Update and add files to milestone 1

0_domain_study/README.md

@@ -1,33 +1,34 @@
 # [Project Name:  "Quantifying Post-Pandemic Math Learning Loss"]  
 
-## Problem Statement
+This folder contains all the research and documentation developed for
+ **Milestone 1** of our Collaborative Data Science Project. The focus
+  of this milestone is to clearly define a real-world, data-driven problem
+   based on our team’s personal experiences and domain understanding.
 
-*"Our team's combined experiences in education, data analysis, and policy
-research reveal a critical challenge: COVID-19 school closures caused Grade 3-5
-students in [Country/Region] to lose 8-12 months of math proficiency,
-disproportionately affecting low-income communities. Through [Teacher Name]'s
-classroom observations, [Analyst Name]'s data work with [Organization], and
-[Researcher Name]'s parent surveys, we identified three compounding factors:*
+## 📌 Project Theme
 
-- **Access Inequality**: 60% of rural students lacked devices for online learning
-- **Pedagogical Gaps**: Teachers received no training for hybrid math instruction
-- **Assessment Blindspots**: Existing data systems failed to capture granular
-learning loss
+**How did students in low- and middle-income countries experience changes
+ in their learning proficiency during the COVID-19 pandemic, and what kind
+  of support systems can address those gaps effectively?**
 
-This created a 40% wider proficiency gap between socioeconomic groups
-post-reopening, rendering traditional one-size-fits-all teaching methods
-ineffective. Our project combines frontline educator insights, policy analysis,
- and data science to identify targeted recovery solutions."
+Our project investigates post-pandemic learning gaps in **mathematics**,
+ especially among **primary school students** in under-resourced contexts
+  where digital remote learning was not available during school closures.
 
-## Research Question  
+---
 
-**"How did the COVID-19 pandemic affect primary math proficiency in
-[Country/Region], and which interventions show the highest potential
-for bridging these gaps?"**  
+## 📂 Contents
 
-## Problem Domain Analysis  
+### `problem-statement`
 
-### Systems Thinking (Iceberg Model)  
+A detailed narrative based on our personal experiences.
+
+---
+
+### `domain-summary`
+
+A summary of our team’s understanding of the problem using systems thinking
+ (**Iceberg Model**). We analyze:
 
 1. **Event:** Declining math test scores post-pandemic.  
 2. **Patterns:**  
@@ -38,3 +39,54 @@ for bridging these gaps?"**
    - Teacher training gaps for remote instruction.  
 4. **Mental Models:**  
    - "Online learning is ineffective for STEM subjects."  
+This helped us identify root causes and shape a more strategic research question.
+
+---
+
+### `literature-review`
+
+A thorough review of the global research landscape on pandemic-era education
+ loss, with a focus on:
+
+- Foundational numeracy decline  
+- Learning poverty in LMICs  
+- Inequity in access to digital learning  
+- Long-term risks if learning gaps are left unaddressed  
+
+This review supports our problem framing with data and literature.
+
+---
+
+### `resources`
+
+A curated list of data sources, articles, and research institutions relevant
+ to our topic. Includes:
+
+- UNESCO, World Bank, PAL Network, ASER, J-PAL, Save the Children  
+- Datasets and dashboards on learning loss, equity, and education recovery
+
+---
+
+### `research-question`
+
+Our final, actionable research question:
+
+**How did students in low- and middle-income countries experience changes in
+ their learning proficiency during the COVID-19 pandemic, and what kind of
+  support systems can address those gaps effectively?**
+
+This question is specific, relevant, and grounded in both personal experience
+ and available data.
+
+---
+
+## 🔜 Next Steps (Milestone 2)
+
+For **Milestone 2**, we will:
+
+- Explore available education data sets (e.g., test scores, dropout rates)  
+- Select a data model appropriate to our question  
+- Begin collecting and documenting relevant data  
+
+We are committed to improving team coordination and building on
+ this foundation with clear, focused analysis.

0_domain_study/domain_summary.md

@@ -1,5 +1,78 @@
 # Domain Research Summary  
 
+## Problem Domain Summary – Using Systems Thinking
+
+Our group’s understanding of the problem is shaped by both personal
+ experiences and the broader context of how the COVID-19 pandemic affected
+  education systems in low- and middle-income countries. To deeply analyze
+   the issue, we applied a systems thinking approach using the **Iceberg Model**,
+    which helped us go beyond the visible events and explore the underlying patterns,
+     structures, and beliefs that sustain the problem.
+
+### Event
+
+The most visible symptom of the problem is the decline in **mathematics proficiency**
+ among primary school students in the aftermath of the COVID-19 pandemic. In some
+  cases, students returned to school with significant gaps in understanding;
+   in others, they dropped out entirely.
+
+### Patterns and Trends
+
+These learning gaps are not isolated. They appear in various forms:
+
+- Reduced math test scores across entire grade levels  
+- Lower student confidence and engagement in math  
+- Increased dropout rates, especially in under-resourced areas  
+- Teachers reporting wide variability in student skill levels  
+
+These trends are consistent across many low- and middle-income countries,
+ especially where digital infrastructure was weak or nonexistent.
+
+### Systemic Structures
+
+Several systemic issues contributed to and sustained the learning loss:
+
+- **Lack of digital infrastructure:** Many students had no access to internet,
+   devices, or electricity during school closures  
+- **Weak national response:** In some regions, no remote learning plan was
+   implemented at all  
+- **Teacher overload:** Even when schools reopened, teachers had limited
+   time, training, or resources to address the wide learning gaps  
+- **Inequality in resource distribution:** Urban schools with more resources
+   recovered faster, while rural or poorer schools continue to struggle  
+- **Inflexible curricula:** Many education systems expected students to move
+   forward without adapting to the gaps created during the pandemic  
+
+### Mental Models
+
+Underlying beliefs and assumptions continue to hold these systems in place:
+
+- The idea that students who fall behind are "not smart"
+   rather than underserved  
+- Belief among decision-makers that learning loss will "fix itself" over time  
+- Cultural norms that may deprioritize girls' education or justify
+   dropping out during a crisis  
+- A lack of trust in data or systems to guide individualized interventions  
+
+### Summary
+
+The math learning gaps caused by the pandemic are not just about missed classes.
+ They are the result of deeply interconnected social, economic, and institutional
+  systems. Without structural changes, these gaps are likely to widen.
+
+A **systems thinking** approach helps us see that solving this problem requires
+ action on multiple levels:
+
+- Changing policies  
+- Improving data access  
+- Supporting teachers  
+- Shifting mindsets  
+
+By recognizing these interconnected layers, we aim to identify the most
+ influential factors in learning loss and explore interventions that
+  can lead to meaningful, equitable recovery in math education for primary
+   students in low- and middle-income countries.
+
 ## Key Findings
 
 1. **Pre-Pandemic Baseline:**  
@@ -9,7 +82,7 @@
 3. **Barriers to Recovery:**  
    - [Z]% of rural schools lacked stable internet (Source: [World Bank]).  
 
-### Data Sources  
+## Data Sources  
 
 - [ UNICEF ]
 - [ World Bank Open Data  ]  

0_domain_study/literature_review.md

@@ -1,5 +1,81 @@
 # Literature Review Summary: Learning Gaps in Math Post-COVID  
 
+## Background Review: Understanding Post-Pandemic Math Learning Gaps in LMICs
+
+The COVID-19 pandemic triggered the largest disruption to education in modern
+history, affecting over 1.6 billion learners worldwide. As schools shut down
+to reduce the spread of the virus, education systems were forced to rapidly
+transition to alternative modes of instruction.
+
+While some high-income countries quickly moved to online learning, many low
+and middle-income countries (LMICs) lacked the digital infrastructure,
+training, and resources to do the same. In some countries, education came
+to a complete halt for months or even years.
+
+According to a 2021 joint report by UNESCO, UNICEF, and the World Bank, school
+closures in LMICs often lasted longer than in wealthier countries.
+The remote learning options provided, if any, were significantly less
+effective. Nearly 463 million children had no access to any form of remote
+education during the height of the pandemic. In regions like Sub-Saharan
+Africa and parts of South Asia, the educational gap between children who
+could and could not continue learning widened drastically.
+
+Among the most negatively affected learning areas was **mathematics**.
+Math learning requires consistent practice, scaffolded instruction, and
+direct teacher feedback. It builds upon itself, so when foundational concepts
+are missed, like place value, operations, or fractions, students find it
+extremely difficult to progress. Unlike reading, which some students can
+continue practicing through informal means like reading at home, math
+instruction typically depends on structured learning environments, materials,
+ and regular teacher input.
+
+### Empirical studies support this decline
+
+- In **Brazil**, a study found that the proportion of grade 2 students who
+   could solve basic math problems fell by over **27%** during the pandemic.
+- In **rural Kenya and Uganda**, Learning Loss Assessment data revealed that
+   students returned to school after closures with math skills far below their
+    previous levels, with some performing at **pre-primary levels**.
+- Similar findings were reported in **Pakistan**, **South Africa**, **India**,
+   and **Bangladesh**.
+
+In addition to learning loss, the pandemic also **intensified educational inequality**.
+Children from marginalized communities, including girls, children with disabilities,
+refugees, and those living in poverty, faced higher risks of dropping out permanently.
+Parents without formal education often could not support their children’s learning
+at home, and in many cases, students had to take on adult responsibilities like
+work or caregiving.
+
+Even after schools reopened, the learning recovery has been **slow and uneven**.
+Teachers are now dealing with classrooms full of students at very different academic
+levels. There is often no diagnostic data to identify where students are struggling.
+In many LMICs, education systems have not provided teachers with remedial tools,
+catch-up programs, or specialized support for math recovery.
+
+Several global organizations have called this a **generational catastrophe**.
+The World Bank estimates that COVID-related learning loss could cost this generation
+of students up to **17 trillion dollars in lifetime earnings** if not addressed
+urgently. They also estimate that the global rate of **learning poverty**,
+defined as the percentage of 10-year-olds who cannot read or understand
+a simple story, rose to **70%** in LMICs during the pandemic.
+Although focused on reading, the same pattern is likely true or worse for math.
+
+The data also shows that **math proficiency** is a key predictor of long-term
+ academic success, employability, and income. This makes recovery in this domain
+  especially urgent. Yet, much of the current literature focuses on high-income
+   contexts or general learning loss, without breaking down the impact on specific
+    subjects like math or specific populations such as primary students in LMICs.
+
+Our research seeks to address this gap. By using **data science tools** and
+a **systems thinking lens**, we aim to examine the extent and causes of math
+learning gaps in primary education across LMICs. This will include analysis
+of performance trends, digital access, school closure lengths, and the
+presence or absence of national recovery plans. Our findings can help
+inform **targeted, evidence-based interventions** that prioritize the
+most affected students and support equitable recovery.
+
+---
+
 ## Key Papers Reviewed  
 
 ### **Paper 1**  

0_domain_study/problem_statement.md

@@ -0,0 +1,49 @@
+# Problem Statement: Based on Personal Experience
+
+During our group discussion, each of us shared personal experiences of how the
+COVID-19 pandemic disrupted math learning in different but connected ways.
+
+**Heba**  
+Heba, a math teacher in a public school, spoke about the difficulties she faced
+teaching during the pandemic. While some schools attempted remote learning,
+many of her students had no stable internet, shared devices with siblings,
+or no space at home to study. Teaching math online was especially challenging
+because it depends on real-time problem-solving and personal interaction.
+Even when students joined, they struggled to follow lessons.
+When schools reopened, she found that her students had widely different
+levels of understanding. Many had forgotten foundational concepts.
+Despite her efforts, it was nearly impossible to give each student
+the support they needed in large classrooms with limited resources.
+
+**May Mon**  
+From Myanmar, May experienced both the COVID-19 pandemic and a coup at the
+same time. When the COVID outbreak occurred, she had just completed high
+school and had to postpone her university education due to the pandemic.
+
+**Nada**  
+As a student, Nada faced a lot of difficulties during the COVID-19 pandemic,
+especially in learning math. One of the main challenges was the weak internet
+connection at home and the lack of enough devices. Her family had to share one
+device among several members. Also, instead of live classes, teachers would
+send recorded videos, but they weren’t clear, and she couldn’t understand
+the lessons properly. It wasn’t just math. She struggled in many subjects.
+She often felt lost and confused, and there wasn’t enough support to help her
+catch up.
+
+**Jubayer and Alexander**  
+Jubayer and Alexander also shared their stories about the challenges they faced
+in their education during the pandemic.
+
+---
+
+Together, these experiences show how the pandemic created serious,
+unequal learning gaps, particularly in math, in many low and middle-income
+settings. Teachers were overwhelmed, and students lacked access to the
+most basic learning tools.
+
+We believe this crisis demands urgent attention. Through data science,
+we aim to explore how deeply these math learning gaps run, which groups
+were most affected, and what interventions could help students recover
+and succeed. Our goal is to inform recovery strategies that are fair,
+targeted, and effective, especially for students and teachers in
+under-resourced contexts.