Behavioural change communication interventions have been shown to be effective at improving infant and young child nutrition knowledge and practices. However, evidence in humanitarian response contexts is scarce. Using data on secondary outcomes of breastfeeding, water treatment, and knowledge from a cluster randomized control trial of the Yemen Cash for Nutrition programme's impact on child nutritional status, this paper shows that the programme significantly improved knowledge and practices for poor women with young children in the pilot districts. The intervention consisted of cash transfers and monthly group nutrition education sessions led by locally recruited community health volunteers. Data are based on self-reports by participants. Estimating impacts among all 1,945 women in 190 clusters randomly assigned to treatment versus control and controlling for baseline levels and community characteristic and adjusting for noncompliance with randomization, the programme increased the probability of breastfeeding initiation within the first hour after delivery by 15.6% points (p < .05; control = 74.4% and treatment = 83.6%), the probability of exclusive breastfeeding during the first 6 months by 14.4% points (control = 13.5% and treatment = 25.3%), the probability of households treating water consumed by adults by 16.7% points (p < .01; control = 13.9% and treatment = 23.4%), and treating water consumed by children under two by 10.3% points (p < .10; control = 31.2% and treatment = 37.9%). Impacts on knowledge and breastfeeding are similar for both literate and illiterate women, and water treatment impacts are significantly larger for literate women. This study was registered at 3ie (RIDIE-STUDY-ID-5b4eff881b29a) and funded by the Nordic Trust Fund and Consultative Group on International Agricultural Research programme on Policies, Institutions, and Markets.

中文翻译：

---

在人道主义背景下进行营养培训：来自一项整群随机试验的证据。

行为改变交流干预措施已被证明可有效改善婴幼儿营养知识和实践。但是，在人道主义对策方面缺乏证据。利用也门现金换营养计划对儿童营养状况的影响的一项随机对照试验的母乳喂养，水处理和知识的次级结果的数据，本文显示该计划显着改善了患有贫困儿童的贫困妇女的知识和实践。试点地区。干预包括现金转移和每月由当地招募的社区卫生志愿者主持的团体营养教育课程。数据基于参与者的自我报告。估算所有1中的影响 在190个群集中的945名妇女被随机分配至治疗组和对照组，并控制基线水平和社区特征，并根据随机分组调整不依从性，该计划使分娩后第一小时内开始母乳喂养的可能性提高了15.6％（p <.05；对照= 74.4％，治疗= 83.6％），头6个月纯母乳喂养的可能性降低了14.4％（对照= 13.5％，治疗= 25.3％），家庭处理成人饮水的概率降低了16.7％ （p <.01；对照组= 13.9％，治疗= 23.4％），将两岁以下儿童所消耗的水处理了10.3％（p <.10；对照组= 31.2％，治疗= 37.9％）。识字和不识字的妇女对知识和母乳喂养的影响相似，和水处理对有文化的妇女的影响要大得多。该研究已在3ie（RIDIE-STUDY-ID-5b4eff881b29a）上注册，并由北欧信托基金和国际农业研究政策，机构和市场咨询小组资助。