Introduction

8 basic indicators for the Iraq MCNA are chosen for comparison with remote sensing data. The list of indicators and their descriptions are provided below.

  • fcs = Food Consumption Score (numeric)
  • household_hunger_scale = Household Hunger Scale (categorical)
  • food_source = Main source of food (categorical)
  • head_seek_work= Head of Household seeking for employment(categorical)
  • how_much_debt = HH debt (numeric)
  • inc_employment_pension = HH income from employment and/or pension (numeric)
  • medical_exp = Medical expense
  • health_issue_chronic = Health issues chronic

Response Rate: The quantity of data for each variable is depicted in the graph below. In the 12839 observations in the Iraq MCNA data, there appear to be no missing values for any of the variables except how much debt However, there are just three missing data in the how much debt column, which is a very small proportion of all the data.

Histograms

Note: As with any statistical investigation, it is advised to use the normally distributed data, hence it is crucial to examine the histograms.

Food Consumption Score

Household Hunger Scale

Debt

Income

Medical expense

NDVI

  • March

  • April

  • May

Distance from water body

  • Using OSM

  • Using XXX

VCI

Satistical test

Food consumption score

Relationship between distance from nearest Water body and FCS

Distance from water source is calculated from two different source -

  • Using OSM data
  • XXXXXXXXXXXXX

The accompanying scatter plot suggests that there may be a relationship between the score for food consumption and the distance to the next water body. Hence In order to verify the hypothesis, a Pearson correlation test was performed.

  • Null hypothesis(H0): There is no correlation between Distance from water body and FCS

According to the results, there is a modest inverse relationship between FCS and distance from the water body, which means that as distance grows, the food consumption score decreases. This also suggests that HH who are located far from a water body should list food as their top necessity.

As priority_needs.food is a binary variable, Hence the following T test has been performed.

Even if the difference is very little, the aforementioned table and boxplot show that there is a significant difference (p is less than.05.) in mean between those who responded “yes” to priority needs.food and those who responded “no”

  • Result:: Households far from water bodies have low food consumption score and there is also evidence that the distance from water bodies are proportional to food needs

Relationship between NDVI and FCS

Interpretation::The food consumption score (high means good) increases for March and April as vegetation health/cover increases, but the trend is the opposite for May-> During FGDs people said that there is less work during winter and spring because of the codl and rain so daily workers could work less

Relationship between VCI and FCS

## `geom_smooth()` using formula 'y ~ x'

Medical Expense

Assumption: Due to chronic diseases brought on by poor water quality, households adjacent to water sources may incur higher medical costs than those distant from water bodies.

The outcome of the aforementioned Pearson correlation test supports the null hypothesis (There is no relationship between the two variables). However, the results indicate a positive association between these two (we were expecting a negative relationship), which also means the farther away from water bodies, the higher the medical costs! The analyst initially finds it incomprehensible, but locals said that it might occur because most of Iraq’s cities are situated along rivers. The rural area doesn’t have a basic medical facility. So when the rural residents became ill, they had to travel to an urban region, which required additional lodging and transit expenses. Based on this supposition, a T-test was run to see if there were any differences between the top demands for a medical facility in urban and rural locations.

Result:

  • There is a significant difference in the mean for those who reported yes to priority_needs.healthcare and those who reported “no”.
  • This means in rural areas (as cities are close to waterbodies), there is a shortage of healthcare centres which results in increasing medical expense

Varifications for findings

    1. Since the findings illustrate that the relationship between FCS/medical costs and distance to the nearest water may be driven by rural-urban differences (urban centres closer to rivers), Hence a Pearson correlation test with the oxford accessibility indicator was performed, and the result aligns with the previous findings.

Logistic regression between

[DO NOT USE THIS TAB! UNDER DEVELOPMENT!!]

    1. A logistic regression is performed between priority_needs.healthcare and Distance (log). The summary of logistic regression is given below
## 
## Call:
## glm(formula = need_priorities.healthcare ~ Distance, family = "binomial", 
##     data = data_logistic)
## 
## Deviance Residuals: 
##    Min      1Q  Median      3Q     Max  
## -1.311  -1.058  -1.030   1.287   1.339  
## 
## Coefficients:
##                 Estimate   Std. Error z value            Pr(>|z|)    
## (Intercept) -0.375179996  0.026106488  -14.37 <0.0000000000000002 ***
## Distance     0.000023852  0.000002843    8.39 <0.0000000000000002 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 15748  on 11462  degrees of freedom
## Residual deviance: 15677  on 11461  degrees of freedom
## AIC: 15681
## 
## Number of Fisher Scoring iterations: 4

Regression plot

## `geom_smooth()` using formula 'y ~ x'