Needs assessment for refugee emergencies (NARE)

Underlying process – how does this work?

A NARE analyses pre-crisis and post-crisis secondary data and suggests what information should be gathered during primary data collection, through visits, observations, interviews, and focus group discussions. It occurs at one point in time and is not a continuous monitoring system.

UNHCR’s role and accountabilities

In an emergency, the UNHCR Representative in a country is responsible for leading and coordinating an overall refugee needs assessment, through sectoral coordination. In an appropriate inter-agency forum, the Representative should request operational partners to nominate staff to join a needs assessment team (see below). A needs assessment coordinator should be identified in UNHCR to contact sector or organizational focal points and follow up. Sectoral leads and emergency coordinators need to decide what information the operational strategy should receive from a needs assessment. An information management officer helps with assessment methodology, design, data collection, analysis and coordination. UNHCR should ensure that the assessment is coordinated with other data collection activities.

The following principles of NARE should be taken into account in all refugee emergency situations:

• A secondary data review should always take place, to determine what information is already available.
• Initial needs assessments should be multi-sectoral and coordinated across sectors. (Detailed sector-specific needs assessments can follow the initial multi-sectoral assessment.)
• Different types of information are best gathered by different data collection techniques, such as facility visits, observations, interviews with key informants and community/focus group discussions.
• Assessments should always be coordinated to ensure data collection activities produce comparable data, are not redundant and generate a shared vision of the needs and situation.

For the initial needs assessment in a new refugee emergency, UNHCR should lead a ‘joint' multi-sectoral needs assessment that involves a range of sectors and humanitarian actors. Joint assessments help establish a common understanding of the situation and make good use of available resources. At the same time, all assessment activities should be harmonized, to ensure comparability of the data and the most efficient use of resources.

Considerations for UNHCR’s engagement at country level

Needs assessments are often situation dependent and many factors may influence their design. They include the level of humanitarian access, whether population movements are stable or dynamic, the amount of time and resources available for the assessment, and the types of interventions that the assessment should inform.

Emergency Needs Assessment Process:
Below are listed the basic steps for designing and conducting an emergency needs assessment. Note that, while they are broadly in order, many steps will run in parallel.

• Identify inter-agency stakeholders and determine coordination arrangements.
• Identify the NARE's purpose, and the kinds of decisions that require needs assessment information. Detail the information needs in a data analysis plan.
• Conduct a secondary data review, compile an assessment registry, a 3W (who, what, when), and identify remaining information gaps.
• Decide if primary data collection is required. If yes, design a methodology for primary data collection, including how communities will be sampled. Calculate the time and resources that will be required (including equipment, translators, data collectors, and data entry personnel).
• Design and test data collection forms. Translate if required.
• Train the data collection team(s).
• Undertake primary data collection.
• Collate, clean and analyse the information produced.
• Undertake joint analysis to identify key findings, priorities and recommendations.
• Disseminate data and information products.
• Begin monitoring.

The general principles described in this entry apply to any type of emergency needs assessment. However, carrying out a needs assessment in urban areas can present particular challenges, because it is difficult to demarcate the geographic areas and populations that need to be surveyed. Remarks relevant to urban assessments can be found throughout this Entry.

Step 1.Who to involve
In the first phase of a refugee emergency, a wide range of humanitarian agencies, NGOs, government counterparts and possibly donors may be involved in the initial assessment planning process. Actors responsible for overseeing and providing assistance in the first phase of a response should be included within the assessment process. It is critical to involve WFP and other relevant partners when the population to be assessed is larger than 5,000 persons.

In urban contexts, identifying key refugee stakeholders is likely to require additional effort. Urban refugees are often mobile and tend to cluster in the poorest areas of the city, or to disperse across many locations; as a result, they are difficult to identify. Refugee focal points can help the team to identify where refugees live, estimate the population and its demography, locate new points of influx, vulnerable groups, infrastructure and security issues, and identify coping mechanisms. Needs too may change rapidly in an urban context. Ask for help from the protection or community-based protection officer to compile the contact addresses of refugee focal points.

The Assessment Team and Refugee Information Management Working Group (RIM WG)
An assessment team is a temporary working group of stakeholders who manage a needs assessment. Members of this team include operational leads who decide what information is required for decision-making, interpret the results of the assessment, and design interventions based on those results; and an information manager, who designs the assessment, selects the sites that will be assessed, compiles the data, and produces final information products. The duties of the assessment team are to:

• Agree on a time frame, a methodology for data collection, a process for cleaning and analysing the results, and how the results will be shared and disseminated.
• Identify an emergency referral system to deal with urgent interventions that need to be made that are uncovered during the assessment process.
• Agree on minimum life-saving sectoral data.
• Finalize the assessment method and design.
• Select sites.
• Coordinate resources, training and logistics.

The assessment team will need to identify what resources will be required. If resources are limited, this may affect the assessment's design. The resources spent on a needs assessment should never exceed 10% of the value of interventions that will be made as a result; ideally, the cost of an assessment should be much lower.
The information manager may be tasked to: compile an assessment registry and do a 3W (Who, what, when - see below); put together and coordinate needs assessment methodologies; make sure that data sets are compatible; and facilitate data sharing in the context of the Refugee Information Management Work Group (RIM WG). The RIM WG is a forum for coordinating multiple assessments, sharing and tracking available data, and keeping the assessment registry up-to-date. Needs assessment coordination functions should be included in the RIM WG terms of reference: see ‘Policy documents, references and tools' below.

Step 2. What to assess
Identify clear and precise objectives and a focus for the needs assessment. Based on the NARE checklist, detail the information needs. The NARE checklist is a multi-sectoral needs assessment list which can be used to develop a country-specific assessment adapted to the local situation. NARE users are not obliged to complete the checklist but can pick and choose information needs depending on the resources available, time, the purpose of the assessment, and the types of interventions that are to be made. List the information needs in order of importance, based on their operational importance and the resources required to collect the data. List the information needs, accompanying indicators and possible information sources in a data analysis plan. (You can see an example of a data analysis template at: https://www.dropbox.com/s/xuhhd346j0rx5kj/Data%20Analysis%20Plan%20-%20Template.xlsx?dl=0.)

Expected outcomes
At the start of a needs assessment process, ask several questions. What are the underlying causes of risk and vulnerability? Have these causes changed and, if so, how? How widespread is the problem (throughout the country or countries, or in specific areas)? How are host communities coping? Which geographical area is most affected, and which affected groups are the most at risk? How do we expect the situation to develop in the next three months? A crucial first step is to agree what operational information will be most relevant to intervention and programme planners.

Note. Do not start a needs assessment process by designing a data-collection form. Begin by identifying the questions that need to be answered, not the ones that will be asked. If you start with the form, your assessment is much less likely to yield actionable information.

Step 3. Secondary data review
The secondary data review provides valuable information. Led by protection officers and information managers, it should consider a range of issues, including:

• Conditions before the crisis (and whether they aggravated the impact of the crisis).
• Underlying vulnerabilities. Which groups were already at risk before the crisis?
• Threats (epidemics, climate, etc.).
• Current conditions (what is already known about the scope and scale of the crisis and the humanitarian consequences)?
• Lessons that can be learned from past crises or refugee operations in the same area.

Secondary data are any information that originates from outside the needs assessment, such as data from the Government, monitoring data, etc. It includes data that are owned by UNHCR, such as registration or ProGres data. Primary data are any kind of time-bound information that is collected by completing an assessment form during an emergency needs assessment. In addition to information from humanitarian partners already in situ, secondary data may be collected from or found through ReliefWeb, Alertnet, ACAPS, OCHA, HumanitarianResponse.info, media, blogs, crowd-sourcing, coordination meeting minutes, census data, etc.

Compiling an assessment registry
To help understand the overall situational context, the information manager should ensure that an ‘assessment registry' is compiled. (You can see an assessment registry template at https://www.dropbox.com/s/6wr85tmvwo4vn6l/1.4%20Assessment% 20Registry_Template_v2.xlsx?dl=0.) This lists the coverage of current and planned assessments and helps the assessment team to decide where to focus subsequent assessment activities. Country operations that have the UNHCR web portal may be able to register needs assessments in an online ‘needs assessment registry' rather than use the Excel template.

Compiling a 3W
To obtain an overview of needs, it is important to understand and map existing services, by type and location, and the capacities of the Government and operational partners, in a detailed 3W (Who, what, when?). As part of the secondary data review, a protection and programme officer should work with the information manager to compile or expand a 3W for the operation. This can be done through a traditional 3W spreadsheet, which may be used to establish a profile of each administrative unit, highlight service coverage (or its absence), and signal areas that should be included or prioritized in the needs assessment.

An urban or rural dispersed needs assessment will need to break the area into smaller geographic units, by agreement with local or national authorities and other partners, after discussion in the RIM WG and other coordination forums. To avoid confusion, units should be based on existing formal or informal boundaries (i.e., municipal administrative units or local neighbourhoods), at least to the extent possible; where no sub-divisions exist, a new system will need to be created.

The division into smaller units should be completed by the information manager and the protection and programme officer and cross-checked with focal points in the community. Once agreed, the names of districts and neighbourhoods should be shared widely with Government and operational partners and used when reports are written and services coordinated.

For additional information on compiling a 3W, see the Entry on the emergency information management strategy. To compile a 3W city map and set up a geo-referenced database, see the section on mapping in the Emergency Information Management Toolkit (Section 15, Annex 5, at http://data.unhcr.org/imtoolkit/ chapters/view/mapping/lang:eng).

Situational analysis
The output of a secondary data review, including the assessment registry and 3W, should be a short report that lists secondary data sources and major findings. The analysis should identify information gaps in the secondary data that a primary data collection should fill.

Step 4: Primary data collection?
Although a needs assessment should always be undertaken when there is a refugee emergency, in some situations primary data should not be collected. Do not collect primary data:

• When collecting data will put collectors or interviewees in danger.
• When the results of an assessment will be incorrect or biased (for example, because interviewees are too afraid to tell the truth or external factors will distort data collection).
• When a population feels over-assessed and may be hostile to additional needs assessments.
• When programmatic decisions have already been made, and the assessment results will have no operational relevance.

In the above situations, it may be appropriate to assess needs by means of a secondary data review only (without collecting primary data).

Emergency referral system
Before collecting primary data, protection staff should arrange an emergency referrals system for the assessment team, by sector, with focal points identified. When field data collection teams come across life-saving issues or cases requiring urgent intervention (a collective centre in need of a water, sanitation and hygiene (WASH) intervention, an individual who needs urgent protection), they should know how to refer them to appropriate service providers. To this end, teams should carry with them UNHCR's Referral for Assistance. In this manner, the emergency referral process and needs assessment fieldwork can be kept separate, and urgent action reports will not be mixed with assessment information for data entry. The emergency referral system needs to be operational when the field assessment occurs.

Needs assessment logistics and equipment
When planning an assessment, it is essential to cover the logistical requirements of the teams that will collect data. Core equipment includes appropriate vehicles, first aid kits, computers, radio or satellite phones, GPS devices, cell phones and chargers with appropriate SIM cards and credit, flashlights, spare batteries for all equipment, physical maps, and stationery (notebooks, pens, pencils and pencil sharpeners). They will also need community interview guides with recording sheets, key informant guides, and any materials required for participatory data collection techniques (flip chart sheets, felt-tipped pens, seasonal calendars, etc.). Ensure that teams carry adequate food and water, particularly if supplies are difficult to obtain in the areas they visit.

Assessment method and survey design
The unit of measurement for an emergency needs assessment should be the community rather than the household or individual. This saves time and reduces the volume of primary data that need to be gathered. Techniques of data collection at community level include direct observation, key informant interviews, and focus group discussions.

The recommended unit of measurement for an urban needs assessment is a neighbourhood or district. In addition to the above techniques, an urban needs assessment might gather information using crowd-sourcing technology or refugee call-in lines (either at the start or later on).

Participatory and AGD approaches
Assessments must also be designed and conducted using participatory approaches that allow populations of concern to voice their opinions about their own needs rather than humanitarian personnel simply deciding for them.

Needs assessments must also adopt an age, gender and diversity (AGD) approach, because the emergency will affect sub-groups in different ways. Field data collection teams need to be gender-balanced and to have had training in participatory approaches and conducting survey questionnaires before they collect primary data.

AGD and participatory approaches may be included in needs assessments in several ways:

• Hold separate focus group discussions for men, women, boys, girls, and groups from different social strata. It may be helpful to conduct these focus group discussions simultaneously so that, for example, men do not influence the women's discussion.
• For closed questionnaires or structured interviews, select key informants from different sections of society.
• When selecting villages or camps to be assessed, map the locations to reflect the diversity of their residents.
• Include questions and themes that are likely to be relevant to marginalized or less vocal segments of communities.
• Collaborate with specialized NGOs who work with hard-to-reach groups (such as older persons, persons with disabilities, children) to obtain information on their specific needs.
• Speak with self-governing associations (of youth, women, etc.) to identify specific needs.
• Use participatory techniques that enable community members themselves to rank problems and prioritize solutions.
• Triangulate the data you gather from different social classes and groups to throw light on how differently they experience the humanitarian situation, and their different needs.

Community participation helps to build on a community's capacities. It also reduces the risk that needs assessments, and the humanitarian interventions to which they give rise, will undermine local coping mechanisms, neglect the needs of marginalized social groups, or waste resources on aid that is not required.

Data collection teams must explicitly request permission from interviewees to use the information they provide (informed consent). Personal information can never be disclosed or transferred for purposes other than those for which it was originally collected and for which consent was explicitly given.

Site selection
Although desirable, it may not be possible to select statistically representative sites in the first weeks of an emergency. For emergency needs assessments, diversity-driven purposive sampling is the preferred sampling method. Purposive sampling aims to sample as many types of site as possible. The team first defines which selection criteria should be considered (in light of the assessment's objectives), and then visits sites that reflect these. Purposive sampling is recommended for rapid assessments because it ensures that the assessment captures the experiences and needs of different groups, and various experiences of impact. An alternative approach is representative sampling, where the population is divided into sub-populations (strata) and random samples are taken in each stratum. Representative sampling requires more time than purposive sampling.

For an emergency needs assessment, sites should be chosen to fill gaps in knowledge (about the severity of impact, geographical factors, the profile of the refugee community, etc.) that were identified during the secondary data review. Consider geography and protection issues but attempt to select sites that reflect the diversity of the situation. You might therefore need to include urban sites and rural camps, sites in mountains and sites in valleys, sites that are dominated by different ethnic or social groups, etc.

For urban assessments, focal points within the refugee community may be able to help identify where refugees in the city are to be found, including hidden or vulnerable sub-groups. Locations should still be chosen to highlight differences - in an emergency's impact on various localities, for example, or in host communities' ability to cope. The core aim remains to provide as comprehensive a picture as possible.

Reducing bias
When collecting and analysing data, a key concern is to reduce bias. ‘Bias' is the systematic skewing of data that are collected. A ‘biased sample', for example, is a sample of the population that does not accurately represent its composition. A ‘biased estimator' is one that systematically over- or under-estimates what is being measured. Bias may occur because only a small number of people in some groups were interviewed, the response rate of certain groups was particularly low, or the team could not gain access to certain areas or people.

Bias can arise for many reasons: the community, Government sources, data collectors, interpreters (if used), key informants, ethnic groups, and both sexes can be responsible for it. The greatest limitation of any key informant interview is that it provides a subjective perspective. Individual responses provide important information, but each will have an individual and a cultural bias that needs to be considered when analysing the information. At all stages of the assessment process, consider whether interviewer bias or informant bias may skew the results, and adjust training and assessment methodologies accordingly.

Certain precautions can help to reduce the bias of any assessment.

• Remember that communities are not homogeneous. Information should always be gathered from different interest groups, including marginalized persons.
• Define the characteristics of those who are consulted. Note those most affected by the crisis, internally displaced persons, minority ethnic groups, women, etc. Record this information when collecting data.
• Record also groups that are absent, whose views the discussion or interviews fail to catch.
• Make sure that affected people are consulted directly and that as many groups as possible are heard (women, children, older persons, persons with disabilities, ethnic and religious minorities, etc.).
• Pay particular attention to the poorest and most socially excluded people, because they are likely to be most affected by a crisis, will have specific needs, and may not speak up easily.
• Jointly analyse results with people from different backgrounds (see below).

Step 5: Designing tools
Data collection questions in the NARE may be further customized, suggested questions may be changed or omitted, and new questions inserted. In addition to sectoral elements, the ‘all sectors' column includes points on population data management, and security and logistics which should be considered.

Primary data-collection forms must be tested before starting a full needs assessment in the field. Testing may be done with local colleagues: what matters is to be sure that the questions are clear and that answers can be analysed. Design of the questionnaire in the appropriate language is the preferred option. Translation is costly, time-consuming and, if done inappropriately, can greatly reduce data quality and accuracy. Plan and budget for translation at an early stage. Agreement on minimum life-saving sectoral data
Needs assessment questions should focus on gathering the core emergency life-saving data that are required for the first phase of an emergency response. All questions should relate directly to operational information that has been prioritized by the assessment team.
When prioritizing assessment questions, consider using a Prioritization Graph (template attached) to facilitate a group discussion of questions to include in the assessment. Prioritize data elements that have life-saving importance and are easy to collect.
Examples of life-saving sectoral data that might be included in the emergency needs assessment can also be found in the minimum and sectoral data sets in the protection, health information system (HIS), WASH, nutrition, mortality, shelter, core relief items (CRI), environment, and livelihood sections of the Emergency Information Management Toolkit. (At: http://data.unhcr.org/imtoolkit/.)

Step 6: Field team training
Needs assessment field team members should be trained in the following:

• Completing the needs assessment data-collection forms.
• Reporting problems.
• Key informant selection.
• Observation techniques.
• Facilitating focus group discussions (if needed).
• Managing community expectations.
• Data confidentiality principles.
• Basic principles of the UNHCR Code of Conduct.

Step 7: Data cleaned, jointly analysed and agreed
Members of the assessment team should agree on a process for cleaning, analysing and interpreting assessment data before data collection starts in the field. Use mobile data collection systems where possible. If paper-based forms are used, field team members should clean and enter data into a spreadsheet at the end of each day. Where possible, data should be broken down into standard UNHCR sex and age groups and other factors of diversity.

Analysis should be done as much as possible jointly with the team involved and other subject and geographic experts. Pooling existing expertise ensures that the analysis is likely to be more comprehensive, robust and accurate. Analysis by staff from different backgrounds can help to mitigate biases, while joint analytical processes can go a long way to create a common understanding of the situation and buy-in for the results.

The team will also need to agree on a communication strategy. In some situations, two versions of the assessment report may need to be produced (one for internal and one for external audiences).

The results of an emergency needs assessment need to be set against any existing contingency plans. The information gathered should also be evaluated, for validity and relevance, the reliability of source(s), and the degree to which it corroborates other information. Any evidence or suspicion of fraud, including falsely reported needs (to attract or deny aid to a place or group) must be reported to senior management.

Unless there are protection concerns, assessment results should be shared at the level of analysis (in most cases, the community). They should be presented in a form that can be easily understood and re-used by others; wherever possible, data should be disaggregated. Emergency priorities, protection concerns and the NARE influence registration planning and provide the foundation of emergency operations' planning.

The information manager should ensure that links between the assessment and monitoring needs are taken into consideration by the RIM WG and operational leads. For example, if the needs assessment shows there are severe problems in a particular sector or in one community, this information can be used to trigger monitoring of those problems. The assessment team should themselves verify the accuracy of the report and acknowledge any weaknesses of documentation or method in the final results.

In addition to the assessment report, the NARE contributes to camp profiles, updates, and funding appeals, as well as other information products. A ‘refugee dashboard' (a one- or two-page visualization of key needs) may be produced to show the results.