Little Thinking Minds

Arabic is the fifth most spoken language globally and the second hardest to learn. What makes it hard to learn is that the spoken language is different from the written language with only 60-70% overlap (World Bank, 2021). According to the World Bank 50% of children in MENA cannot read or comprehend an age appropriate text (World Bank, 2021). Currently, there are over 15 million children who are out of school due to ongoing regional conflicts; this has been compounded by millions being out of school due to COVID (UNICEF, n.d.). Given that the ability to read and understand a simple text provides a foundation for further education, it is imperative that children get the language and cognition development they need in the early grades of schooling (Duke, 2019; Hernandez, 2011). Progress toward eliminating this learning crisis can only be made with the conviction that literacy is at the core of human development and necessitates innovative, context-specific, and culturally responsive approaches. This can be done by leveraging technology to design innovative learning experiences that are both equitable and impactful on learning outcomes (OECD, 2020). 

I Read Arabic (IRA) is a platform that aims to accelerate learning in Arabic literacy by providing an engaging and individualized learning experience. It offers self-paced and digitized educational content for schools and at-home users. All too often, schools in the region incentivize rote memorization and meek recitation of text written on a blackboard (UNICEF MENA, 2017). Curriculum in the region promote and reinforce grammatical and spelling rules in a rote and abstract way that discourages children from making up words or misspelling, while traditional whole-word approaches to teaching the alphabetic system prevail (Saiegh-Haddad & Schiff, 2016; Taha & Saiegh-Haddad, 2016). Research indicates that a scaffolded and differentiated learning approach is more effective than the traditional approaches to instruction (Bondie et al., 2019; Gregory & Chapman, 2012; Romanelli et al., 2009; Smale-Jacobse et al., 2019). However, tailoring instruction to each student’s needs and preferences can be a challenge to practice as it entails far more than the adaptation of curricula ( Dosch & Zidon, 2014; Turner et al., 2017). IRA tailors learning experiences to fit students’ strengths, needs, goals, and pace of learning while supporting and extending the capacity of teachers. It places learners in appropriate learning levels, tailor content delivery, provide timely feedback and resources, and allow learners to advance or regress individually based on learners' performance (like the number of errors made and type of errors). 

IRA spans the entire range of early reading; it begins with letter naming and sound recognition, to being able to read books through connected text. The platform offers 23 levels based on the Queen Rania Teaching Academy Standards for Teaching Arabic for Native Speakers. The learning content is presented in a fun and engaging way and follows a gamified approach to assessment. IRA also aids teachers with the teaching-learning process by facilitating real-time monitoring of student/cohort progress while easing administrative tasks like grading reading exercises. It translates the performance data into actionable insights so teachers can intervene when a student is struggling.

The tailored content and the personalized and timely feedback will ensure that individual learning gaps are adequately addressed. Our target learner population are learners that come from low socioeconomic backgrounds and low-middle income countries in the MENA region. To make our platform accessible to a larger population of learners we offer it at affordable prices (and at times for free) for projects with ministries of education in the region, NGOs that focus on remedial education and displaced communities, as well as non-formal education settings. We also offer an offline version of our platform for low-resource settings where access to the internet is not feasible. 

In user experience (UX) design, it’s crucial to develop and refine skills to understand and address rapid changes in users’ environments and behaviors. For that we follow the Design Thinking (DT) approach. DT is a nonlinear five-stage process. In the 1st stage (i.e., Empathize) we gain an empathetic understanding of the user experience through one-on-one interviews. Here we set aside our own assumptions about what works and gain real insight into students, teachers and parents and their needs. Empathy is crucial to a human-centered design process such as the design thinking. We then analyze our observations from the interviews and synthesize them to define the core problems users have shared with us. We then look for alternative ways to view the problems and identify innovative solutions. Once we identify the best possible solution for each problem, our team produces scaled-down versions of the specific features to investigate the ideas we’ve generated with our users. Based on users’ feedback we make further iterations, alterations and refinements. This is an iterative process that gets repeated till we get to a solution that works for our users.

We are currently at phase 4 of the Nesta Standards of Evidence. We used the EGRA assessment tool (Early Grade Reading Assessment to measure the impact of our platform, specifically literacy gains pertaining to (letter sound identification, syllable identification, non-word reading, oral reading fluency, reading comprehension and listening comprehension). A third party impact evaluation using EGRA showed a 30% improvement in literacy gains (results here). 

• Our education team identifies the learning outcomes for each level, and tie them to the book quizzes and end of level assessments that the students have to go through in order to progress in their journey

• The tracking is done real time based on the questions and answers by the student

• We have been tracking the data for 3 academic years

Some statistics and interesting data points: 

• Grade 3 students seem to be taking longer to achieve their assigned learning outcome goals than any other grade. The data further shows that the same group of students start off their academic year at a lower level than any other student group in other grades, however they catch up towards the end of the year.

• On average children complete 1.8 levels

• Active students perform on average 34 reading activities per year, 32 quizzes per year, and 30 listening activities. Children in grade one spend on average 225 minutes per year reading, this is very interesting because official statistics tell us that Arab children read six minutes per year; this shows the reason is lack of access and availability of books and level-appropriate books  and not lack of interest in reading.  https://bit.ly/3lD0EUc

• Most active students are Arabs in Diaspora 

We measure the impact by: 

1- Time spent reading and numbers of books read per year

 (on average Arab children read 6 minutes a year according to the Arab Thought Foundation https://en.unesco.org/creativity/policy-monitoring-platform/arab-reading-challenge#:~:text=Reports%20and%20studies%20have%20shown,Arab%20Report%20for%20Cultural%20Development.(

2- Improvement in learning outcomes (how is it measured)

3- Levels progression

Our solution is evidence based and scientifically proven to improve literacy outcomes.. A third party impact evaluation found students who participated in the Qysas project achieved statistically significant greater gains than comparison group students who did not participate in the project on three EGRA subtasks:  syllable identification, oral reading fluency (ORF), and reading comprehension. The study took place among 20 public schools in Jordan comparing treatment students (those using our app for one year twice a week) with a control group (students in the same grade who are not using the app) over the course of a year, The study was conducted by San Francisco based School to School Research and was part of the All Children Reading Grand Challenge for Development funded by USAID, Australian Aid and World Vision.

(Please refer to the final report document.). 

(Full study: https://allchildrenreading.org/wordpress/wp-content/uploads/2017/10/Little-Thinking-Minds-End-of-Project-Report.pdf)

I Read Arabic is an Arabic literacy platform that provides an engaging and individualized learning experience.

While access to technology alone does not advance literacy, adaptive and personalized learning (PAL) holds the promise of leveling the playing field when the available resources and the need for this technology are taken into consideration (McCarthy et al., 2020; Wagner et al., 2014). Grounded in pedagogical differentiation, PAL that tailors instruction to each student’s learning ability, skill level, and interest can be more effective than the one size fits all curricular approach prevalent in many classrooms around the world (Gregory & Chapman, 2012; Turner et al., 2017; Smale-Jacobse et al., 2019). PAL draws on earlier research on the Zone of Proximal Development (ZPD; Vygotsky, 1978) and places learners in the right reading level, allowing them to learn in their preferred learning style, while identifying areas which the learner is yet to master. To better gauge learners' progress in the classroom, PAL provides teachers with individual learning and class-based reports pinpointing challenging areas for individuals or groups of students. Parents are important facilitators of children’s learning, this tool will enable parents to take an active role in their child’s learning journey by providing them with a detailed progress report so they can monitor their performance and provide support when needed. 

Our current platform does not adapt the content frequently, it only places the student at the right level based on a diagnostic test. All learners placed at a certain level are presented with the same learning content and activities, and progress linearly through one predetermined learning pathway. This is where the current platform falls short. Learners have to demonstrate some mastery to be presented with new content, but in the current version learner’s individual learning gaps are not identified beyond the initial placement test. It will be crucial for students to demonstrate their understanding of skills and concepts before moving on to more complex objectives. 

Based on the feedback we got from our users, we realized the need to provide adaptive content. Following the Design Thinking approach, we created a prototype that adapts the content to investigate its effectiveness before we implement it. This prototype leverages digital content used in a recent study that produced positive effects of a locally-contextualized learning multimedia platform for improving early grade reading outcomes (see Figure 1). This study was conducted as part of the All Children Reading Grand Challenge for Development where Little Thinking Minds  was one of 13 global grantees.  School to School Research (a San Francisco based company) conducted an experiment that focused on the assessment of early grade reading skills to understand the ability of technology-based innovations to improve the literacy skills of early grade learners using the Early Grade Reading Assessment (EGRA) tool. The EGRA is an oral assessment that measures students’ most basic foundational literacy skills in the early grades—specifically, recognizing letters of the alphabet, reading simple words, understanding sentences and paragraphs, and listening with comprehension. It began in January 2015 and ended on July 31, 2017. Students who participated in the project achieved statistically significant greater gains than comparison group students who did not participate in the project on three EGRA subtasks: syllable identification, oral reading fluency (ORF), and reading comprehension.

While we found that personalized learning (placing a student in the right level), we are yet to explore the impact of using an Adaptive and Personalized Learning (PAL) technology that adapts the content, feedback and resources frequently based on students engagement with the learning material on early grade literacy skills in Arabic. Currently, the literature on the effectiveness of adaptation within K-12 settings presents an inconclusive assessment of its impact on student’s reading achievement (Shamir et al., 2019). We hope through the help of the LEAP Fellows to establish a proof of concept by hypothesizing and answering the following questions: 

How does the adaptive and personalized-mediated instruction compare with the non-adaptive mediated instruction (i.e., personalized learning) in terms of increasing the overall Arabic early grade literacy? 

How do students view their overall experience with the system?

Creating adaptive platforms is not an easy task. Exploring the impact of this technology on early grade literacy skills in the region will help us decide whether we should invest in turning the entire platform into an adaptive one. This investigation may also offer international development actors, policy and decision-makers, and other stakeholders insight into not only the extent to which PAL improves literacy skills, but also whether it perpetuates, or even exacerbates, inequity. The findings can then be used to drive change in policy and budget allocations to ensure equitable quality literacy instruction, and to promote lifelong learning opportunities for all children as envisioned by the Sustainable Development Goals (UN General Assembly, 2015).