Can Pupils Teach Themselves to Code using Minecraft?

The following is a dissertation proposal and am looking for a school to develop this in. If this sounds interesting, please get in touch.

Abstract
There is a growing need in schools in the UK to prepare pupils for the digital economy by developing 21st-century soft skills. There is also a shortage in England of specialist computing teachers. Traditional approaches to teaching and learning rooted in the industrial age are being challenged by ones which leverage the use of EdTech. Evidence-driven pedagogies such as peer-peer learning, data-driven instruction, gamification and constructivist approaches to learning that allow for greater levels of personalisation, may help resolve Benjamin Bloom’s 2-sigma problem. There is also a shortage of special education needs teaching assistance in the schools. On the other hand, companies like Microsoft and Google are celebrating neurodiversity by using non-traditional approaches to selecting prospective candidates. e.g. Microsoft use Minecraft as part of their interview processes.

The aim of this research project is to evaluate the conditions which may influence peer-peer learning in Minecraft as primary and secondary school pupils are given the challenge of learning to code and programme in Python. The method will be based on the research of peer to peer learning (Betts, 2013; Mitra, 2019), gamification ( Betts, 2013; Resnick, 2017), cognitive presence (Kanuka, Rourke and Laflamme, 2007). Pupils will be given a variety of tasks ranging from skills building to group project work. Assessments will include peer-assessed evaluations and will conclude with a final test. The analysis will explore the potential influence of demographics, neurodiversity and the development of 21st-century skills. SPSS will be used to conduct the following three statistical analysis methods: ANOVA, two-sample t-test and correlation analysis.

Keywords: gamification, constructivism, peer-peer learning, Minecraft, mastery learning, coding

Background
21st Century Skills
The World Economic Forum (2019) , has set out eight characteristics which define high-quality learning in the Fourth Industrial Revolution – Education 4.0, which includes student-driven learning, problem-based and collaborative learning, interpersonal, innovation and creativity, and technology skills.

This is further reinforced by the Innovative Teaching and Learning project (ITL Research, 2014), which investigated the impact of innovative teaching on student learning and produced the following six strands:

1. Knowledge Construction
2. Skilled Communication
3. Collaboration
4. Real-Life Problem Solving
5. Use of IT
6. Self Regulation

The ITL research was used by Microsoft (2020) to develop their 21st Century Learning Design 21CLD rubric and by Michael Fullan in his Deep Learning model (Fullan and Quinn, 2017). Studies have shown young people often over estimate their digital skills and that there is a gap between ‘lifestyle’ and ‘workplace’ skills (DLGI, 2016). Given the need to develop such skills in pupils for the digital economy, the recent warning from Ofsted on the lack of suitably qualified computing teachers in England is concerning: there is a need for Continue Professional Development (CPD) for primary school teachers and 46% of secondary computing teachers holding a computing qualification (TES, 2022).

Pedagogy
Benjamin Bloom demonstrated how in his 2-Sigma Problem, pupils who received a combination of mastery learning and 1-1 tuition, performed 2 grades or standard deviations better than those taught traditionally (Bloom, 1984). In his quest to solve Bloom’s 2-Sigma Problem, Betts (2013), suggests peer-peer learning and gamification may be the key. Interestingly, it has been shown constructivist approach can raise achievement more than a mastery learning approach alone (Gokalp, 2017). There are cases, however, where constructivist approaches are combined with mastery learning. 42, a coding University in Paris, offer a free, project-based learning environment that relies on peer-peer learning pedagogy without teachers or courses (Beard, 2018); the model has been influenced by the work of Resnick, (1998, 2017). In addition to supporting self-regulation, peer-peer learning has also been linked to improved educational outcomes. It was found collaborative use of the internet in a Self Organised Learning Environment (SOLE), positively affected reading comprehension in children (Mitra, 2019).

Gamification and Constructivism
Minecraft has its roots in Piaget’s constructivist theory of learning in which learning occurs through (re)construction rather than knowledge transmission and knowledge is socially constructed (Schifter and Cipollone, 2015). As an instructional manual for Minecraft wasn’t officially developed by its developer, this raised the level of problem-solving needed to play the game, resulting in communities of YouTubers who share their knowledge through modelling (Niemeyer and Gerber, 2015). Microsoft recently used Minecraft for more inclusive hiring to allow neurodivergent applicants to easily demonstrate their skills (Microsoft, 2022), comes as no surprise, given the research around the benefits of Minecraft for pupils with autism (Rutkin, 2016; Wolf et al., 2016). In addition to the shortage of qualified computing teachers mentioned earlier, there has been a recent crisis of teaching assistants for special education needs and disabilities (SEND) quitting the profession ‘in droves’ due to pay (Fazackerley, 2022).

I would therefore like to test peer-peer pedagogy, gamification and self-organised learning environments to see if pupils can teach themselves to code in Minecraft and evaluate factors which may influence it. I will explore the following questions in my analysis

1. Do demographics impact student engagement levels in a peer-peer learning environment?
2. Are neurodiverse learners more prone to excel in online peer-peer learning environments than through paper-based tests?
3. Does peer-peer learning in Minecraft support the development of 21st Century Learning skills in pupils?

Method

Piscine
42 have an incubation session called piscine, French for swimming pool with the implication being pupils are thrown in the deep end. Prior attainment of students is not a requirement for entry at 42 and therefore the piscine stage is used to assess students with practical coding tasks. I may not have the option of selective intake if working with pupils during school hours. Betts (2013) suggests pupils performing at less than half of the average score in his gamified environment can be used to predict at-risk students. Such students may need more teacher-directed support. A prior assignment in the form of a pilot hour-of-code task, could therefore be devised to ensure pupils have the necessary skills to engage in the project or have challenges tailored at the right level. If the project is run as an after school club, then the ‘piscine’ phase can be used to resolve teething issues and find the right cohort of pupils if it is oversubscribed.

Teacher as Learner

"Through dialogue, the teacher-of-the-students and the students-of-the-teacher cease to exist and a new term emerges: teacher-student with
student-teachers. The teacher is no longer merely the one-who-teaches, but
one who is himself taught in dialogue with the students, who in turn while
being taught also teach. They become jointly responsible for a process in
which all grow" - Paulo Freire

Paulo Freire has suggested the teacher is also a learner and the roles can be interdependent and interchangeable (Freire, 2017). The role of a teacher will no longer be the provider of knowledge but that of facilitating and managing the environment and establishing rules for engagement, especially as pupils are not adult learners and therefore can’t be left unsupervised in an online environment. Where they may need to break off into teams over an online call, the breakout room meeting will be recorded. This may well be an opportunity for the ‘teacher’ to learn about Minecraft from pupils by posing questions to encourage critical thinking. Part of the role may also involve monitoring pupil progress through data and making strategic interventions to the game design (see Curriculum section below)

Gamification, Mastery Learning and Experience Points
One of Betts’ (2013) conclusions is that gamified environment works most successfully when participants have been rewarded for competence and not completion. To measure competence, peer-peer engagement was analysed; meaningful engagement was only found when he applied the 4 levels of cognitive presence approach developed by Kanuka et al (2007) to evaluate critical thinking levels and award points accordingly:

The advantage of having Minecraft sessions online over a Microsoft Teams call, where pupils can go into breakout rooms, is that the transcript can be analysed based on Kanuka’s (2007) cognitive presence model. Using the 42 model, I would encourage mastery learning so pupils only progress to the next challenge once they have ‘passed’ a challenge. Pupils will therefore be free to work at their own pace on individual challenges but will need to work in groups during their teamwork tasks. For effective engagement levels to be evaluated, XP points won't be given to pupils for the completion of a task.

At the end of each session, pupils will take two minutes to complete an exit ticket and answer questions about their learning journey.

Peer Assessment
Evidence suggests metacognition strategies in which pupils think about their own learning can be equivalent to an additional 7 months of progress (EEF, 2018). Using the 42 model, peer assessment will be carried out throughout the course as students complete each challenge and submit evidence of understanding through short, recorded videos demonstrating meta-cognition. They would need to request a peer review whenever they submit work. They will be allocated their peer who would have an assessment framework to offer feedback and transfer their knowledge. They will need to successfully complete a peer-review by offering a working solution which is explained in their own words to move to the next challenge.

Pupils will be allocated five experience points (XP); each time they request a peer review, a point will be deducted; each time they conduct a peer review they will gain an experience point.

Curriculum
I would like to explore the following three cases to allow for a full spectrum of coding and programming possibilities in Minecraft and scope to stretch the more able pupils.

Level 1: Block-based programming in Minecraft (key stage 2)
Level 2: Transitioning from block-based programming to Python (key stage 3)
Level 3: Python programming in Minecraft (key stage 4/5)

The Minecraft activities should be aligned with the national curriculum computing requirements for each key stage. Using the 42 model, there will be a series of tasks designed to acquire skills in which pupils can ask each other for help; in addition to this, there will be a group project in which those skills are applied. A weekly synchronous session either in person or online followed by opportunities for asynchronous collaboration would be provided to the pupils. A data-driven curriculum approach could be taken to make necessary adjustments to the environment, curriculum and pedagogy, if needed to cater for greater personalisation of pupils’ needs (Fletcher-Wood, 2018; Bambrick-Santoyo, 2019). Some pupils for example may need greater or lower levels of challenge or the XP and peer-assessment approach may need to be further refined to drive higher levels of engagement. This would be a data-driven interventive measure by the ’teacher’ provided the peer-peer dynamics of student learning are not disrupted.

Test score: At the beginning of the project, a baseline test will be used to bench-mark pupils’ knowledge of coding. At the end of the project, a timed test and a short practical in Minecraft will be conducted to evaluate the comprehension of the learning outcomes. A comparison will be made between the two assessments to determine if pupils have learned to code.

**Strengths and Weaknesses of the Approach **
The key advantage of the approach of peer-peer learning and constructivist pedagogy is it lends itself more to the gamified environment and less to teacher-led instruction.

Potential issues with the approach will be related to the level of skills needed for peer learning and self-regulation. In my experience of working with primary school children, this isn’t a major concern as many pupils play online games, including Minecraft, and naturally, play by collaborating online.
Data Analysis

The analysis will have two parts:

1. Descriptive analysis: basic summary statistics represented through graphs to establish an overview of the data e.g. mean, median, mode, outliers score range

2. Inferential analysis:

Based on the research question, the following three methods can be used:

Question Method
1 ANOVA: This consists of doing an analysis of variance (ANOVA); the method compares the variance of means for the data. ANOVA tests the mean of multiple independent variables at once.

2 Two-sample t-test

3 Correlation analysis

Question 1: Do demographics impact student engagement levels in a peer-peer learning environment?

The dependent variable will be the XP Score to measure engagement levels.
The independent variable is demographics: gender, English as an Additional Language, age, and Pupil Premium. An analysis of variance (ANOVA) can be conducted as there are more than two independent variables. The null hypothesis H0 and alternative hypothesis H1 are as follows:

Ho: The mean of the independent variables is equal
H1: The mean of the independent variables is not equal – at least one of them may vary

The hypothesis will be tested at 5% level of significance for possible acceptance or rejection of the null hypothesis.

Question 2: Are neurodiverse learners more prone to excel in online peer-peer learning than through paper-based tests?

Neurodiverse pupils will be identified through their SEND profile and potential outlier values.

Variable 1: test score of Minecraft
Variable 2: average test score in English and Maths (paper-based).

Two-sample t-tests can be conducted to determine the difference in mean between the two variables and observe their p-value.

H0: the mean difference between the two variables is insignificant
H1: the mean difference between the two variables is significant

The hypothesis will be tested at 5% level of significance for possible acceptance or rejection of the null hypothesis.

Question 3: Does peer-peer learning in Minecraft support the development of 21st Century Learning skills in pupils?

Correlation analysis can be used to look for an association between peer-peer learning (XP points and test score) and 21st-century learning skills (21CLD score); the 21CLD Rubric can be used to give a score which measures the pupils’ 21st-century skills across various levels based on the following strands:

Ho: there is no association between the two variables
H1: There is an association between the two variables (either positive or negative)

An association of greater than 5% level of significance will imply a significant impact of the variables on the dependent variable.
The analysis will conduct using SPSS software package.

Data Storage and Management

All participants will be working in the same Microsoft 0365 tenancy. Data will therefore be collected, analysed and stored within the tenancy. Any third party applications will require users to sign in using their school’s account. Applications to be used
• Minecraft Education Edition
• MS 0365 Apps (MS Teams, OneNote, Word, Excel, Stream, Forms)
• FlipGrid for screen recording and peer-peer feedback will require users to authenticate with their MS0365 school login; a private space will be made for E21C Trust users only.

Ethics
To minimise any potential ethical issues which may arise in the proposed research, I will consider the following principles proposed by Hammersley and Traianou ( 2012):

1. Minimising Harm
   Students could be ‘at harm’ if they are not learning over an extended period of time or experience cyberbullying. Each pupil’s progress and interaction will be monitored every week. Online sessions will be recorded and permissions will be set up in MS Teams so as not to allow pupils to delete comments once written. In the event a pupil is disengaged or simply not learning, I would have planned an alternative provision or activity for them to complete. This could include more teacher-directed support or a change of activity altogether, once the pupil's needs have been identified.

2. Protecting privacy
   Data will be stored and analysed within the same MS0365 tenancy. In order to protect student identity and maintain confidentiality, consideration should be taken when using social media to promote the research proposal and any photos taken and shared are done so on a school device under the school’s social media policy. I will use UPN and/or pseudo names to identify pupils if there is a need to share results more publicly e.g. during a presentation at a Microsoft event.

3. Respecting Autonomy
   An ethical consent form will be used to gain the consent of pupils to participate in this research project while making clear their autonomy to withdraw at any point should they choose to.

Risk Analysis
Some potential pitfalls, plans for avoidance and contingency plans are as follows:

1. IT issues around the firewall to ensure students can access YouTube and Minecraft; also students have enough bandwidth or processing capacity to join an MS Teams call at the same time, especially if they are collaborating with other students in the Trust in a Minecraft world.

Such IT issues can be avoided by
a) Ensuring all students have the correct MS 0365 license and IT access
b) URLs used by Minecraft are whitelisted
c) The computers used have required specification to run Minecraft and MS Teams simultaneously.
d) There is on-hand IT support during the first session to resolve any teething issues
e) A trial run is conducted for all the IT checks

In the event there are still IT issues preventing pupils from engaging, alternative provisions would need to be made. e.g. students borrow laptop if they lack access at home or library computers are made available before/after school.

2. Level of challenge: Some students may find the level of challenge too difficult or easy.

This can be resolved by offering such students a breakout session in MS Teams which is more teacher-directed or by simplifying the tasks. For the more able, there should be enough to challenge them with Python coding and being able to work collaboratively with their peers.

In the event of a lack of challenge, students will be encouraged to complete their key stage level and move on to the next level e.g. key stage 2 can move to key stage 3 and so on.

In the event other barriers, such as language or SEND needs, prevent a pupil from learning, despite more teacher-directed learning, in-class teaching assistance will be used to provide an alternative learning task altogether.

3. Accessibility. Pupils may have difficulties accessing the content due to language barriers or Special Education Needs.

I will use Immersive Reader in Minecraft and MS Teams to support pupils with dyslexia or English as an Additional Language. MS Translate can be used in the session (online or in person) to translate text and live speech into multiple languages.

Part of an induction task would be to assess accessibility requirements. In the event a pupil is struggling to engage with the task, a breakout session can be conducted with their teaching assistant in which they get targeted one-one support.

4. Managing online behaviour, particularly if students are able to use breakout rooms in Microsoft (MS) Teams to work in their group

MS Teams meetings, including breakout rooms, will be recorded. Ground rules and consequences will be established during the first session.

In the event of an online behaviour incident the school’s behaviour policy will be followed and sanctions given as they would be in any lesson.

5. For Level 3 which involves actual Python scripting to manipulate Minecraft relies on Azure notebooks. This may need setting up or potentially a cost. I will have a testing period prior to launching this part of the research project. In the event this cannot be achieved, I would consider alternative approaches such as using Raspberry Pi for Python-Minecraft coding.

6. Pupils at E21C Trust who can engage with the project. IT requirements or curriculum schemes of work may impact which groups of pupils I can involve in the research project. A solution may be to run the project as an after-school club rather than during lesson time; this offers the added advantage of pupils collaborating from a range of schools via MS Teams.