Praktik pemberian makan bayi dan anak yang sesuai diberikan pada anak usia 6-23 bulan sangat penting untuk pertumbuhannya dan dapat mencegah stunting. Adanya variasi yang tinggi antar provinsi mengenai nutrisi menyebabkan pentingnya menganalisis berdasarkan variabel kontekstual pada level provinsi. Tujuan dari penelitian ini adalah mempelajari pengaruh variabel kontekstual dan variabel otonomi perempuan dan mempelajari efek perbedaan dari dua variabel tersebut terhadap praktik pemberian makan bayi dan anak pada tiga indikator utama yaitu minimum dietary diversity, minimum meal frequency, dan minimum acceptable diet. Menggunakan analisis regresi logistik multilevel dengan data Survei Demografi dan Kesehatan Indonesia (SDKI 2017) menunjukkan bahwa variabel kontekstual yang berasosiasi dengan praktik pemberian makan bayi dan anak adalah persentase penduduk miskin dan persentase fasilitas kesehatan. Sedangkan pada level individu beberapa indikator variabel otonomi perempuan (economic decision making, access to control dan freedom from threat) dan beberapa variabel kontrol (umur ibu, klasifikasi tempat tinggal, akses ibu terhadap internet, umur anak dan pendidikan suami) memiliki asosiasi terhadap praktik pemberian makan bayi dan anak pada 3 indikator yang digunakan. Variabel interaksi menunjukkan adanya efek perbedaan pengaruh variabel otonomi perempuan dan variabel kontekstual terhadap praktik pemberian makan bayi dan anak.

---

Appropriate infant and young child feeding practices given to children aged 6-23 months are very important for their growth and can prevent stunting. The high variation between provinces regarding nutrition makes it important to analyze based on contextual variables at the provincial level. The purpose of this study was to study the effect of contextual variables and women's autonomy variables and to study the different effects of these two variables on infant and child feeding practices on the main indicators, namely minimum dietary diversity, minimum meal frequency, and minimum acceptable diet. Using multilevel logistic regression analysis with data from the Indonesian Demographic and Health Survey (IDHS 2017) shows that the contextual variables associated with infant and child feeding practices are the percentage of poor people and the percentage of working women. While at the individual level several indicators of women's autonomy variables (economic decision making, access to control and freedom from threat) and several control variables (mother's age, classification of residence, mother's access to the internet, age of children and husband's education) have associations with the practice of giving feeding infants and children on the 3 indicators used.Interaction variables indicate that there are differences in the effects of women's autonomy variables and contextual variables on infant and young child feeding practices.