Sample size and sampling design
The sample size was estimated based on the expected prevalence of exclusive breastfeeding rates in the district at each time point. In the pre-intervention stage with assumed prevalence of EBF of 40% and absolute precision of 5% and a design effect of 1.2 (for clustering), the sample size calculated was 427. The corresponding sample sizes for the 1st and 2nd post intervention stages were 447 and 440 respectively, assuming the same design effect and expected prevalence of EBF to be 50% and 55% in the two time periods.
At the first stage, panchayats were selected from each block using a PPS sampling technique. Thus, in all the 6 blocks were selected for collecting information from mothers. At the second stage, mothers who have delivered a child within 0–3 months, 3–6, 6–12 and 12–24 months were selected from the sample villages. Thus, a total sample of size 421 (base line), 480 (second phase) and 597 (third phase) mothers with children between zero to twenty four months were selected from the 6 blocks and interviewed to assess the initiation of breastfeeding and complementary feeding practices in the area. In case sample was not achieved in any particular village, adjoining village was considered to achieve the required sample.