(image credit: Overton2012 via Wikimedia Commons, Creative Commons Attribution-Share Alike 3.0 Unported license)

An interesting study of online tutorials from Ada Kim and Andrew Ko at the University of Washington. The researchers sampled 30 popular and diverse online coding tutorials and analyzed what and how the sites taught learners. The sites sampled taught six programming languages Java, Python, PHP, JavaScript, C#, C, and included interactive learning including Scratch and Alice. The sites varied as far as aiming at specific ages.

Reviewing the approaches to the 30 site sample, they realized:

• Transforming information into knowledge is a key principle for effective learning. Binding a large set of disconnected facts is important as well as connecting prior knowledge into new knowledge. Establishing a learning framework helps students gain a deeper understanding of the material.
• The educational games provided the most immediate and personalized feedback, likely improving from deliberate practice. From a teaching perspective, games such as Gidget, Lightbot, CodeHunt, and tutorials provided by Code.org are currently most likely to be effective in learning.
• Deliberate practice helps learners achieve mastery in a particular domain.
• Clearly structured and articulated goals are critical to enhancing the effectiveness of deliberate practice, however, practice should be coupled with appropriately targeted feedback, including information about learners’ progress, to guide them toward goals.

The study suggests most online coding tutorials do not yet achieve many key principles in learning sciences. Development needs to better emphasize personalized support and contextualized feedback. They need to explain to learners why and when to use particular coding concepts.

The researchers recommend that teachers be very selective in their use of materials, focusing on the more evidence-based tutorials, particularly educational games. The games provide hierarchical structure, immediate feedback, and opportunities that learners actively write code and use subsequent knowledge for coding throughout the tutorial.

Interesting Findings in the Study:

1. Few tutorials explain when and why a particular concept is useful in programming.
   1. Few provide guidance for common errors.
   2. None provide personalization based on prior coding experience or learner goals, other than rudimentary age-based differentiation
2. Most of the tutorials did not teach the concept of objects or object-orientation.
3. Some sites start bottom-up (build on elementary principles) and others top down (Scratch: make the cat dance”)
4. Only a few tutorials, primarily educational games and creative platforms, offered project-based contexts that provided an explicit goal of a stage or a module.
5. Gidget equipped a sophisticated editor panel so that learners even could see the error messages and syntax errors in the editor, which was more instructive than just providing a text guideline for practices.
6. All interactive tutorials and educational games with a code editor provided some form of immediate feedback, but much of this was shallow. Almost half of the tutorials did not provide feedback when learners made errors. These tutorials fell into two cases:
   1. Some tutorials like Scratch provided open-ended practice, but did not provide feedback about right or wrong code relative to a goal or
   2. A tutorial’s code editor did not produce feedback about error messages.
7. Some tutorials provided feedback through instructor or peer communication. For example, MOOCs provided some opportunities to communicate with instructors or peers, and some resources had online communities in which learners could ask questions. While this feedback was available, none of it was immediate and learners had no guarantee of receiving answers to their questions. Assistance was rarely precise enough to improve learners’ conceptions of the material, and it was not customized at all to learners’ prior knowledge.
8. Coding tutorials gave more attention to emphasizing how to practice particular commands and functions rather than to provide information like when and why to use them.
9. Most of the web references did not establish a specific learning context for what they taught, whether an authoritative lecture based context, a goal-driven context, or a story-based context.

Study: Kim, Ada S and Andrew J. Koh, Pedagogical Analysis of Online Coding Tutorials, SIGCSE’17, March 8–11, 2017, Seattle, WA, USA. ISBN 978-1-4503-4698-6/17/03

What are you looking for in tutorials? Post your thoughts in the comments.