Food security


Variables in analysis


dependent variables


HDDs



Descriptive statistics


by country










by gender



by income






by education level






Meet Food Needs


Descriptive statistics


by country










by gender



by income






by education level






Lacking

Lacking variables (Q4: lacking_cereal, lacking_veg, lacking_fruit, lacking_leg, lacking_fish, lacking_oil) has been summarised in a single variable (Q4.lacking) with a confirmatory factor analysis




Descriptive statistics


by country










by gender



by income






by education level






Independent variables


Computed explanatory variables


descriptive statistics


Univariate analysis of the relationship between dependent variables and explanatory variables




Regression models


project level


country level


HDDs



Q16.meet_food_needs


Q4.lacking


Regression models food security - experiments


project level


Risk


Time


PGG


Trust


HDDs - experiments country level


Risk


Time


PGG


Trust


DQQ


Food Group Diversity Score (FGDS) and Minimum Dietary Diversity for Women of Reproductive Age (MDD-W)



MDD-W: if FGDS ge 5 than Yes otherwise No



descriptive statistics


by country




by income



by education level



Univariate analysis of the relationship between MDD.W and explanatory variables



logistic model


project level


with all explanatory variables


stepwise model selection (BIC criterion)


country level


MDD.W - experiments


project level


country level


risk


Time


PGG


Trust


NCD-Protect score, NCD-Risk score, Global Dietary Recommendations (GDR) score


GDR score = NCD-Protect - NCD-Risk + 9



descriptive statistics


by country






by gender





by income







by education level







Univariate analysis of the relationship between GDR and explanatory variables



regression model


project level


with all explanatory variables


stepwise model selection (BIC criterion)


country level


GDR - experiments


project level


country level


risk


Time


PGG


Trust


Innovation

innovation as mix between Q6 local food interest & Q13 nutri food interest

The propensity to introduce the consumption of innovative products is calculated as the average of Q6.local_food_ _interest and Q13.nutri_food_interest

Reasons for including new food products (local or nutrient-rich) were grouped using a factor analysis, as well as obstacles.





descriptive statistics


by country





by gender




by income




by education



Univariate analysis of the relationship between Propensity to introduce new food products and explanatory variables



regression model

project level


with all explanatory variables



stepwise model selection (BIC criterion)



country level



Regression models innovation - experiments


project level



country level


risk


Time


PGG


Trust


Experiments


Risk


Univariate analysis




regression analysis


feature selection (Boruta algorithm)


KE

## 
## Call:
##  randomForest(formula = frml, data = datB, importance = T) 
##                Type of random forest: regression
##                      Number of trees: 500
## No. of variables tried at each split: 5
## 
##           Mean of squared residuals: 5.174913
##                     % Var explained: 44.79
##                   variable    IncMSE IncNodePurity
## 1       Q2.7.imp_tradition 21.800549     174.92474
## 2           Q2.8.imp_local 18.072497     188.89968
## 3        Q2.3.imp_provider 16.616380     169.12884
## 4         Q2.9.imp_product 16.560927     157.19317
## 5          Q11.healty_diet 14.292595     115.43544
## 6           Q3.income_food 14.252738     164.75299
## 7   Q6.local_food_interest 14.154527      79.89880
## 8       Q2.5.imp_envfriend 13.926696     127.74895
## 9            Q5.local_food 10.561493      70.84496
## 10      Q2.2.imp_nutrition  9.682119      87.39528
## 11 Q13.nutri_food_interest  9.196815      52.70723
## 12         Q2.6.imp_divers  8.855920      82.63194
## 13              Q4.lacking  8.846952     219.87668
## 14         Q19.4.hh_salary  8.451514     102.27890
## 15   Q2.1.imp_availability  4.447112      43.02512
## 16             Q19.HH_size  3.879172     113.06789
## 17          Q2.4.imp_price  2.985620      42.54572
## [1] "1 Q2.7.imp_tradition"
## [1] "2 Q2.8.imp_local"
## [1] "3 Q2.3.imp_provider"
## [1] "4 Q2.9.imp_product"
## [1] "5 Q11.healty_diet"
## [1] "6 Q3.income_food"


MO

## 
## Call:
##  randomForest(formula = frml, data = datB, importance = T) 
##                Type of random forest: regression
##                      Number of trees: 500
## No. of variables tried at each split: 2
## 
##           Mean of squared residuals: 3.07358
##                     % Var explained: 3.92
##                  variable   IncMSE IncNodePurity
## 1            session.code 8.066338     381.78131
## 2                     GDR 6.396866     146.67868
## 3          Q2.8.imp_local 4.289426      73.00984
## 4              Q4.lacking 4.153730     183.62809
## 5 Q13.nutri_food_interest 4.110316      84.66416
## 6         Q2.6.imp_divers 3.978559      69.76876
## [1] "1 session.code"
## [1] "2 GDR"
## [1] "3 Q2.8.imp_local"
## [1] "4 Q4.lacking"
## [1] "5 Q13.nutri_food_interest"
## [1] "6 Q2.6.imp_divers"