The experimental design to compare self-beliefs and enjoyment across mouse and tangible versions of Bots & (Main)Frames.
After submitting a pre-test survey, participants were randomized to a condition, played either the mouse version or tangibles version
and then supplied post-test responses to the programming self-beliefs questionnaire. Participants then played the other game version
and provided comparison ratings of enjoyment.
both game versions, and answered questions about
their play experience in a concluding in-person
interview. For the full experiment design, see Figure 6.
A total of 34 university participants (ages 17-35,
median: 18.5) were randomly allocated to one of two
conditions. The mouse game condition consisted of 18
(5 male) participants and the tangibles game condition
consisted of 16 (8 male) participants. All participants
had 6 months or less of programming experience and
worked collaboratively in pairs of two for each test.
Only two pairs knew each other before the study. The
pair programming approach was used to examine
collaboration differences between conditions, however
(due to space concerns) the analysis will be excluded
from this paper and presented in future work.
We analyzed pre- and post-game questionnaires to
assess changes in programming self-beliefs subscales
between- and within-participants, and ratings of
enjoyment for both game versions. We found that all
participants, regardless of game version, improved on
the 3 subscales of programming self-beliefs. However,
the tangibles received higher enjoyment ratings.
A series of t-tests and Chi-Squared tests found no
significant condition difference in pre-existing
experience or demographic characteristics (all ps
>.05). However, there were pre-test condition
differences for ratings of programming interest in the
tangible condition (Mean: 11.92, SD: 1.82) compared
to the mouse condition (Mean: 9.91, SD: 3.41),
t(26.373)=-2.194, p=.037. Due to this difference, the
statistical tests we used to assess condition differences
control for pre-test ratings of programming interest.
Programming Self-Beliefs: General Improvements on
Programming Self-Beliefs Subscales
A series of Wilcoxon signed-rank tests found overall
significant changes in median pre-post scores for the
three subscales across both conditions: programming
self-concept (z=-3.687, p<.001), programming interest
(z=-4.062, p<.001), and programming aptitude (z=-
2.876, p=.004) (see Figure 7). However, there were no
significant differences in scores between conditions.
Figure 7: Median scores on
programming self-belief subscales:
self-concept, interest & aptitude.
Across both conditions there were
significant improvements in scores
for each subscale. NOTE: Lower
programming aptitude scores
indicate stronger growth mindset.
Enjoyment Ratings: Tangibles Version Rated as
Significantly more Enjoyable than the Mouse Version
After playing both games, participants supplied
enjoyment ratings for both the mouse version and
tangibles version. Ratings were given on a 7-point scale
ranging from 1 (very unenjoyable) to 7 (very
enjoyable). A Wilcoxon signed-rank test shows that,
across all participant ratings, the tangibles version had
higher median post-test ratings of enjoyment (Median:
7.0) compared to the enjoyment ratings of the mouse
version (Median: 6.0), z=-4.451, p<.001 (see Table 1).
Table 1: Enjoyment ratings for each game across all 34
participants. Ratings of enjoyment were significantly higher for
the tangibles version, relative to the mouse version.
This study revealed two important results regarding the
impact of puzzle-based educational programming
games and tangibles on crucial educational factors.
First, the Wilcoxon signed-rank tests used to evaluate
the efficacy of mouse and tangible input for improving
programming self-beliefs shows an improvement in
median pre-post scores for programming self-concept,
programming interest, and programming aptitude
(lower programming aptitude scores indicate more ideal
growth mindset). Our results partially confirm H1 since
both versions of Bots & (Main)Frames improve
programming self-beliefs for novice programmers.
However, there were no condition differences showing
greater improvement for the tangibles version. This
rejects part of H1 and suggests that both mouse and
tangibles can be used in educational programming
games to enhance self-beliefs for novice programmers.
Second, we found that median enjoyment ratings
across all participants were significantly higher for the
tangibles version relative to the mouse version. This
confirms H2 and suggests tangibles are more enjoyable
to use than mouse input for novice programmers.
Furthermore, since enjoyment has been shown to
improve long term interest in learning about science
, the use of tangibles could also potentially improve
long-term interest and compliance with programming.
Conclusion and Future Work
Our results show that while puzzle-based educational
programming games in general may improve players'
programming self-beliefs, the use of alternative forms
of physical input and interaction (e.g., tangibles) prove
more effective for important learning factors such as
enjoyment. However, there are many directions for
future work due to the small sample size and limited
analysis. For instance, our data suggested potential
greater benefits for non-STEM academic majors and
females using the tangibles version, but the sample
size was too small to perform an analysis of such
potential interaction effects. We also need to perform
an in-depth qualitative analysis to examine whether the
tangible approach better aids in collaborative aspects of
learning programming concepts. Additionally, it would
be useful to examine both conditions' impact on
remaining programming self-beliefs subscales such as
programming anxiety and debugging self-efficacy.
We would like to thank Connor Harada for his help
running studies as well as Kyle and Kaitlyn Kliewer for
their assistance in building the tangible blocks.
1. Ainley, M. and Ainley, J. 2011. Student engagement with
science in early adolescence: The contribution of
enjoyment to students’ continuing interest in learning
about science. Contemporary Educational Psychology. 36,
1 (2011), 4–12.
2. Antle, A.N. et al. 2008. Playing with The Sound Maker: Do
Embodied Metaphors Help Children Learn? Proceedings of
the 7th international conference on Interaction design and
children - IDC ’08 (2008), 178.
3. Bakker, S. et al. 2012. Embodied metaphors in tangible
interaction design. Personal and Ubiquitous Computing
4. Barr, D. et al. 2011. Computational Thinking: A Digital
Age Skill for Everyone. Learning & Leading with
Technology. 38, 6 (2011), 20–23.
5. Blackwell, L.S. et al. 2007. Implicit theories of intelligence
predict achievement across an adolescent transition: A
longitudinal study and an intervention. Child development.
78, 1 (2007), 246–263.
6. Blockly: A visual programming editor:
https://developers.google.com/blockly/. Accessed: 2016-
7. Computer Science For All: 2016.
science-all. Accessed: 2016-09-21.
8. Conradi, B. et al. 2011. Flow of electrons: an augmented
workspace for learning physical computing experientially.
Proceedings of the ACM International Conference on
Interactive Tabletops and Surfaces - ITS ’11 (2011), 182–
9. Dijkstra, E.W. and others 1989. On the cruelty of really
teaching computing science. Communications of the ACM.
32, 12 (1989), 1398–1404.
10. Dweck, C.S. 2000. Self-theories: Their role in motivation,
personality, and development. Psychology Press.
11. Gnoli, A. et al. 2014. Back to the future: Embodied
Classroom Simulations of Animal Foraging. Proceedings of
TEI ’14 (2014), 275–282.
12. Güldenpfennig, F. et al. 2016. Toward Thingy Oriented
Programming: Recording Marcos With Tangibles.
Proceedings of the TEI’16 (2016), 455–461.
13. Harteveld, C. et al. 2014. A Design-Focused Analysis of
Games Teaching Computer Science. Proceedings of
Games+ Learning+ Society 10 (2014).
14. Hicks, A. et al. 2014. Building Games to Learn from Their
Players: Generating Hints in a Serious Game. Intelligent
Tutoring Systems: 12th International Conference, ITS
2014, Honolulu, HI, USA, June 5-9, 2014. Proceedings. S.
Trausan-Matu et al., eds. Springer International
15. Hicks, A. 2012. Creation, evaluation, and presentation of
user-generated content in community game-based tutors.
Proceedings of the International Conference on the
Foundations of Digital Games - FDG ’12 (2012).
16. Hu, F. et al. 2015. Strawbies: Explorations in Tangible
Programming. Proceedings of the 14th International
Conference on Interaction Design and Children - IDC ’15
17. Huggard, M. 2004. Programming trauma: Can it be
avoided. Proceedings of the BCS Grand Challenges in
Computing: Education. (2004), 50–51.
18. Ishii, H. 2008. Tangible bits: beyond pixels. Proceedings
of the 2nd international conference on Tangible and
Embedded Intreaction (TEI ’08) (2008).
19. Kaltenbrunner, M. and Bencina, R. 2007. reacTIVision: a
computer-vision framework for table-based tangible
interaction. Proceedings of the 1st international
conference on Tangible and embedded interaction.
20. Kao, D. and Harrell, D.F. 2015. Mazzy: A STEM Learning
Game. Foundations of Digital Games (2015).
21. Kinnunen, P. and Simon, B. 2010. Experiencing
programming assignments in CS1: the emotional toll.
Proceedings of the Sixth international workshop on
Computing education research (2010), 77–86.
22. Kumar, V. et al. 2015. Note Code – A Tangible Music
Programming Puzzle Tool. Proceedings of TEI ’15 (2015),
23. Lode, H. et al. 2013. Machineers: playfully introducing
programming to children. CHI ’13 Human Factors in
Computing Systems (2013), 2639–2642.
24. Marsh, H.W. and Martin, A.J. 2011. Academic self-concept
and academic achievement: Relations and causal
ordering. British Journal of Educational Psychology. 81, 1
25. Melcer, E. and Isbister, K. 2016. Bridging the Physical
Divide: A Design Framework for Embodied Learning
Games and Simulations. CHI’16 Extended Abstracts
26. Melcer, E. and Isbister, K. 2016. Bridging the Physical
Learning Divides: A Design Framework for Embodied
Learning Games and Simulations. Proceedings of the 1st
International Joint Conference of DiGRA and FDG (2016).
27. Papert, S. 1980. Mindstorms: Children, computers, and
powerful ideas. Basic Books, Inc.
28. Pekrun, R. 2006. The control-value theory of achievement
emotions: Assumptions, corollaries, and implications for
educational research and practice. Educational psychology
review. 18, 4 (2006), 315–341.
29. Pekrun, R. and Stephens, E.J. 2010. Achievement
emotions: A control-value approach. Social and
Personality Psychology Compass. 4, 4 (2010), 238–255.
30. Pouw, W.T.J.L. et al. 2014. An Embedded and Embodied
Cognition Review of Instructional Manipulatives.
Educational Psychology Review. 26, 1 (2014), 51–72.
31. Price, S. 2008. A representation approach to
conceptualizing tangible learning environments.
Proceedings of TEI ’08 (2008).
32. Price, S. et al. 2010. Action and representation in tangible
systems: implications for design of learning interactions.
Proceedings of TEI ’10 (2010), 145–152.
33. Price, S. et al. 2008. Towards a framework for
investigating tangible environments for learning.
International Journal of Arts and Technology. 1, 3/4
34. Resnick, M. et al. 2009. Scratch: Programming for All.
Communications of the ACM. 52, (2009), 60–67.
35. Rogerson, C. and Scott, E. 2010. The fear factor: How it
affects students learning to program in a tertiary
environment. Journal of Information Technology
Education. 9, 1 (2010), 147–171.
36. Schweikardt, E. and Gross, M. 2008. The robot is the
program: interacting with roBlocks. Proceedings of TEI ’08
37. Schweikardt, E. and Gross, M.D. 2006. roBlocks: a robotic
construction kit for mathematics and science education.
Proceedings of the 8th international conference on
Multimodal interfaces (2006), 72–75.
38. Scott, M.J. and Ghinea, G. 2013. Educating programmers:
A reflection on barriers to deliberate practice. Proceedings
of the 2nd Annual HEA STEM Conference (2013).
39. Scott, M.J. and Ghinea, G. 2014. Measuring enrichment:
the assembly and validation of an instrument to assess
student self-beliefs in CS1. Proceedings of the tenth
annual conference on International computing education
research (2014), 123–130.
40. Scott, M.J. and Ghinea, G. 2014. On the domain-
specificity of mindsets: The relationship between aptitude
beliefs and programming practice. IEEE Transactions on
Education. 57, 3 (2014), 169–174.
41. Shelley, T. et al. 2011. Evaluating the Embodiment
Benefits of a paper-based TUI for Spatially Sensitive
Simulations. Extended Abstracts of the 2011 Conference
on Human Factors in Computing Systems (2011), 1375.
42. Skulmowski, A. et al. 2016. Embodied learning using a
tangible user interface: The effects of haptic perception
and selective pointing on a spatial learning task.
Computers & Education. 92–93, (2016), 64–75.
43. Wang, D. et al. 2016. A Tangible Embedded Programming
System to Convey Event-Handling Concept. Proceedings
of TEI’16 (2016), 133–140.
44. Wyeth, P. 2008. How Young Children Learn to Program
With Sensor, Action, and Logic Blocks. Journal of the
Learning Sciences. 17, 4 (2008), 517–550.
45. Wyeth, P. and Purchase, H.C. 2002. Tangible
programming elements for young children. CHI ’02
extended abstracts on Human factors in computing
systems - CHI ’02 (2002), 774.
46. Yannier, N. et al. 2016. Adding Physicality to an
Interactive Game Improves Learning and Enjoyment :
Evidence from EarthShake. ACM Transactions on
Computer-Human Interaction (TOCHI). 23, 4 (2016).
47. Year of Code and £500,000 Fund to Inspire Future Tech
Experts Launched: 2014.