PENERAPAN ANALISIS DISKRIMINAN UNTUK MENGKAJI PERBEDAAN KONSUMSI PANGAN HEWANI DI WILAYAH PERKOTAAN DAN PERDESAAN DI INDONESIA

Abstract

This research aims to examine the differences between animal food consumption in urban and rural areas in Indonesia using discriminant analysis. The data used are secondary data obtained from the National socio-economic Survey (Susenas) held Central Bureau of statistics of the year 2015. The data analyzed includes data on average consumption of protein (grams) per capita a day according to the dining area and group housing in 34 provinces in Indonesia. Animal food consumption are examined, namely fish consumption, the consumption of meat, eggs and milk consumption. Based on the results of the analysis it can be concluded that the normality test,  matrix test  of  the  variety  of  uniform  and   test  the   average   vector  value is fulfilled, so the discriminant function is obtained as follows: where is consumption of fish, is the consumption of meat, eggs and milk consumption was and is the value of the discriminant score. The result of the discriminant functions that form validation indicates that the function is accurate and stable. The precision of the discriminant function was 85.08% or there are 2 errors in classifying a group of villages and 8 errors in classifying a group of 67 regions of cities and villages throughout the province in Indonesia. Based on the magnitude of the coefficients of the function, that the consumption of eggs and milk that contributes the largest animal food consumption differences across cities and villages.

Key words : animal food consumption, discriminant analysis, test validation, assumptions