DepEd School Sanitation & Utilities 360° Statistical Audit: SY 2023-24
Comprehensive analysis of WASH facilities, sanitation infrastructure, and utility access across Philippine public schools, revealing critical inequalities and infrastructure gaps. Note: Sanitation data covers public schools only.
Photo by Mark Anthony Llego. Pictured: Desmond Grey at Basey I Central
Elementary School, Basey, Samar.
🔴 CRITICAL DATA LIMITATION: This analysis focuses EXCLUSIVELY on PUBLIC SCHOOLS. The National School Building Inventory (NSBI) contains NULL values for all sanitation metrics pertaining to the 12,349 private schools. All findings and comparisons in this report pertain ONLY to the public school system.
Executive Summary
Context & Data Scope
This comprehensive statistical audit analyzes 47,818 Philippine public schools for School Year 2023-24, representing the complete census of DepEd-recognized public schools nationwide. The analysis employs 35+ statistical techniques spanning distribution fitting, inequality measurement, hypothesis testing, regression modeling, clustering, and dimensionality reduction.
Headline Statistics
| Key Metric | Value | Interpretation |
|---|
| Schools Analyzed | 47,818 | Complete public school census |
| Average toilets per school | 17.5 (median: 11) | Right-skewed distribution |
| WASH functionality rate | 91.0% | High functionality |
| Schools with all three utilities | 29,370 (61.4%) | Majority have full access |
| Schools with ZERO utilities | 8,776 (18.4%) | CRITICAL |
| Gender Parity Index | 1.15 | Slight female bias |
Inequality Indices
| Index | Value | Interpretation |
|---|
| Gini Coefficient | 0.518 | High inequality (>0.4 threshold) |
| Palma Ratio | 3.31 | Top 10% have 3.3× bottom 40% |
| Theil Index | 0.465 | Decomposable inequality measure |
| Hoover Index | 0.377 | 37.7% redistribution needed for equality |
| Atkinson (ε=0.5) | 0.208 | Moderate aversion measure |
| Atkinson (ε=1.0) | 0.376 | High aversion measure |
Critical Findings
- High Facility Inequality: Gini coefficient of 0.518 indicates significant disparities in toilet facility distribution across schools
- Within-Group Dominance: Theil decomposition reveals 99.8% of inequality is WITHIN management types, only 0.2% BETWEEN types
- Infrastructure Poverty Crisis: 8,776 schools (18.4%) operate with ZERO utilities (water, electricity, or internet)
- Gender Parity Achievement: Mean GPI of 1.15 with 73.7% of schools achieving near-parity (0.9-1.1)
- Utility Access Gaps: Water (80.3%), Electricity (80.1%), Internet (62.7%) - internet lags significantly
Priority Recommendations
- Immediate: Target 8,776 zero-utility schools for emergency infrastructure intervention
- Short-term: Address within-group inequality through school-level needs assessment
- Medium-term: Expand internet connectivity to bridge digital divide (37.3% without access)
- Policy: Revise resource allocation formulas to reduce Gini below 0.40
Section 1: Introduction & Context
1.1 Background & Policy Context
1.1.1 Global & National WASH Policy Framework
International Commitments:
- SDG 6.2: Universal access to adequate sanitation by 2030
- WHO/UNICEF Joint Monitoring Programme: Standards for school WASH
- UN Convention on Rights of the Child (Article 24): Right to health
National Policies:
- DepEd Order No. 10, s. 2016: Policy and Guidelines for Comprehensive Water, Sanitation and Hygiene in Schools (WinS) Program
- DepEd Memorandum No. 194, s. 2018: Implementing Guidelines with Three-Star Approach
- Philippine WASH in Schools (WinS) Three-Star Approach: Step-wise quality standards
Key Policy Standard (DO 10, s. 2016): All schools shall have adequate, clean, functional, safe, and accessible toilet facilities meeting the 50:1 pupil-to-bowl ratio as stipulated in the Philippine Sanitation Code.
1.1.2 Importance of School WASH
Educational Impact:
- Attendance rates (especially menstruating girls)
- Learning outcomes correlation
- School retention and completion rates
Health Impact:
- Disease prevention (diarrhea, parasitic infections)
- Reduced soil-transmitted helminth infections
- COVID-19 and pandemic preparedness
Equity & Dignity:
- Gender equality in facility provision
- Inclusion of persons with disabilities (PWD)
- Menstrual hygiene management (MHM)
1.2 Research Objectives
This statistical audit addresses the following objectives:
- Quantify Current State: Measure toilet facilities, WASH infrastructure, and utility access across all public schools
- Assess Inequality: Determine distribution equity using 8 inequality indices with decomposition analysis
- Identify Disparities: Detect vulnerable school populations and infrastructure gaps
- Predict Outcomes: Model relationships between utilities and WASH functionality
- Provide Evidence: Support targeted policy interventions with rigorous statistical analysis
1.3 Scope & Limitations
Data Scope
| Category | Specification |
|---|
| Schools Included | 47,818 public schools |
| Schools Excluded | 12,349 private schools (NULL data) |
| Variables | 114+ covering toilets, WASH, utilities |
| School Levels | ES (39,346), JHS (10,561), SHS (7,743) |
| Management Types | 3 (DepEd, LGU, SUC-Lab, Other) |
| School Year | 2023-24 |
⚠️ CRITICAL DATA LIMITATION - PRIVATE SCHOOL DATA GAP:
- Issue: All 12,349 private schools contain NULL values for sanitation metrics
- Implication: NO public-private comparisons possible
- Technical Note: NSBI sanitation module covers public schools only
- All findings: Pertain EXCLUSIVELY to the public school system
Other Limitations:
- Self-reported data (potential reporting bias)
- Cross-sectional snapshot (no longitudinal trends)
- No quality assessment beyond functionality
- No behavioral/usage data (infrastructure only)
- Geographic coordinates unavailable for spatial analysis
Section 2: Methodology
2.1 Data Source & Collection
Source: DepEd National School Building Inventory (NSBI), SY 2023-24
Dataset Characteristics:
- Population: All DepEd-recognized public schools
- Sample: Census (100% of 47,818 schools)
- Collection Method: Online reporting system
- Validation: DepEd quality assurance processes
2.2 Variable Definitions
Sanitation Variables
| Variable Category | Naming Convention | Example |
|---|
| Attached Toilets | san_attached_LEVEL_FACILITY | san_attached_es_toilet_male |
| Standalone Toilets | san_standalone_LEVEL_FACILITY | san_standalone_jhs_toilet_female |
| Facility Types | toilet_male, toilet_female, toilet_pwd, toilet_shared | - |
WASH Variables
| Variable | Description |
|---|
wash_LEVEL_group_func_with_soap | Functional group facilities with soap |
wash_LEVEL_group_func_no_soap | Functional group facilities without soap |
wash_LEVEL_group_nonfunc | Non-functional group facilities |
wash_LEVEL_indiv_func_with_soap | Functional individual facilities with soap |
Definitions:
- Functional: Accessible, daily water supply, learner-appropriate height, appropriate drainage
- Group facility: Accommodates ≥10 learners (ES) or ≥4 learners (JHS/SHS)
- Individual facility: Classroom, drinking points, near toilets/canteen for critical-time use
Utility Variables
| Category | Variables |
|---|
| Water Supply | es_has_water_supply, jhs_has_water_supply, shs_has_water_supply |
| Electricity | LEVEL_has_electricity_grid, LEVEL_has_electricity_offgrid |
| Internet | LEVEL_has_internet, LEVEL_internet_purpose_instructional, LEVEL_internet_purpose_admin |
Derived Variables (Created for Analysis)
| Variable | Formula |
|---|
total_toilets | Sum of all toilet types across all levels |
wash_functional_total | Sum of all functional WASH facilities |
wash_with_soap_total | Sum of functional facilities with soap |
wash_functionality_rate | Functional ÷ Total WASH |
gender_parity_index | Female toilets ÷ Male toilets |
infrastructure_score | Sum of (water + electricity + internet) [0-3] |
has_water | Any level has water supply |
has_electricity | Any level has electricity (grid or off-grid) |
has_internet | Any level has internet |
2.3 Statistical Methods (35+ Techniques)
Distribution Analysis (4 methods)
- Shapiro-Wilk Test - Normality testing (n≤5,000)
- Kolmogorov-Smirnov Test - Goodness-of-fit to theoretical distributions
- Anderson-Darling Test - Tail-weighted normality test
- Chi-Square Goodness-of-Fit - Count data distribution fitting
Inequality Measures (8 indices)
- Gini Coefficient (with Bootstrap 95% CI)
- Palma Ratio (with Bootstrap 95% CI)
- Theil Index (with between/within decomposition)
- Atkinson Index (ε = 0.5, 1.0)
- Hoover Index (Robin Hood Index)
- Concentration Index
- Entropy-Based Inequality
- Lorenz Curve Analysis
Hypothesis Testing (6 methods)
- Mann-Whitney U (with Rank-Biserial Correlation)
- Kruskal-Wallis (with η² Effect Size)
- Dunn's Post-Hoc Test (Bonferroni correction)
- Wilcoxon Signed-Rank (Gender parity vs 1.0)
- Levene's Test (Variance homogeneity)
- FDR Correction (Benjamini-Hochberg)
Regression Models (6 approaches)
- OLS Regression (with Robust SE, HC3)
- Quantile Regression (10th, 50th, 90th percentiles)
- Zero-Inflated Poisson (excess zeros)
- Negative Binomial (overdispersion)
- Hurdle Models (zero vs. positive counts)
- Interaction Effects (Water × Electricity)
Clustering Methods (7 metrics)
- K-Means Clustering (with validation)
- Silhouette Score
- Calinski-Harabasz Index
- Davies-Bouldin Index
- Hierarchical Clustering (Ward linkage)
- Cophenetic Correlation
- Gap Statistic
Data Quality Assessment
- Missing Data Analysis - Pattern identification
- Outlier Detection - Tukey's fences (1.5×IQR)
- Sensitivity Testing
- PCA - Dimensionality reduction
- Factor Analysis
Section 3: Descriptive Findings
3.1 Overall Sanitation Infrastructure
3.1.1 Aggregate Statistics
Table 3.1: Toilet Facility Descriptive Statistics
| Variable | N | Mean | Median | Mode | Std Dev | IQR | CV (%) | Min | Max | Skewness | Kurtosis |
|---|
| total_toilets | 47,818 | 17.5 | 11 | 8 | 22.04 | 15 | 125.96 | 0 | 864 | 5.41 | 81.21 |
| total_toilets_es | 39,346 | 14.27 | 10 | 8 | 17.02 | 12 | 119.30 | 0 | 864 | 7.73 | 214.50 |
| total_toilets_jhs | 10,561 | 18.38 | 12 | 0 | 21.83 | 19 | 118.81 | 0 | 268 | 3.23 | 17.47 |
| total_toilets_shs | 7,743 | 10.48 | 6 | 0 | 13.66 | 10 | 130.29 | 0 | 259 | 4.29 | 36.99 |
Distribution Interpretation:
- Right-skewed: Skewness 5.41 indicates long right tail (few schools with many toilets)
- High CV (125.96%): Substantial variation relative to mean
- Median vs Mean: Median (11) better represents typical school than mean (17.5)
- Leptokurtic: Kurtosis 81.21 indicates heavy tails and peaked distribution
3.1.2 WASH Facility Functionality
Table 3.2: WASH Facility Statistics
| Variable | N | Mean | Median | Std Dev | Func Rate |
|---|
| wash_functional_total | 47,818 | 15.11 | 9 | 23.92 | - |
| wash_nonfunctional_total | 47,818 | 1.07 | 0 | 4.51 | - |
| wash_functionality_rate | 45,554 | 0.91 | 1.0 | 0.22 | 91.0% |
| wash_with_soap_total | 47,818 | 13.99 | 8 | 22.99 | - |
| wash_soap_rate | 44,186 | 0.92 | 1.0 | 0.23 | 92.0% |
Key Findings:
- 91% functionality rate: High proportion of WASH facilities are operational
- 92% soap availability: Among functional facilities, most have soap
- High variability: CV exceeds 150% for functional WASH count
3.1.3 Normality Assessment
Table 3.3: Shapiro-Wilk Normality Tests
| Variable | W-Statistic | p-value | Normal? | Interpretation |
|---|
| total_toilets | 0.6077 | <0.0001 | No | Non-normal distribution |
| wash_functional_total | 0.5012 | <0.0001 | No | Non-normal distribution |
| wash_with_soap_total | 0.4885 | <0.0001 | No | Non-normal distribution |
Statistical Implication: All key variables significantly depart from normality. This justifies the use of non-parametric methods (Mann-Whitney, Kruskal-Wallis, Spearman) throughout this analysis.
3.2 Utility Access
Table 3.4: Utility Access Rates
| Utility | Schools with Access | % of Total |
|---|
| Water Supply | 38,407 | 80.3% |
| Electricity | 38,306 | 80.1% |
| Internet | 29,971 | 62.7% |
| All Three Utilities | 29,370 | 61.4% |
| ZERO Utilities | 8,776 | 18.4% |
🚨 CRITICAL FINDING: 8,776 schools (18.4%) operate without ANY of the three basic utilities. These schools face compounded infrastructure poverty and require immediate intervention.
Utility Combination Patterns:
| Score | Utilities | Count | Percentage | Description |
|---|
| 3 | Water + Electricity + Internet | 29,370 | 61.42% | Full Access |
| 2 | Two of three utilities | 8,890 | 18.59% | Moderate Access |
| 1 | One utility only | 782 | 1.64% | Limited Access |
| 0 | NONE | 8,776 | 18.35% | CRITICAL Poverty |
Figure 3.1: Infrastructure Poverty Score Distribution Across Public Schools.
Nearly 1 in 5 schools (18.4%) operate with zero utilities.
3.3 School Level Distribution
Table 3.5: School Level Offerings
| Level | N Schools | % of Total |
|---|
| Elementary School (ES) | 39,346 | 82.3% |
| Junior High School (JHS) | 10,561 | 22.1% |
| Senior High School (SHS) | 7,743 | 16.2% |
Note: Percentages exceed 100% because schools can offer multiple levels simultaneously (e.g., integrated schools offering ES through SHS).
Section 4: Inequality Analysis
4.1 Comprehensive Inequality Measures
This analysis employs 8 inequality indices to provide a multi-dimensional view of facility distribution equity.
Table 4.1: All Inequality Measures
| Variable | Gini | Palma | Theil | Atkinson (ε=0.5) | Atkinson (ε=1.0) | Hoover | Concentration | Entropy |
|---|
| total_toilets | 0.518 | 3.31 | 0.465 | 0.208 | 0.376 | 0.377 | 0.518 | 0.043 |
| wash_functional | 0.573 | 4.16 | 0.560 | 0.242 | 0.428 | 0.414 | 0.573 | 0.052 |
Interpretation Guide:
| Index | Range | Interpretation |
|---|
| Gini (0.518) | 0-1 | >0.4 = HIGH inequality - CONCERNING |
| Palma (3.31) | 0-∞ | Top 10% have 3.3× the facilities of bottom 40% |
| Theil (0.465) | 0-∞ | Decomposable measure - see Section 4.2 |
| Hoover (0.377) | 0-1 | 37.7% of facilities need redistribution for equality |
| Atkinson (ε=0.5) | 0-1 | 20.8% "wasted" at moderate aversion |
| Atkinson (ε=1.0) | 0-1 | 37.6% "wasted" at high aversion |
4.2 Theil Index Decomposition
A key advantage of the Theil Index is its decomposability into between-group and within-group components.
Table 4.2: Theil Decomposition by Management Type
| Component | Value | Share (%) |
|---|
| Overall Theil | 0.465 | 100% |
| Between-Group Theil | 0.001 | 0.2% |
| Within-Group Theil | 0.464 | 99.8% |
📊 CRITICAL POLICY INSIGHT:
The decomposition reveals that 99.8% of inequality exists WITHIN management types, while only 0.2% is attributable to differences BETWEEN management types (DepEd, LGU, SUC-Lab).
Policy Implication: Reducing inequality requires targeting variation WITHIN each management type rather than equalizing across management types. This suggests school-level factors (size, age, location, enrollment) drive inequality more than systemic differences between governance structures.
4.3 Lorenz Curve Analysis
Figure 4.1: Lorenz Curves for Toilet Distribution - Public Schools. Distance
from 45° equality line reflects Gini coefficient of 0.518.
Lorenz Curve Interpretation:
The Lorenz curve plots cumulative percentage of facilities (Y-axis) against cumulative percentage of schools (X-axis). The 45° line represents perfect equality.
- Area between curve and diagonal: Represents inequality (Gini coefficient)
- Curve position: Further from diagonal = more inequality
- Key reading: Bottom 50% of schools share a disproportionately small share of total facilities
Section 5: Gender Parity Analysis
5.1 Gender Parity Index Overview
The Gender Parity Index (GPI) measures the ratio of female to male toilet facilities.
Formula: GPI = Female Toilets ÷ Male Toilets
Table 5.1: GPI Descriptive Statistics
| Metric | Value |
|---|
| Schools with GPI data | 32,080 |
| Mean GPI | 1.148 |
| Median GPI | 1.000 |
| Standard Deviation | 0.520 |
| Minimum | 0.000 |
| Maximum | 9.500 |
Table 5.2: GPI Distribution Categories
| Category | GPI Range | N Schools | Percentage |
|---|
| Male Bias | <0.9 | 1,339 | 4.2% |
| Near Parity | 0.9-1.1 | 23,650 | 73.7% |
| Female Bias | >1.1 | 7,091 | 22.1% |
Figure 5.1: Gender Parity Index Distribution. 73.7% of schools achieve
near-parity (GPI 0.9-1.1).
5.2 Statistical Test of Gender Parity
Wilcoxon Signed-Rank Test:
- Null Hypothesis: Median GPI = 1.0 (perfect parity)
- Test Result: p-value < 0.0001
- Conclusion: REJECT null hypothesis - significant deviation from perfect parity
Interpretation: While 73.7% of schools achieve near-parity (GPI 0.9-1.1), the system as a whole shows statistically significant deviation from perfect parity, with a slight bias toward female facilities (mean GPI = 1.15). This may reflect policy efforts to address menstrual hygiene management needs.
Section 6: Hypothesis Testing & Comparative Analysis
6.1 Management Type Comparisons
Table 6.1: Summary Statistics by Management Type
| Management | N | Mean Toilets | Median | SD | Mean WASH | Median WASH |
|---|
| DepEd | 47,802 | 17.48 | 11 | 21.98 | 15.1 | 9 |
Figure 6.1: Toilet Distribution by Management Type. DepEd-operated schools
dominate the dataset (47,802 of 47,818).
Note: The dataset is dominated by DepEd-operated schools (47,802 of 47,818), limiting the power of between-group comparisons. The Theil decomposition (Section 4.2) confirms minimal between-group variation.
Section 7: Regression Analysis
7.1 Quantile Regression
Quantile regression reveals how utility access affects different parts of the facility distribution.
Table 7.1: Quantile Regression Coefficients
| Predictor | Q10 (10th %ile) | Q50 (Median) | Q90 (90th %ile) |
|---|
| Constant | 4.00*** | 19.00*** | 64.00*** |
| has_water | -1.02** | -8.00*** | -33.00*** |
| has_electricity | -0.98** | -4.00*** | -14.00*** |
| has_internet | 1.00*** | 5.00*** | 19.00*** |
| Pseudo R² | 0.007 | 0.042 | 0.108 |
*Note: *** p<0.001, ** p<0.01, _ p<0.05_
⚠️ INTERPRETATION CAUTION:
The negative coefficients for water and electricity are counterintuitive and warrant careful interpretation:
- Possible confounding: Schools without water/electricity may have received recent infrastructure investments that included toilets
- Selection effects: Older, established schools with utilities may have aging infrastructure
- Internet as proxy: Schools with internet tend to be newer/better-resourced (positive relationship maintained)
- Low R²: Models explain only 0.7-10.8% of variance, indicating other factors dominate
Recommendation: These results should inform further investigation rather than direct policy conclusions.
Section 8: Clustering & School Typologies
8.1 K-Means Clustering Results
Table 8.1: Cluster Profiles
| Cluster | N Schools | Mean Toilets | Median Toilets | Description |
|---|
| Basic Level | 36,288 | 11.48 | 9 | Resource-constrained |
| Well-Equipped | 3,058 | 53.07 | 45 | Above-average facilities |
8.2 Cluster Validation
Cophenetic Correlation: 0.578
| Threshold | Interpretation |
|---|
| >0.75 | Excellent hierarchical structure |
| 0.60-0.75 | Good structure |
| <0.60 | Poor hierarchical fit |
Assessment: The cophenetic correlation of 0.578 falls slightly below the "good" threshold (0.60), suggesting moderate hierarchical structure. The 2-cluster solution (Basic vs. Well-Equipped) provides the clearest typology for policy targeting.
Figure 8.1: Cluster Validation Metrics. Silhouette and elbow analysis
support 2-cluster solution.
Figure 8.2: Hierarchical Clustering Dendrogram. Cophenetic correlation of
0.578 indicates moderate hierarchical structure.
Section 9: Dimension Reduction (PCA)
9.1 Principal Component Analysis
Variance Explained:
- PC1: 45.9% of variance
- PC1 + PC2: 80.8% of variance (cumulative)
Table 9.1: Component Loadings
| Variable | PC1 | PC2 | PC3 |
|---|
| total_toilets | -0.377 | 0.282 | 0.800 |
| wash_functional_total | -0.407 | 0.475 | -0.304 |
| wash_with_soap_total | -0.391 | 0.484 | -0.349 |
| has_water | 0.465 | 0.376 | -0.092 |
| has_electricity | 0.464 | 0.381 | -0.076 |
| has_internet | 0.329 | 0.417 | 0.363 |
Component Interpretation:
- PC1 ("Utility Access"): High loadings on water, electricity, internet
- Higher scores = Better utility access
- PC2 ("WASH Quality"): High loadings on soap availability
- Higher scores = Better hygiene infrastructure
- PC3 ("Toilet Quantity"): High loading on total toilets
- Higher scores = More toilet facilities
Figure 9.1: PCA Scree Plot and Cumulative Variance. First two components
explain 80.8% of total variance.
Section 10: Infrastructure Poverty Analysis
10.1 Infrastructure Score Distribution
The Infrastructure Poverty Score sums access to water (0/1), electricity (0/1), and internet (0/1), ranging from 0 (no utilities) to 3 (all utilities).
Table 10.1: Infrastructure Score Distribution
| Score | Description | Count | Percentage |
|---|
| 0 | No Utilities | 8,776 | 18.35% |
| 1 | One Utility | 782 | 1.64% |
| 2 | Two Utilities | 8,890 | 18.59% |
| 3 | All Three Utilities | 29,370 | 61.42% |
🚨 CRITICAL INFRASTRUCTURE POVERTY:
Nearly 1 in 5 public schools (18.35%) operates without ANY basic utilities. These 8,776 schools face compounded disadvantages:
- No water supply → Cannot operate WASH facilities properly
- No electricity → Limited operational hours, no powered equipment
- No internet → Disconnected from digital learning, administrative systems
These schools require Priority 1 intervention in infrastructure planning.
Section 11: Missing Data Analysis
11.1 Variables with Highest Missing Rates
Table 11.1: Top 10 Variables by Missing Rate
| Variable | Missing % | Missing Count |
|---|
| wash_shs_indiv_func_no_soap | 98.0% | 46,869 |
| wash_shs_indiv_nonfunc | 97.8% | 46,756 |
| wash_shs_group_func_no_soap | 97.7% | 46,714 |
| wash_shs_group_nonfunc | 97.5% | 46,636 |
| wash_jhs_indiv_func_no_soap | 97.3% | 46,540 |
| wash_jhs_group_func_no_soap | 96.9% | 46,345 |
| wash_jhs_indiv_nonfunc | 96.8% | 46,290 |
| wash_jhs_group_nonfunc | 96.0% | 45,921 |
| wash_es_indiv_func_no_soap | 94.9% | 45,382 |
| wash_es_group_func_no_soap | 93.5% | 44,731 |
Figure 11.1: Missing Data Analysis. High missing rates for JHS/SHS variables
reflect schools not offering those levels.
Missing Data Interpretation:
High missing rates for JHS and SHS variables are expected because:
- Only 22.1% of schools offer JHS (10,561 schools)
- Only 16.2% of schools offer SHS (7,743 schools)
Missing values primarily reflect not applicable rather than missing data - schools that don't offer a level cannot report facilities for that level.
Total variables with missing data: 123
Section 12: Distribution Fitting
12.1 Goodness-of-Fit Tests
Table 12.1: Distribution Fitting Results
| Distribution | Variable | K-S Statistic | K-S p-value | Fits? | χ² Statistic | χ² p-value | χ² Fits? | A-D Statistic | A-D Critical (5%) | A-D Fits? |
|---|
| Poisson | total_toilets | 0.449 | <0.001 | No | 643,484 | <0.001 | No | - | - | - |
| Normal | total_toilets | 0.214 | <0.001 | No | - | - | - | 4,305 | 0.787 | No |
| Poisson | wash_functional | 0.448 | <0.001 | No | 1,363,010 | <0.001 | No | - | - | - |
| Normal | wash_functional | 0.264 | <0.001 | No | - | - | - | 5,280 | 0.787 | No |
Conclusion: Data do NOT follow standard Poisson or Normal distributions. The high skewness and overdispersion (variance >> mean) suggest Negative Binomial or Zero-Inflated models are more appropriate for count data modeling.
Section 13: Policy Recommendations
13.1 Immediate Interventions (0-12 months)
Priority 1: Zero-Utility Schools
- Target: 8,776 schools with ZERO utilities
- Deploy emergency water solutions (rainwater collection, tank delivery)
- Solar panel installation for off-grid electricity
- Community connectivity hubs for internet access
- Budget Focus: MOOE + Special infrastructure fund
Priority 2: Gender Parity Gaps
- Target: 1,339 schools with GPI < 0.9
- Audit male/female enrollment ratios
- Add female-specific facilities where warranted
- Ensure MHM compliance per DO 10, s. 2016
Priority 3: Non-Functional WASH
- Target: Schools with >30% non-functional WASH
- Repair/replace broken fixtures
- Establish maintenance protocols
- Train school personnel on preventive maintenance
13.2 Short-Term Goals (1-2 years)
| Goal | Current State | 2-Year Target | Action |
|---|
| Reduce zero-utility schools | 18.4% | <5% | Prioritized infrastructure program |
| Improve internet access | 62.7% | 80% | DICT partnership, DepEd Computerization |
| Reduce Gini coefficient | 0.518 | <0.45 | Needs-based allocation formula |
| Achieve 95% WASH functionality | 91% | 95% | Maintenance fund + training |
13.3 Medium-Term Goals (3-5 years)
-
Universal Minimum Standards:
- All schools meet 50:1 pupil-to-bowl ratio (pending enrollment data integration)
- 100% of schools with water supply
- 100% of schools with electricity
- 90% of schools with internet
-
Reduce Within-Group Inequality:
- Target Theil Index within-group component <0.35
- School-level needs assessment framework
- Performance-based facility grants
-
Monitoring & Accountability:
- Real-time WASH monitoring dashboard
- Annual Three-Star Approach assessment
- Public reporting of school-level indicators
Section 14: Methodology Summary
14.1 Statistical Methods Applied
35+ Techniques Across 8 Categories:
| Category | Methods | Purpose |
|---|
| Distribution Analysis | Shapiro-Wilk, K-S, Anderson-Darling, Chi-Square GOF | Assess normality, fit theoretical distributions |
| Inequality Measures | Gini (Bootstrap CI), Palma (Bootstrap CI), Theil (Decomposition), Atkinson, Hoover, Concentration, Entropy | Quantify distribution fairness |
| Hypothesis Testing | Mann-Whitney, Kruskal-Wallis (η²), Levene's, Wilcoxon, FDR Correction | Test group differences |
| Regression Models | OLS (Robust SE), Quantile, ZIP, Negative Binomial, Hurdle, Interactions | Predict outcomes |
| Clustering | K-Means, Hierarchical (Ward), Silhouette, CH, DB, Cophenetic, Gap | Identify school typologies |
| Dimension Reduction | PCA, Factor Analysis | Identify latent structure |
| Equity Measures | GPI, Rate Ratios, Infrastructure Score | Gender and utility equity |
| Data Quality | Missing Data Analysis, Outlier Detection, Sensitivity | Ensure data integrity |
Appendices
Appendix A: Chart Index
Appendix B: References
Policy Documents:
- DepEd Order No. 10, s. 2016: Policy and Guidelines for Comprehensive Water, Sanitation and Hygiene in Schools (WinS) Program
- DepEd Memorandum No. 194, s. 2018: Implementing Guidelines to DO 10, s. 2016 (Three-Star Approach)
Data Source:
- DepEd National School Building Inventory (NSBI), SY 2023-24
Appendix C: Acronyms
| Acronym | Full Form |
|---|
| WASH | Water, Sanitation, and Hygiene |
| WinS | WASH in Schools |
| GPI | Gender Parity Index |
| ES | Elementary School |
| JHS | Junior High School |
| SHS | Senior High School |
| NSBI | National School Building Inventory |
| MOOE | Maintenance and Other Operating Expenses |
| MHM | Menstrual Hygiene Management |
| PWD | Persons with Disabilities |
| CI | Confidence Interval |
| FDR | False Discovery Rate |
| PCA | Principal Component Analysis |
| ZIP | Zero-Inflated Poisson |
| SDG | Sustainable Development Goal |