Many kids describe their perception of programming to be “hard and boring.” Computational Thinking, including programming, is considered by many to be an important skill for the 21st century workforce. However, in order for kids to get interested in Computational Thinking is it essential to understand why kids perceive programming to be hard, i.e., to understand cognitive challenges, and why kids perceive programming to be boring, i.e., to understand affective challenges. Over the last 20 years our Computer Science education research creating Computational Thinking tools such as AgentSheets and AgentCubes, developing curricula and providing teacher professional development systematically explored what we call the Cognitive/Affective Challenges space. We have not only found strategies to gradually move from “hard and boring” towards “accessible and exciting” but in the process also developed research instruments to computationally assess cognitive as well as affective challenges. This paper outlines the Cognitive/Affective Challenges space, briefly describes Computational Thinking Pattern analysis as cognitive instrument, and illustrates Retention of Flow as affective instrument to assess motivation.
by Google [
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
• Affective Dimension: Making Programming more Exciting through
DomainOriented Tools. Domain-Oriented Tools [
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.
Some tools such as Inflatable Icons (Fig. 4) focus on the notions of ownership and
creativity by providing highly accessible mechanisms to create 2D, 3D and even
physical artwork, which then can be turned into programmable games. These ideas
can contribute significantly towards the broadening of participation [
What are indicators that cognitive/affective tools have been sufficiently compelling to
transform participation from “have to” to “want to?” The Scalable Game Design
project [
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.
CTPA can detect patterns independent of application (e.g., game versus
simulation). This enables the computation of indicators suggesting transfer of concepts
relevant to Computational Thinking from one domain to another. CTPA has been
validated and published.
• Affective Assessment: Retention of Flow (RoF). Retention of Flow uses education
data mining to assess Flow of students following instructions to create a game
project similar to IKEA customers following instructions to assemble furniture. A
Markov-chain model is employed to predict retention functions. With thousands of
students building games this model can determine motivational levels, i.e., Flow,
over time. Discrepancies between the model and actual data is indicative of
potential instructional challenges that may result in either boredom or anxiety. This kind
of retention data does exist for 3rd party activities such as the Angry Birds Hour of
Code 2013 tutorial created by Code.org. RoF has been used to compare
motivational levels across some of these tutorials suggesting very high levels of
motivation for Scalable Game Design projects. This research is in an early stage but has
already been published [
Retention of Flow is particularly relevant to the discussion of “have to” versus “want to” because it is based on voluntary tasks such as programming a game and measures how far participants will progress without being forced. Our 3D Frogger Hour of Code tutorial was intentionally designed to be a cliffhanger activity. That is, within the one-hour time limit of the Hour of Code event participants would be able to create the first couple of agents (the Frog, the road and trucks), create a first game level and program the frog to be controllable by cursor keys. The participants would be able to see that the tutorial has addition chapters raising the question of how likely they would “want to” continue.
In a first stage we just analyzed the retention data exploring how many participants would write at least 1 line of code, at least 2 lines of code, etc. This analysis revealed that the data closely matched negative exponential distributions characterized by survival functions and also found in participation drop off in MOOCs and other programming environments. Discrepancies between the negative exponential fit curve and the actual data (Fig. 2 blue versus black line) could be explained through three types of challenges: cognitive challenges, e.g., confusing instructions, technical challenges, e.g., browsers crashing when typing in unrecognized characters, and practical challenges, e.g., schools only allocating exactly one hour for the activity.
Most interestingly the negative exponential curve continued beyond the first hour of instructions and even beyond all instructions. That is, our cliffhanger approach worked in the sense that participants continued beyond the time typically allocated in schools at rates matching the rates of the first hour. Moreover, the rates were actually higher than the retention rates of the Angry Birds tutorial built by code.org. We can only speculate that the game design process including the design of their own characters and worlds was key to reach these high levels of motivation.
In a second step we wanted to model the decision process that participants must have gone through. Following instructions such as the instructions to build a LEGO Proc. of Fourth International Workshop on Cultures of Participation in the Digital Age - CoPDA 2016 Gothenburg (Sweden), October 23, 2016 (published at http://ceur-ws.org).
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.
construction, e.g., a LEGO Star Wars spaceship, can be conceptualized as sequence of instructions to build an artifact. Each instruction will only be followed with a certain probability. We have formulated a good design conjecture suggesting that it makes sense to design tutorials in a way that each step poses roughly the same level of challenge. A Markov chain can be used to model this process of equal probabilities to continue after each step. Using this model we can derive the probability to continue from the retention data to find anomalies.
+Kink
-Kink Drop
Drop
Drop
Drop +Kink
The data resulting from computing the probability (Fig. 5) to continue is very fine grained and can be used to quickly identify potential trouble spots in the instructional material. These so called drops and kinks can be interpreted as deviations from a Flow state, which, in turn, can shed light on the threshold between “have to” and “want to.” 3
Scientifically speaking it is typically desirable to be able to comprehend individual conditions of complex systems isolated from each other. The Cognitive/Affective space, ideally, would enable the orthogonal investigation of cognitive and affective concerns. However, in reality, it is difficult to separate probable interactions between Proc. of Fourth International Workshop on Cultures of Participation in the Digital Age - CoPDA 2016 Gothenburg (Sweden), October 23, 2016 (published at http://ceur-ws.org).
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.
these two dimensions. For example, it is likely that people are willing to tackle hard
problems in the presence of a compelling incentive. These incentives, in turn, may
interact with intrinsic and extrinsic motivation [
The Cognitive/Affective Space can be segmented into intrinsic motivation (Fig. 6, green = “want to”) and extrinsic motivation (Fig. 6, red = “have to”). Extrinsic motivation is often based on rewards such as receiving money for accomplishing a task. However, as suggested by Benabou, the impact of rewards onto intrinsic and extrinsic motivation is surprisingly difficult to predict. For instance, the impact of a reward is likely to depend on the context in which the reward is provided. An “ex ante” award, that is an award that is promised ahead of accomplishing the task may actually diminish intrinsic motivation by signaling that the task will be boring. “Ex post” rewards, in contrast, such as providing a bicycle to a hard working child, may suggest that a task was considered difficult and that the person accomplishing the task exhibited talent. In other words, unlike the “ex ante” reward, the “ex post” reward may have a positive impact onto intrinsic motivation.
Four examples illustrate the “want to” “have to” segmentation illustrated in Fig. 6: 1. Writing C++ Program to compute prime numbers. This is our classical example of a task that is hard and boring. Perhaps with the exception of mathematicians the intrinsic motivation of most people to compute prime numbers is minimal. Moreover, writing a C++ program to compute prime numbers is difficult tasks for people with no prior programming experience. 2. Cleaning up your room. There is no high cognitive load on cleaning up a room but this task is commonly perceived to be very boring. It may be difficult to develop intrinsic motivation for this kind of task due to its Sisyphean nature. It takes considerable effort to clean up the room but the chance that the room will quickly get messy again is high. 3. Watch cat videos. The immense number of people who have watched cat videos on YouTube or even shared cat videos through social media is astronomic and is indicative of how excited most people appear to get by watching cat videos. Watching these videos is certainly not hard. Tasks that are this easy and this exciting, i.e., tasks that are in this deep green section of the “want to” space are not typically well rewarded. In other words, it would probably be difficult to find a job based on tasks like this. 4. Solving Crossword Puzzles. Many find solving crossword puzzles interesting but some crossword puzzles can also be quite hard. The transition from 4a to 4b captures a gradual increase along the cognitive challenge dimension. From Monday to Saturday the crossword puzzle featured daily in the New York Times gradually increases from simple to super hard. Proc. of Fourth International Workshop on Cultures of Participation in the Digital Age - CoPDA 2016 Gothenburg (Sweden), October 23, 2016 (published at http://ceur-ws.org).
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
Revisiting Fig. 1 and comparing it to Fig. 6 suggest a potential contradiction. In Fig. 1
the upper right corner, i.e., the Holy Grail of Computer Science education appears to
be roughly aligned with “watching cat videos on YouTube” in Fig. 6. Importantly,
however, one should not conclude from this alignment that Computer Science
education focuses on tasks that are necessarily easy and exciting. Instead, the Holy Grail
should be understood as a low threshold starting point for the apprehensive masses
with no experience in programming sharing a negative perception of programming.
Perhaps the threshold of entry for novices should try to be as low as watching cat
videos in order to be become something that people “want to” do. Ultimately, however,
Computer Science education must include a well-designed path to move from simpler
towards more complex cognitive challenges. Similar to the notion of the gradually
more sophisticated New York Times crossword puzzles Scalable Game Design, as
Computer Science education strategy, is based on an approach to gradually move
from easy to hard Computer Science education challenges [
The boundary between “have to” and “want to” can be explored through the Cognitive/Affective Challenges framework. On the one hand, tools can be used to mitigate these challenges. On the other hand, instruments can be devised to measure cognitive and affective challenges. There are complex interactions between cognitive and affecProc. of Fourth International Workshop on Cultures of Participation in the Digital Age - CoPDA 2016 Gothenburg (Sweden), October 23, 2016 (published at http://ceur-ws.org).
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.
tive challenges but Retention of Flow is a research instrument that can identify and even measure concrete challenges. Once these challenges have been identified through research instruments they can be addressed to improve tools and instructions to gradually shift from activities that are “hard and boring” to ones that are “accessible and exciting.” 5
This work is supported by the National Science Foundation under Grant Numbers 0833612, 1345523, and 0848962, by the Hasler Foundation, the Swiss National Science Foundation under grant CRAGP2_158545, and. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of these foundations. 6