In order to collect assessment data to inform the child protection response to the emergency, analysis will focus on communities (defined as camps and nearby villages IDPs have relocated to) as the unit of measurement. While households would also be a valuable unit of measurement, given limited human resources, it will be more efficient to focus on communities.

From the sampling frame of 150 camps and few nearby villages, the evaluation team will sample sites using the criteria of accessibly severely affected vs. less severely affected camps and villages. Areas targeted will be rural vs. urban communities and potentially distinct ethnic/racial groups. Three camps and 2-3 nearby villages will be selected from each site using purposive sampling. Prioritizing accessible affected areas based on secondary sources of information and experience is preferable to attempting to assess non-affected areas due to limited resources. Purposive sampling will be used due to resource and logistical constraints, and can be adjusted if needed. Random sampling is too costly and time consuming in this situation.

The primary objectives for the assessment include determining patterns of separation; types of care arrangements for separated children; capacities in community to respond to child separation; patterns and levels of institutionalization of children; nature and extent of any hazards for children in the environment and risk to their safety as a result; types and levels of violence towards children in the community; causes and level of risk of death and/or Injury to children; risks and available responses to sexual or gender based violence for children; existing patterns and scale of child labor; likely new risks as a result of the emergency; sources of stress and available coping mechanisms for children and their caregivers; and capacities for provision of people/resources at community level to provide support for children.

It will be more effective to use a smaller sample and collect more thorough information about as many objectives as possible rather than having a large sample size. Even though this would involve a trade off for high validity, purposive sampling will provide an approximate sense of the scale and priorities necessary to consider in planning the protection response. Potentially different impacts of the emergency among the categories of affected groups in the sample may also be brought to light.

Given the limited number of human resources available (10 social workers), an analysis will be conducted using a combination of key informant interviews, direct observation, and a desk review. The key informant interviews, while less thorough an interview technique than a focus group, will provide information that is “good enough” despite the lack of expertise of the interviewers. Key informant interviews will take place with 3-4 persons with extensive experience with children from each community sampled (teachers, medical personnel, community leaders), and will be triangulated with direct observation and a desk review of materials (to confirm what is already known about the area, community, and capacities of the region).

To prepare the social workers to be an effective assessor team, training must be provided. Ideally, the team would be trained as part of preparation activities in anticipation of a crisis/disaster with at least 1 month of basic training required as well as periodic refresher training. However, given the time constraints during a complex emergency, 2-3 days of classroom training and 2-3 days of supervised interviewing in the field should be adequate.

The data analysis will be led by the team supervisor – an expert in child protection – with input from the assessors. The data will be analyzed using descriptive analysis, which is considered best when using purposive sampling. Specifically, frequency analysis and cross tabulation analysis would be employed.

The rapid assessment team is responsible for collecting the data. The specific questions to be answered by the data should be determined by the supervisor and guided by the principal of “What We Need to Know”. The purpose of the assessment is to provide a snapshot of urgent child protection related needs among the affected population, and therefore data is collected not simply for the sake of collecting data but rather with the understanding that it will clarify and prioritize child protection activities.

Data will be collected by the assessors under the supervision of the team supervisor and then cleaned daily through debriefing sessions with the team-leader and assessors. The next stage of data cleaning will occur during data entry, ideally done in parallel with data collection, which will not only save time but will allow data entry personnel to ask timely questions to the data collection team. Also, if possible, data should be entered into a computer database to facilitate data manipulation, analysis and sharing.

The supervisor and team leaders will have a daily plan and goals for the assessment. The morning meeting will include all team members to explain goals for the day and discuss issues that arose from data collected the previous day. In the evening, supervisors and team leaders will review and clean the data before sending it to data entry personnel. The supervisor and team leaders will then discuss any issues that arose during the day, any adjustments needed to the goals and overall direction of the assessment, and will analyze the data using triangulation techniques.

In order to facilitate sharing of information with other clusters operating in the emergency theater, a mini-workshop with other child protection actors and key partners will organized to disseminate the major findings and their implications. This will allow other organizations quick access to data that might inform their activities. Afterwards detailed reports – including an executive summary, methodology used, key findings, identification of priorities, and key recommendations – should be made available to Child Protection Working Group (CPWG) members. Raw data should be made available to CPGW members as well as to members of other clusters.

The length of the assessment is somewhat dependent on the goals of the assessment as well as the situation on the ground. Since this is meant to be a rapid assessment, it should ideally be completed within 2-3 weeks. However, this time frame will also depend on the evolving nature of the crisis, including access to crisis affected areas, security considerations, etc.

There will be periodic real time analysis of the data as it is entered into the database. This is important so that early trends can be detected and acted upon. The mini-workshop held to disseminate results will occur within one week of the end of the assessment so as to disseminate any clear and present trends that require immediate attention. A more thorough analysis will be completed within one month of the end of the assessment.

The assessment methodology will be focused on triangulation, the process by which to confirm whether or not the data collected is accurate and falls within the “good enough” threshold. This initial triangulation of data should occur during the daily team debriefs. Triangulation will also be used once all the data has been aggregated as well. Once the data has been “validated”, descriptive analysis methods such as frequency and cross tabular analysis should be employed.

Two survey options were considered and discarded for not being ideal in this situation were directly interviewing of children and use of PRM focus groups. Given the limited experience of the social workers in working with children in emergency situations and from different areas of the country, the assessment leadership deemed it might be more harmful to (rather than beneficial to) directly interview children about threats to their security and well being. Instead, interviews will take place with adults that have extensive experience in working with the children within their communities, allowing the assessment to gather the information needed while still working within the “do no harm” principal. Secondly, while the PRM focus group methodology would allow a much richer understanding of the threats to and priorities of each community interviewed, the level of expertise and time required to properly do such an exercise was simply unrealistic given the human resource and time constraints faced by the assessment team.