Stringent and detailed methodology to avoid ambiguity Siddharudha Shivalli, Ynepoya Medical College, Yenepoya University, Mangaluru-575018, Karnataka, India 14 October 2015 I read this article with curiosity. Authors’ efforts are commendable. It reiterates malaria as a major public health problem in the study area. However, following issues need to be addressed. Authors have used 7 open ended questions to assess the malaria knowledge among the study participants. The participants’ knowledge level was categorized as good (responded well to all questions), adequate (responded to at least five questions) and poor (could not answer >3 questions). However, ambiguity persists when categorizing the responses to questions 4, 5 and 6. Authors should have mentioned the minimum expected response to each question to consider the participant as 'knowledgeable'. For instance, mentioning at least 3 measures to avoid malaria for question 4 was regarded as knowledgeable. Furthermore, such categorization should be done by 2 people independently and should sought 3rd opinion in case of disagreement in order to avoid inter observer bias. However, authors have not mentioned about who did the categorization and how the bias was tackled. Data regarding correct responses to all the 7 questions would have helped to fix the priority while designing health education material. I would have also considered the ‘knowledge of availability of free diagnosis and treatment in government healthcare setup’ while assessing the knowledge component. In addition, treatment seeking pattern (preferred source and type of treatment) would have been more interesting. Inclusion of 'body complexion, odour, and clothing [Table 1] as study variables need to be justified. Is it related to personal hygiene and/or vector bionomics? If so, their association with occurrence of malaria should have been mentioned. Definition for house type (bamboo, kacha and pucca) should have been explicitly mentioned in the method to avoid ambiguity. Use of existing guidelines to define the economic status would have been more apt.Use of ‘socio-economic determinants’ in the title is debatable as the study was cross sectional and multivariate analysis was not done. None the less, I must congratulate the authors for investigating an important public health problem in the study area. Competing interests The author declares that there is no conflict of interest about this publication.