UCL Masters research project: Designing Elaborative Game Feedback for Struggling Readers

Welcome to the final of a series of Blogs featuring 2017/2018 Masters projects on Education and Technology supervised at the UCL Knowledge Lab (partner and coordinator of iRead). Each of the projects featured has focused on a key research area of  iRead.

Alline Kobayashi

I come from a Computer Science background with a main interest in the Human-Computer Interaction field. After doing some research about usability and accessibility and working in an online learning platform as a web developer, I was invited to work on a project with children with difficulties related to reading and writing. Throughout my work with these children, I did my first Masters in Linguistics based on a case study of a boy who was a struggling learner. I investigated how the use of technology interfered or not in his literacy process. This experience working with children gave me an idea of a customizable crossword game that I have developed and released on App Store in 2016. Combining my background in Computer Science and my interest in Education, last year I started a Master’s degree in Education and Technology at UCL with the aim of deepening my knowledge in this field of interdisciplinary study.

Problem addressed

Think about the latest educational games that you have seen. How do they give feedback to the learner? When the learner gives the wrong response, does the app provide a hint or provide the child with an opportunity to try again? After analysing the feedback of five widely-used reading games for children, Benton and her colleagues (2018) found that most of these commercially available games were inconsistent with the education literature on feedback: at the end of a game activity, most of the games only indicated if the answer is right or wrong (for example, using sounds or colours), without giving more information to the learner about their performance, or indicating why the answer is right/wrong. Moreover, when the child made a mistake, the feedback was not clear and encouraged a trial and error strategy. These findings led the authors to identify an opportunity for research on the use of elaborative feedback to allow children to learn how to recover from their errors, what I call error feedback. Inspired by this study, the present research investigated the relationship between error feedback, learning, and motivation in an early literacy writing app for children.

Approach

The mixed methods approach adopted in the research combined three methods (observation of gameplay, questionnaire, and interview) through which six children evaluated feedback designs delivered when the learner makes an error in the context of of an iPad game. The study was conducted in Portuguese, and the participants were struggling learners who have difficulties related to reading and writing. Although all the participants were identified by their school as struggling with literacy, they had different backgrounds, studied in various schools and were in different stages of the literacy process. Therefore, they represented a heterogeneous and small group that allowed me an in-depth investigation.

The game used in this research is a free crossword game for children called “Língua de Gato”. Concerning error feedback, the game indicates visually if the answer is right or wrong: the letter changes from blue to green if it is in the right position and from blue to red if it is in the wrong place. The main reason for choosing this game as the focus of my research is the fact that its feedback (verification feedback) illustrates the typical behaviour of educational games and does not present elaborative feedback as recommended by the education literature.

Língua de Gato app: current feedback design

Methods

The research had two phases with two goals. The first phase focused on the evaluation of the current feedback of the app to understand its limitations. As the focus of the research was on error feedback, I wanted to observe events such as: when receiving the error feedback, will the child continue writing the same word? Will he/she replace the wrong letter? Also, in case of error, what strategies does the child use?

In the second phase I designed three new game feedback options:

• Option 1 immediate correct response feedback: when adding a letter in the wrong position, it highlights the wrong letter and the correct one.
• Option 2 immediate response-contingent feedback: instead of indirectly highlighting the correct or wrong letter, it shows a hint to the learner related to possible reasons why the mistake occurs with the aim of guiding him to the correct solution.
• Option 3 delayed topic-contingent feedback: it waits until the learner finishes writing the word to highlight the wrong syllable and show a message that has the aim of improving the learner’s understanding about the content of that syllable.

I aimed to understand the children’s perception of the efficacy of the three feedback types concerning their understanding and motivation to continue with the game task. After the explanation of how each new feedback design would work, the participants individually answered a questionnaire and participated in a semi-structured interview allowing me to understand their feedback preference.

Key highlights of research 

The findings of this study show the limitations of verification feedback, and provide new insights about the design of elaborative feedback. Specifically:

• Verification feedback did not engage child participants in a reflection about the ongoing learning or facilitate corrective behaviours. When the feedback merely validated the answer of the learner without encouraging him/her to reread what s/he wrote, it culminated in the child’s use of a trial and error strategy.
• Child participants with more severe literacy difficulties preferred correct response feedback (option 1), whereas more advanced learners preferred elaborative feedback (options 2 and 3). This finding showed a diversity within our participants’ preferences highlighting the importance of considering the learner’s skills when studying and designing feedback for games. Some challenging questions that emerged are: how to give feedback to learners who struggle the most in reading and writing? Should we use audio messages instead of writing cues? Should we mix or even sequence verification and elaborative feedback with the former building the child’s confidence to engage in the latter?
• Observing in depth a group of heterogeneous learners made me realise that the current literature on feedback should better examine the relationship between feedback and specific learner characteristics from a qualitative point of view in order to establish how general quantitative trends vary according to particular groups.

Relevance to iRead

We are developing a set of games that provide systematic elaborative game feedback to support children’s recovery from errors (Work package 6). This MA project connects with this work allowing us to problematise the needs and responsiveness of feedback of the different learner groups we are involving in the project.

Want to keep in touch with our project developments? Sign up to our termly newsletter!