Food Consumption Patterns: Which Groups Really Eat Differently?

A district nutrition survey collected monthly per-capita food consumption across 4 demographic groups and 10 food groups. The headline: "Urban households consume 22% more than rural." But when the analyst breaks it down, the gap is driven almost entirely by 2 luxury food groups — the other 8 show minimal difference. She must manually identify which food groups drive the apparent gap before the report misleads policy.

# Aplly.xyz Case Study Submission

## Title
Food Consumption Patterns: Which Groups Really Eat Differently?

## Type
Data Analytics

## Difficulty
Intermediate

## Estimated Time
75 minutes

## Overview
A district nutrition survey collected monthly per-capita food consumption across 4 demographic groups and 10 food groups. The headline: "Urban households consume 22% more than rural." But when the analyst breaks it down, the gap is driven almost entirely by 2 luxury food groups — the other 8 show minimal difference. She must manually identify which food groups drive the apparent gap before the report misleads policy.

## Case Details

Function Focus: Multi-group consumption comparison, gap decomposition by food group, identifying which categories drive a headline difference, demographic profiling

Scenario:
The district nutrition survey report has a draft headline: "Urban households consume 22% more food per capita than rural households." The data shows total monthly consumption of 22.4 kg per capita for urban vs 18.4 kg for rural. But the analyst suspects the gap is not uniform — it might be driven entirely by 2 expensive food groups (milk and meat/fish) that urban households can afford. If you remove those, the gap may nearly disappear. She must manually decompose the 4-percentage-point gap by food group and demographic segment to determine whether the headline is truthful or misleading.

Dataset Structure:
- 4 demographic groups: Urban High Income, Urban Low Income, Rural High Income, Rural Low Income
- 10 food groups with monthly per-capita consumption (kg or liters)
- Total consumption per group
- Urban average and Rural average for comparison

Tasks:
1. Compute the total monthly per-capita consumption for each of the 4 demographic groups by summing all 10 food groups — confirm the Urban (22.4 kg) and Rural (18.4 kg) averages and the 22% gap
2. For each food group, compute the urban-rural consumption difference in absolute terms (kg) and as a percentage of the total gap — identify the top 2 food groups that together explain at least 70% of the total gap
3. Compare Urban Low Income vs Rural High Income (the two middle groups) — is the gap still 22% when income is controlled? Compute the difference and state whether the gap is driven by income or by urban/rural location
4. Identify the food group where all 4 groups consume nearly identical amounts (coefficient of variation < 10%) — this is the "staple" that income and location don't affect
5. Only after submitting your manual analysis, re-run the decomposition in a spreadsheet or AI tool and report any discrepancies — then recommend whether the headline should say "22% gap" or something more nuanced

Expected Output:
An analytical memo containing: (a) total consumption table for all 4 groups confirming the headline, (b) per-food-group gap decomposition identifying the top 2 drivers, (c) income-controlled comparison (Urban Low vs Rural High) with a corrected gap, (d) identification of the constant food group across demographics, (e) a headline recommendation, and (f) a round-trip discrepancy note.

Evaluation Criteria:
Correct total consumption computation and gap confirmation, accurate per-food-group gap decomposition with correct percentage attribution, a meaningful income-controlled comparison that isolates the real driver of the gap, correct identification of the invariant food group, and a well-justified headline recommendation.

## Data Sources

Monthly Per-Capita Food Consumption (kg or liters):
| Food Group | Urban High Income | Urban Low Income | Rural High Income | Rural Low Income |
|---|---|---|---|---|
| Cereals (rice/wheat) | 8.2 | 8.8 | 9.5 | 10.0 |
| Pulses | 1.8 | 1.6 | 1.7 | 1.5 |
| Vegetables | 4.5 | 4.0 | 3.8 | 3.5 |
| Fruits | 2.2 | 1.2 | 1.0 | 0.6 |
| Milk & products (liters) | 5.5 | 3.0 | 2.5 | 1.8 |
| Meat & Fish | 2.8 | 1.2 | 0.8 | 0.5 |
| Eggs | 0.6 | 0.4 | 0.3 | 0.2 |
| Oils & Fats | 1.2 | 1.0 | 1.1 | 0.9 |
| Sugar & Jaggery | 1.0 | 0.9 | 0.8 | 0.7 |
| Spices & Condiments | 0.4 | 0.3 | 0.3 | 0.2 |

Group Sizes (survey sample):
| Group | Households Sampled |
|---|---|
| Urban High Income | 250 |
| Urban Low Income | 250 |
| Rural High Income | 250 |
| Rural Low Income | 250 |

Context:
- Draft headline: "Urban households consume 22% more food per capita than rural"
- RDA recommends minimum 14.5 kg per capita per month for a balanced diet
- All 4 groups meet the minimum, but quality composition differs
- Policy question: should government focus on total consumption or diet quality?

## Solution Frameworks
Multi-group comparison, gap decomposition, demographic profiling, income-controlled analysis, headline accuracy auditing

## Solver Guidance & Tutorials
Link to: "Decomposing Group Differences — Which Categories Drive the Gap?" tutorial

## What You'll Learn
- Decomposing a headline gap into its component drivers
- Controlling for income when comparing urban vs rural populations
- Identifying staple foods that are invariant across demographics
- Determining whether a reported difference is meaningful or misleading

## Tags
food consumption, demographic analysis, gap decomposition, nutrition survey, urban-rural comparison

## Registration Links
- Register as Solver
- Register as Evaluator

Monthly Per-Capita Food Consumption (kg or liters):
| Food Group | Urban High Income | Urban Low Income | Rural High Income | Rural Low Income |
|---|---|---|---|---|
| Cereals (rice/wheat) | 8.2 | 8.8 | 9.5 | 10.0 |
| Pulses | 1.8 | 1.6 | 1.7 | 1.5 |
| Vegetables | 4.5 | 4.0 | 3.8 | 3.5 |
| Fruits | 2.2 | 1.2 | 1.0 | 0.6 |
| Milk & products (liters) | 5.5 | 3.0 | 2.5 | 1.8 |
| Meat & Fish | 2.8 | 1.2 | 0.8 | 0.5 |
| Eggs | 0.6 | 0.4 | 0.3 | 0.2 |
| Oils & Fats | 1.2 | 1.0 | 1.1 | 0.9 |
| Sugar & Jaggery | 1.0 | 0.9 | 0.8 | 0.7 |
| Spices & Condiments | 0.4 | 0.3 | 0.3 | 0.2 |

Group Sizes (survey sample):
| Group | Households Sampled |
|---|---|
| Urban High Income | 250 |
| Urban Low Income | 250 |
| Rural High Income | 250 |
| Rural Low Income | 250 |

Context:
- Draft headline: "Urban households consume 22% more food per capita than rural"
- RDA recommends minimum 14.5 kg per capita per month for a balanced diet
- All 4 groups meet the minimum, but quality composition differs
- Policy question: should government focus on total consumption or diet quality?

Multi-group comparison, gap decomposition, demographic profiling, income-controlled analysis, headline accuracy auditing