October 2011

Technical Brief:

Direct Observation and Key Informant

Interview Techniques for primary data
collection during rapid assessments
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Contents

Contents .................................................................................................................................. 2
1. Introduction ....................................................................................................................... 3
1.1 Information and strategic interaction .......................................................................... 3
1.2 In practice .................................................................................................................. 4
2. Objectives ......................................................................................................................... 4
3. Primary Data Collection .................................................................................................... 5
4. Data Collection Techniques .............................................................................................. 6
4.1 Direct Observation ..................................................................................................... 6
4.1.1 Strengths and Limitations of Direct Observation ..................................................... 6
4.1.2 Basic Principles of Direct Observation .................................................................... 7
4.1.3 Do’s and Don’ts of Direct Observation .................................................................. 10
4.2 Key Informant interviews .......................................................................................... 11
4.2.1 Strengths and Limitations of Key Informant Interviews ......................................... 12
4.2.2 Choosing Semi-structured or Structured Interviews .............................................. 12
4.2.3 How to select Key Informants ............................................................................... 14
4.2.4 Basic Principles of Key Informant Interviews ........................................................ 14
4.2.5 Do’s and Don’ts for Key Informant Interviews ....................................................... 17
5. Reference documents and further reading ...................................................................... 18

Page 2 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Direct observation and key informant techniques for primary data

collection during rapid assessments

1. Introduction

1.1 Information and strategic interaction

Observation and assessment of external information are so fundamental to human nature that
it is difficult to describe their application to data collection methods in disaster needs
assessments. To successfully understand the roles that observation and communicated
information play in data collection during rapid needs assessments, it helps to consider two
distinctions: firstly, the difference between data and information; and secondly, the distinction
between ‘normal’ and “strategic” interaction.

Systematic observation and consultation of affected populations to elicit community

knowledge and perceptions during the primary data collection aspect of rapid needs
assessments produces information. Assessment team members, while observing an affected
community, will ask questions directly to the affected population through interviews and group
discussion and of the affected area through observation. Only by using tools such as check-
lists do these information gathering processes generate data.

Assessment teams have limited control over who they interact with, and in what composition
and sequence. The initial contact with an affected community is usually with community
leaders. This first meeting establishes mutual trust and understanding between the
assessment team and the community. It also communicates how the community leaders
perceive and define the initial impact of the disaster upon their community.

Members of the affected community become key informants (KIs) and provide information on
behalf of the community when the following expectations are operational:
 When the assessment team expects the KIs to know local conditions and have the
authority to share this information.
 When the community expects the KIs to handle the contact with the assessment team
for the benefit of the community.

These two sets of expectations are not identical and both the assessment team and the
community will act and interact in ways that are strategic. The assessment team, by definition,
will have limited time to spend with individual KIs and will need to verify or cross-check
community input. A close-ended communication style is imposed by the team's goal to collect
relevant disaster impact information rather than capture local narratives.

From the community and KI perspective, strategic concerns may be more complex. They will
depend on political, cultural, gender, and other considerations impacting community
organisation and leadership at the affected location. Their expectation and anticipation of
potential humanitarian support will also affect the information they communicate1. This may
be further affected when one or more community members have an awareness of
humanitarian agencies and emergency response processes.

1
Even with the assessment team communicating that participation in a rapid needs assessment does not indicate
preferential humanitarian assistance, during an emergency situation, affected populations will still hope that it
might.

Page 3 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

1.2 In practice

These observations outline a basic canvas for structured observation and the use of key
informants. The practical consequences of this understanding for rapid needs assessments
and their use of observation and key informants are threefold.

Firstly, because learning is an ongoing process for both assessment teams and the
communities they are assessing, both parties may benefit from a two visit assessment
process (it may happen in phase 1-2, or in phase 2-3). This enables both the assessment
teams and the communities they are assessing to revise their initial messages and types,
detail, and content of information they are sharing regarding priority needs resulting from the
disaster in light of exposure to the assessment process.

Secondly, because conceptual learning for assessment teams occurs within a context of
changing environment (both physical and human). During an assessment, team members
may benefit from modifying data collection tools. They may also benefit from adjusting their
communication style with communities as the assessment is ongoing and they are able to
integrate positive, and negative, feedback from community leaders and assessment
participants. Check lists and questionnaires should remain open for modification as should
methods of communication with community leaders and types (and style) of data collected
during KIs.

Finally, as the assessment process is ongoing, the need to collect general and contextual
information on the affected population will continue to be contained by the need to capture
that information in concise and comparable data formats. Certain disaster types have
similar consequences wherever they happen and data collection tends to be standardized, but
in the same time, it is necessary to continue to collect qualitative information which identifies
the subjective nature of the disaster’s impact. Assessment teams will need to combine both
general/contextual information and specific data to ensure a complete picture of the needs of
affected communities and avoid discovery failure.

Ultimately, the key to using direct observation and key informants most effectively during data
collection in a rapid needs assessment is to maximize observation, comparing as much as
possible, as openly as possible, while restricting information gathering to those fundamental
elements which can be used to inform decision makers on:
 who is most affected,
 where they are,
 what their key priority needs are.

2. Objectives

Effective information gathering and data collection during the initial stages of an emergency
depend on the optimal and appropriate application of tools and methods. The use of the
wrong tools at the wrong time results in useless and extraneous information which draws
valuable time and effort from the assessment process and vice versa.

Experience from disaster affected communities shows that direct observation and key
informant interviews are effective data collection methods for the initial phases of needs
assessment. Direct observation provides a snapshot picture of an affected location while the
tools for recording direct observation impressions enable the assessment team to make
critical sense of those impressions. Likewise, key informant interviews provide the

Page 4 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

assessment team with the impressions given by a community spokesperson on their behalf.
Information from key informant interviews can be combined to create a shared impression of
community members as to the impact of a disaster upon their community.

Both direct observation and key informant interviews can be carried out quickly and with
relatively few resources during an emergency. They are typically used together during primary
field data collection (phases 1 and 2) for maximum impact.

This technical brief provides an overview of these two commonly used techniques for rapid
primary data collection while recognizing the operational constraints of crisis situations. The
brief outlines a detailed step-by-step approach on how to undertake direct observation and
key informant interviews.

This technical brief is intended for use by assessment team members who aspire to improve
their primary data collection techniques for rapid coordinated assessments as well as for
training/briefing purposes.

3. Primary Data Collection

Phase 1 of an assessment is defined as a Preliminary Scenario Definition and is achieved

through initial assessments where estimates of the scale and severity of the disaster’s impact
are determined to support initial response decisions. This phase is completed in the first days
following a disaster. Phase 2 of assessment is defined as a multi-cluster/sector rapid
assessment and completed within two weeks of a disaster.

Primary data is most generally understood as data collected directly from the information
source itself and which has not undergone analysis before reaching the analytical phase of the
needs assessment2. Primary data is collected directly from members of the affected
population by the assessment team through field work.

Primary data collection during rapid assessments can have different or multiple purposes. Box
1 below describes some reasons to undertake primary data collection.

Box 1: The purpose of primary data collection

 Gather information not available through review of secondary data.

 Confirm or refute information provided by secondary data.

 Provide a qualitative picture of the range of impacts of the disaster and identify risk
factors.

 Identify priority groups and locations requiring immediate humanitarian response.

 Ensure that the affected population participates in identifying priorities.

 Identify key informants and priority sites for further data collection or monitoring.

2
ACAPS 2011, Technical Brief on Qualitative and Quantitative Data.

Page 5 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

It is recommended that primary data collection during an initial and rapid needs assessment
(phase 1 and 2 of assessment) take place at the community level3. Constraints of time,
access and logistics mean that collecting meaningful quantities of data at the household or
individual level is generally not possible at this stage. Moreover, the collection and recording
of large sets of qualitative data have proven difficult to analyse within the tight timeframes
required for phase 1 and 2 assessment and decision making.

4. Data Collection Techniques

This section describes both direct observation and key informant interview techniques and
approaches.

4.1 Direct Observation

Observation is often underrated as a data collection method. Everyone collects direct

observation information, knowingly or unknowingly. However, employing direct observation as
an effective assessment tool requires consciously using, and recording, what is seen, heard,
and smelled to help shape our understanding of a situation or a problem.

Observation is also a good way to cross-check people’s answers to questions. Its use may
generate questions for further investigation and help form future discussions or frame
questions in case of inconsistency between what the interviewer of a key informant observes
and what the respondents are saying.

There are two approaches to Direct Observation4. Firstly, during structured observation, the
observer is looking for a specific behaviour, object or event. For example, when an observer
looks to see if the population uses soap before and after meals, structured observation can
help answer the question. Structured observation can also be used to detect the non-
existence of a specific issue (e.g. to see if a population is not using soap before and after
meals). To guide structured observation, a checklist is normally developed to function both as
a reminder and a recording tool.

Secondly, during unstructured observation, the observer is looking at how things are done
and what issues exist. For instance, if an observer is interested in knowing how people move
in and out of a camp, unstructured observation is an appropriate method. To guide
unstructured observation, a short set of open ended questions can be developed that will be
answered based on observations.

4.1.1 Strengths and Limitations of Direct Observation

Direct observation can be used to rapidly collect different types of information in an

emergency situation. It does not require costly resources, or detailed training, which makes it
a quick data collection process that is easy to implement.

However, because direct observation as a data collection technique provides a snapshot of

the situation, it has limited power in a rapidly changing situation or where there is substantial
population movement. Furthermore, it provides limited information about capacities and

3
The Operational Guidance for Coordinated Assessments in Humanitarian crises (NATF, Feb 2011).
4
Child Protection Rapid Assessment Toolkit as of January 2011.

Page 6 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

priorities of the people5. Finally, while specific training is not a prerequisite for effective direct
observation, some preparation is necessary to ensure that the observers are aware that their
own perceptions and expectations are subjective and impact upon how they report and
interpret their observations. The gender, age, ethnicity and previous disaster response
experience of the observer can all effect the interpretation of data collected during
observation. Particular sector specializations (e.g. protection, WASH, shelter, etc.) of
observers also may influence their observation findings, as observers may focus observation
on their own area of specialization or misinterpret information outside of their specialization.
The technical expertise required to answer particular observation questions should match the
level of technical expertise of the observers.

4.1.2 Basic Principles of Direct Observation

This section outlines basic principles for direct observation before, during and after primary
data collection.

Before the field assessment

Every data collection instrument (e.g. questionnaire, interview checklist) should make
provision and space for direct observation comments and notes as they help add context and
meaning to the data collected.

Table 1: Recording Observations6

Example of form for recording observations
Location Observation Significance Follow-up
 Poor drainage around  Water contamination  Investigate
well; spilled water flowing likely to lead to household water
Village X back into the well diarrheal disease, usage: do people
 Animals walking around particularly among boil and/or treat
the well young children water?

Data collectors must be informed of the value of their observations through pre-field visit
preparation, and understand how direct observation links with other data collection tools.
There will be further benefit from instruction on why and how to systematically record direct
observation in questionnaires or through separate checklists, while ensuring that their
observations are separated from the respondents’ comments or responses.

It can be useful to hire an interpreter to help make sense of local observations and clarify
assumptions about issues raised during direct observation. However, as with the data
collectors themselves, interpreters will have their own cultural prejudices and biases which
need to be made overt when analysing information collected during direct observation. For
example, an urban, educated translator from a differing ethnic group may have different
perceptions about an affected population than someone with less education who is from a
rural community and is of an ethnic group present in the assessment locality.

During the field assessment

Direct observation starts upon initial entry into an assessment site, much before an interview
or discussion. Field assessors should observe conditions and particular features from a range

5
Adopted from the Guidance on Profiling Internally Displaced Persons, NRC’s Internal Displacement Monitoring
Centre and UN OCHA, April 2008 Edition.
6
Modified from the 2009 WFP EFSA Handbook.

Page 7 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

of viewpoints and places to provide a representative view of the affected area. Often things
seen on the drive into the area or upon entering the village on foot provide valuable contextual
data. If there is a high point, such as a hill or a tall building, the site should be observed from
above to get a sense of the conditions and variations across the site.

After an introduction to relevant community leaders, a community assessment should start

with a walk around the location. Walking through the area with local people facilitates
discussion and can be an excellent way to come across unexpected information.

Assessors should also spend time in communal or public places (cafés, tea shops, markets,
religious buildings). Look around and talk to people. A local market is an excellent first stop in
the observation process as it gives a useful picture of what is available, what people produce,
buy and sell as well as what the prices are for basic commodities.

Observation provides immediate information for assessing the status of existing infrastructure.
Driving along a road is a sure way of finding out if it is passable, but be careful in conflict areas
where landmines and explosive remnants of war may pose security problems.

During the assessment, take the opportunity to observe with an open mind, compare as
much as possible, but restrict the information gathering to what can be processed,
condensed and analysed within the assessment time frame.

Observe:
o People’s physical condition and
activities Box 2: Key sites for observation
o Children, older persons, the
chronically ill, and those persons
Water collection points, latrines, communal
with disabilities
o Housing, properties, livestock, washing areas, schools, storage facilities,
assets, etc. grave sites, markets, health facilities and
o Where appropriate, the daily lives religious centres.
of women (be aware that in some
cultural settings, it is inappropriate and disrespectful for men to observe and/or
interview women)
o The state of public services, sanitation systems, and infrastructure (e.g. schools, water
points, health posts etc.)
o If possible, power relationships within the community and whether people from
different groups have different coping mechanisms or access to aid.

Record both what was expected to be evident in the community as well as what was not
observed. The absence of people in the market, of children in the schools, of men or women
in displaced population groups is as important as their presence.

Where culturally acceptable and the security situation permits, take pictures. Photos, video
footage and even sketches can be useful in communicating to others the reality of the
situation7. When photographing individuals or photographing when physically in an affected
community, always ask permission before taking pictures. Be sensitive to the fact that taking
photographs of affected persons can both endanger them (in conflict settings) or be highly
inappropriate (such as men photographing women). Do not endanger the assessment team
by attempting to take photos where they are prohibited, e.g. in military installations.

7
Modified from 2000 IFRC Disaster assessment guideline.

Page 8 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Cross-check information. If discussing water, ask to see the water source. If people describe
unfamiliar foods or building methods, ask to see them. Direct observation can be used for on
the spot triangulation for the responses, discussion, and explanations given by affected
persons. However, there is only so much that a team can do in a day, in terms of trying out
different methods and integrating information across them. Fatigue, as the day lengthens,
interferes with the team members' capacity for note-taking, mental review and comparing, and
sensible further questioning of KIs. It is often more productive to have fewer observations and
meetings, but to conduct these more slowly, with careful note-taking and opportunities for both
the team and the community to make revisions and then actually use the precious information.

Meet with the whole assessment team at least once during the fieldwork day at each site to
review progress and decide which important places still need attention before leaving the site.
This helps avoid gaps in gathering essential data about important points.

At the end of the field assessment visit, meet with community representatives. Explain what
has been done and seen, share the initial conclusions, and inform the community how this
information will be used. Be sure not to make commitments or promises regarding assistance.

After the field assessment

A debrief between assessment team members should be organised by the team leader to
collect observations from the team, triangulate information and wrap up final conclusions of
the field visit. Direct observations must be transferred from individual checklists to a data
summary sheet where necessary.

Highlight areas where team observations and population responses do not match to enable
further analysis of discrepancies and identify triangulation needs.

Page 9 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

4.1.3 Do’s and Don’ts of Direct Observation

Box 3: Do’s and Don’ts of Direct Observation

Do:
 Enter the observation process without pre-conceived notions and fixed
expectations.

 Note observations made and information volunteered that are related to subjects
beyond formal assessment concerns. Be prepared to follow advice from people met
in the locations, and use the opportunity to observe things which were not planned.

 Walk across the community outside of predefined routes such as roads, paths or
natural boundaries to obtain a cross-section of points for observation and provide a
balanced view of conditions.

 Record information which is contradictory or surprising to expectations8.

 Keep focused to make useful comparisons.

 Be active and curious in the observation process. Observation is not just about
seeing, but also about hearing, smelling, tasting, feeling and touching.

 Be aware of what was not seen. Note the absence of services and infrastructure.

 Respect local culture. Community members are observing you just as much as you
are observing them. Follow local rules of behaviour, e.g. do not smoke during
interviews. Be aware of gender dynamics and ensure that the teams reflect this.
Be sensitive to local concerns, e.g. if there is a shortage of food and water, do not
consume food in front of affected community members.

Don’t:
 Begin the observation process with a set of expectations or seek to record data
primarily to prove a pre-existing hypothesis.

 Rely on remembering information. Record observations on a checklist.

 Focus solely on misery and destitution. Be aware of capacities, opportunities, and

social capital within the affected community.

 Be intrusive. Take steps to be as sensitive and respectful as possible9.

 Take a photograph without asking prior permission.

8
Roberto, Michael, 2010, Principles for effective observation.
9
Oxfam, Food Security Assessment Guidelines, 2003.

Page 10 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

4.2 Key Informant interviews

In addition to direct observation, KI interviews are a commonly used data collection technique
for rapid assessments. A KI interview is one where an individual with prior knowledge of the
affected community is questioned to gather key information on the impact of the disaster and
on priority community needs. The crucial element of a KI interview is that the informant is well
versed in information about his/her community, its inhabitants10, the site visited, or the
emergency11 either because of his professional background, leadership role or personal
experience. Typically a KI is a local leader whether civil, government or religious.

This initial group of KIs may be highly diverse in perspective and competency. Their personal
knowledge of the disaster situation may also greatly vary as well as their expectations of what
relief and reconstruction agencies will do for their community. KIs may know surprisingly little
or may contradict each other’s response. KIs may further retract their initial descriptions of the
impact of the disaster when a more powerful individual joins the discussion and offers a
different assessment.

The heterogeneity of KIs and strategic interaction of both assessment team members and KI
respondents needs to be considered. In theory, the community would need more time than
the one day assessment to agree on and harmonize key community messages. Follow-up
meetings in short order would be desirable. However, this is not practical under the time
pressure of phase 1 and 2 assessments. Therefore, the one-point information exchange tends
to focus on areas of ready agreement. For example, in coastal cyclones in Bangladesh,
assessment teams and KIs may emphasize the destruction of homes and public buildings. By
contrast, the problem of salinity in rice paddies filled with sea water, and therefore the issue of
resuming traditional farming vs. converting to shrimp ponds (where poor people regularly lose
out), may take much longer time to rise to proper attention.

Second, the sophistication of communities in working with assessment teams varies greatly,
and sometimes inversely with needs (because communities with more sophisticated key
informants are probably those which even in normal times have greater diversified human and
social capital, and thus should be more resilient to disasters). Under time pressure, teams
visiting sophisticated communities may be happy to accept an already widely elaborated self-
assessment wholesale, with some of the most vocal key informants literally looking over the
shoulders of team members and directing them how to fill out check lists.

In less sophisticated communities, assessment teams (particularly locally hired enumerators)

may studiously work with community leaders on plausible scenarios until a "cognitive
consensus" is reached. For example, claimed measles vaccination coverage of 80 percent
before the disaster may reflect favourably on the village, but may be sufficiently problematic in
the current circumstances to attract assistance. In other words, from a traditional survey
quality perspective, measurement error may be major, yet is socially accepted.

10
Adopted from the Guidance on Profiling Internally Displaced Persons, NRC’s Internal Displacement Monitoring
Centre and UN OCHA, April 2008 Edition.
11
2009 WFP EFSA Handbook.

Page 11 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

In communities least tuned into administrative discourse, observation of physical events

including destruction, illness symptoms may be favoured by teams who cannot find key
informants speaking a language they can understand themselves or through translation.
Alternatively, influential persons from neighbouring communities may accompany teams to
such sites, e.g. teachers from a central village with students from an outlying village to serve
as interpreters. They would interpret for teams linguistically as well as substantively, by
helping define impacts and needs. In northern Iraq in 2003, landmine survey teams elevated
multiple villages to new survey targets after visiting the central villages and taking guidance as
to where to delineate distinct settlements12.

However, regular citizens can also be valuable KIs because they can share their
representative and personal experience. For example, a young female household head may
be able to highlight priority needs from the perspective of a mother; likewise a person who is
unable to walk without assistance may be able to highlight challenges that certain strata of the
community face in accessing aid due to mobility challenges. While not traditionally considered
to be KIs, these individuals can provide a unique perspective of the experience of typical
members of the affected community.

Key informant interviews may be used to:

 Obtain technical information from people representing specific professions, such as health
workers or school teachers
 Gain specific knowledge about a specific topic or sector (e.g. interviewing a water
committee representative)
 Delve into sensitive and protection issues that may not be appropriate for group discussion.

4.2.1 Strengths and Limitations of Key Informant Interviews

KI interviews can be organised quickly and carried out with few resources. They have
particular value in gaining a perspective of the impact of the disaster on a community where
access to affected populations has been compromised or is difficult. They also provide a
holistic and qualitative overview of the impact of a disaster on community members.

The greatest limitation of a KI interview is that it provides a subjective perspective on the

impact of a disaster. As with all individual responses, information will have both an individual
and a cultural bias which needs to be considered when analysing KI interview responses.

4.2.2 Choosing Semi-structured or Structured Interviews

This section provides an overview of how to undertake a KI interview and which issues need
to be taken into account. A KI interview can be semi-structured or structured.

12
Even if chiefly for the purpose of lengthening their employment.

Page 12 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Semi-structured interview (checklist):

A semi-structured interview is a guided interview in which a limited set of questions are

decided ahead of time13. The questions are open ended, with the aim of stimulating discussion
on a given topic. Box 4 describes open and closed ended questions. When conducting a semi-
structured interview, the interviewer uses a checklist or question outline instead of a
questionnaire and tries to build a relaxed and constructive relationship with the KI through a
conversational approach. This requires the interviewer to be familiar with general cultural
considerations, be sensitive to the
interviewee, and not be judgmental or Box 4: Types of questions
too set in their pre conceived ideas.
While understanding the language can Close ended questions have specific answers,
be an advantage14, the more important which are normally short, with yes/no or categorical
concern is to cross-check translation to answers. They are generally easy to aggregate and
ensure that the concerns of the analyse as they do not required complex recoding
interviewee rather than those of the operations.
translator are being captured.
Open ended questions have no fixed set of
Analysing findings from semi-structured responses allowing the respondent to answer as
interviews entails a labour-intensive s/he sees fit. They allow the respondent to think and
process given that there is often a reflect, and give the respondent a chance to voice
greater range of answers and 1
their answers in their own words . Analysis
responses than in other forms of
demands the ability to rank priorities and compare
information collection. One way to do
qualitative responses holistically.
this is to try to summarize answers by
main points raised and then create a
limited number of sub categories. This will help determine how interviewees prioritised or de-
prioritised certain issues. Another option is to summarise the interview into a single summary
sheet listing the sectors and sub sectors affected as well as the concerns and priorities
expressed by the population. When comparing different interviews across affected
communities, the findings of the different summary sheets can be aggregated into one and
information analysed to identify patterns and areas of concern.

Structured interview (questionnaire):

In its simplest form, a structured interview involves one person asking another person a list of
pre-determined questions about selected topics using a questionnaire. The aim of a structured
interview is to ensure that each KI is asked exactly the same questions preferably in the same
order. This ensures that answers can be accurately aggregated and that comparisons can be
made with confidence between sample sub-groups or between different assessment periods.

A list of predicted options for answers to the questions can be included so that assessors
simply need to tick the box. This saves time and increases accuracy in the field. However,
the assessment teams need to take care not to lead respondents by reading out the options.
The questionnaire must also always have a space for assessors to include options other than
those which have been pre-defined.

Design of a good questionnaire demands technical expertise, experience and a good

understanding of the context. Structured interviews are recommended for phase 2

13
2006 UNDAC Handbook
14
World Food Program, EFSA, 2009

Page 13 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

assessments when findings of phase 1 assessment provide practical recommendations on

information needs and areas requiring further investigation15.

Structured interviews can be time consuming and care should be taken to keep them focused
and limit the number of questions asked. Experience from post disaster responses indicates
that spending roughly an hour per interview and selecting a cross-section of KIs maximises
the range and quality of information gathered.

4.2.3 How to select Key Informants

During rapid assessments, KIs are selected to provide general information about population
profiles and movement trends, security, and sector issues (water, environment and sanitation,
food security/nutrition, shelter, health, protection, environment, education, etc.).

The number and type of KIs selected per location will depend on the range of expertise or
perspective available from the pool of KIs, the nature of the disaster, the availability of people
and the time that can be spent at the
Box 5. Useful resource persons for rapid
site. When identifying the KIs,
remember to arrange interviews with assessment in emergencies may include
individuals of different genders, ages,
and religious and/or ethnic minorities At district/local level, representative(s) of:
to ensure a full picture of the affected Government authorities, local leaders, village elders,
community. It is important that the police, army, fire service, rescue services, NGOs, civil
assessors take into account power defence, IFRC/ICRC, international and national relief
dynamics within a community and teams/organizations officers, religious leaders, UN
that opposing social groups (strata) national staff, doctors, nurses, TBAs, evacuation
do not speak for one another. centre focal point, teachers, etc.

As noted in section 4.2, a KI can also At capital level, representative(s) of:

be an individual who represents National authorities, UNDAC and UN agencies
certain aspects of the community and members, geographical institutes, departments of
can provide meaningful indications meteorology/hydrology, agencies, NGOs, embassies,
about access, risks, priorities, OCHA staff, etc.
vulnerabilities and capacities at the
community level.

Where an affected community includes different population groups, such as a host population
and a displaced population, key informants should be selected from all groups of interest16.
Groups should be divided based on heterogeneity of experience: if one group is likely to
experience the humanitarian crisis in a significantly different way than another group, each
group should have its own key informant. Box 5 provides examples of people who can be
useful key informants.

4.2.4 Basic Principles of Key Informant Interviews17

Before the assessment

Involve experts in the design and planning of the assessment, especially for the sampling, the
site selection process, the design and translation of the questionnaire.

15
Phase 1 assessment will support the design of an appropriate and adapted questionnaire as well as to guide the
site selection process in phase 2 of assessment.
16
2009 Initial Rapid Assessment Guidance Notes.
17
2007 Initial Rapid Assessment Guidance Notes.

Page 14 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Field test the data collection instrument and refine it as necessary. A field test will provide a
good indication of the complexity of the data collection instrument and the time required to
complete it. Questions which KIs have difficulty understanding or which make them feel
uncomfortable or prone to providing false answers should be re-worded, replaced or removed
following the field test.

Plan the field data collection carefully. Ensure that there is enough time to carry out KI
interviews. Remember to inform the authorities of the assessment itinerary and bring
credential letters to the assessment locations which explain the assessment objectives.

Choose an appropriate and experienced assessment team. Ensure gender and age balance
within the team(s) as well as translators if necessary. Divide tasks according to the expertise
of team members.

Ensure team members are properly trained to achieve accurate and precise assessments.
Team members should be briefed on and understand the objectives, methodology and
principles of the rapid assessment and the possible interventions that could be implemented
as a result of it. Each interviewer should be thoroughly familiar with the data collection process
and the information being elicited by each question. Provide field notes that explain and define
key terminology and outline site sampling.

During the assessment

Box 6: Reduce bias while selecting respondents
Select the identified KI (see box Remember that communities are not homogeneous.
618). Be aware of the KI’s situation, Gather and weigh information from local sources that
what activities they are engaged in, represent different interest groups, including marginalised
and what their surroundings are. persons. For example, select informants from both host
and displaced populations, where present.
Interview people in a safe place
that is convenient to them and Define the different characteristics of people to
adapt to their needs. consult (e.g. those most affected by the crisis, IDPs,
minority ethnic groups, etc). When conducting key
informant interviews, check who is present against this
Make sure people understand why criteria. Note groups that are not represented.
they were asked to participate as a
KI and what will be done with the Wherever possible, consult the affected population
directly, including women, children, older persons,
information they share. Do not raise
persons with disabilities, and ethnic or religious
expectations. minorities. The poorest and most socially excluded
people in the affected community are likely to be worst hit
Ensure good communication and by the crisis. Do not rely only on information from official
sources and those in power as they do not always
informed consent. Participants represent excluded groups in their communities.
must understand that they are not
required to participate in the
interview. Make sure they understand that a lack of participation will not negatively impact
them.

The full list of questions may not necessarily be covered during the interview. The order in
which questions are asked may change according to which KI is interviewed first. However,
18
Source: 2010 UNICEF JENA.

Page 15 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

consistency in the order of question asked does make data compilation and analysis easier.
Nevertheless, it is important to be flexible and adapt according to specific situations. Be aware
that the more difference there is between the way information is collected at different sites, the
more challenging and time consuming it will be to build an overall picture of the humanitarian
impact in the affected area during the analysis phase.

Start the interview with general questions about the situation and allow the interviewee to raise
issues of concern to them before guiding the conversation to the subjects of interest to the
assessment team. Ask questions that are factual and relatively straightforward to answer first,
and move on to more sensitive issues only when the KI is more at ease (build trust before
asking intimate, private or potentially endangering questions).

Ensure that the translator understands the subject and vocabulary of the interview and is able
to forge a respectful relationship with interviewees.

Take notes throughout the interview; ensure that the information is transferred safely and
without distortion.

Combine interviews with observation to verify information and correct inconsistencies19.

Consult the people affected and not only their official representative. Consider the needs of
different groups and individuals, seek out marginalised groups and ensure their interests are
taken into account.

When an interview does not yield the overall perspective needed, politely bring the discussion
to an end, thank the interviewees for their time, and seek other KIs to talk with20.

Structure the interview with each KI informant with care. Make sure KIs know that their time
and participation is valued. Do not end the interview too abruptly.

Give KIs the opportunity to ask questions or share their thoughts on issues that have not yet
been discussed. Be careful not to raise unrealistic expectations of aid.

Record metadata (such as date, location of interview, social role of interviewee, group
represented by the interviewee, etc.) for each KI, as this information will be used in the
interpretation of the data.

After the assessment

Conduct a debriefing at the end of the assessment day to give assessment team members the
opportunity to discuss the strengths and weaknesses of the interviews and the interview
process. Compare findings, views and impressions. Gather observational information,
anecdotes, or concerns not captured in the data collection form. Consider the reliability of the
key informants as well as the team bias. All of this information should be considered and
included in the final report.

19
2007 Initial Rapid Assessment Guidance Notes.
20
2009 Initial Rapid Assessment Guidance Notes.

Page 16 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

Endeavour to maintain communication with the visited communities and inform KIs about how
the information they provided is being used and what follow up actions are being taken. Share
the final report with them if possible.

4.2.5 Do’s and Don’ts for Key Informant Interviews21

Box 7 – Do’s and Don’ts for Key Informant Interviews

Do:
 Introduce the assessment team and ask permission to carry out the interview.
Build trust with the KIs, give them time to talk about their priorities or express
grief.

 Ask to take notes or use a PDA to record the interview. With a PDA, explain what
it is and how it works.

 Be sensitive to the time needed to complete the interview. Be aware that the KI
may have pressing obligations or may have no obligations and just needs to talk.

 Make sure the data collection instrument has space for capturing direct
observation comments and notes. Keep the data collection instrument brief.

 Avoid/limit open ended questions in the data collection instrument for phase 2.

 Choose your key informants well. Know the information gap and identify the KI
best able to finding the missing information.

 Choose a limited number of critical topics to discuss the KI.

 Be alert to non-verbal signs and behaviours which indicate how comfortable the
KI is with the interview. If the KI is uncomfortable with the questions, do not insist
they answer.

 Be consistent. Use the same methods in each community visited. Record data
consistently to ease comparisons and highlight obvious differences that stand out.

 Record access routes, time taken and other logistical tips to help future plans.

 Give voice to all vulnerable groups, specifically women, children, older persons,
persons with disabilities and religious and ethnic minorities.

Don’t:
 Waste time talking as a whole team to one KI (apart from initial introduction to
authorities or other gatekeepers).

 Substitute direct observation for the KI’s answer or explanation to a question22.

Where direct observation differs from a respondent’s answer, note this and try to
determine potential reasons why this may be the case.

21
2008 and 2009 Initial Rapid Assessment Guidance Notes

Page 17 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

 Put the KI in a compromising situation by conducting an individual interview.

Explain to community observers why the specific KI was chosen and what topic
you want to discuss.

 Interrogate respondents as an extractive process.

 Create expectations about future humanitarian support.

 Monopolise the time of individual interviewees. Especially during times of crisis,

people have their own priorities.

 Limit information to one KI’s response. Triangulate by asking other KIs until you
are confident that there is consensus on this point

 Induce particular answers by helping an interviewee to respond.

 Ask questions that may stigmatise people or endanger them.

 Use people’s names when collecting information. Ensure the anonymity of the
data collected, but if key protection risks are observed, refer them confidentially to
Protection Cluster colleagues for appropriate and confidential follow up.

 Prevent KIs from asking you questions at the end of the interview.

 Let a translator answer a question for the interviewee or dominate the interview
process.

5. Reference documents and further reading

 ACAPS 2011, Technical Brief on Qualitative and Quantitative Data.

 Global Protection Cluster, Child Protection Rapid Assessment Toolkit, 2011
 Inter Agency Standing Committee, Initial Rapid Assessment Guidance Notes, 2007; 2008;
2009
 International Federation of Red Cross and Red Crescent Societies, Disaster Emergency
and Needs Assessment, 2000,
http://www.ifrc.org/Docs/pubs/disasters/resources/corner/dp-manual/Disemnas.pdf
 International Federation of Red Cross and Red Crescent Societies, Vulnerability and
Capacities Assessment Toolbox, (VAC), 2007
 International Federation of Red Cross and Red Crescent Societies, Guidelines for
assessments in emergencies, 2008,
http://www.ifrc.org/Docs/pubs/disasters/resources/responding-disasters/guidelines-for-
emergency-en.pdf
 International Mine Action Standards, Data Collection and Needs Assessment, 2005,
http://www.mineaction.org/downloads/1/2%20-
%20Data%20Collection%20and%20Needs%20Assessment.pdf

22
2005 IMAS Data collection and needs assessment.

Page 18 of 19
 Technical Brief – Direct Observation and Key Informant Interview Techniques

 REDLAC Work Group on Risk, Emergency and Disaster of the Interagency Standing
Committee for the American and Caribbean region, Methodology Rapid Assessment for
Humanitarian Assistance, 2006,
http://www.humanitarianreform.org/humanitarianreform/Portals/1/cluster%20approach%20
page/training/CSLT%20July%2007/Day4/REDLAC%20Rapid%20Needs%20Assessment.
pdf
 UNDAC Handbook, 2006,
http://ochaonline.un.org/OCHAHome/AboutUs/Coordination/UNDACSystem/UNDACHand
book/tabid/6012/language/en-US/Default.aspx
 UNICEF, Joint Education Needs Assessment Toolkit, 2010,
http://oneresponse.info/GlobalClusters/Education/publicdocuments/Ed_NA_Toolkit_Final.p
df
 UNICEF, Short Guide to Education Assessment, 2010
 UNFPA, Guidelines on Data Issues in Humanitarian Crisis Situations, 2010,
http://www.unfpa.org/public/cache/offonce/home/publications/pid/6253;jsessionid=5A6402
366EBCC70ED26BCF59296A976F
 UNHCR/WFP, Joint Assessment Mission (JAM) Guidelines, 2008,
http://home.wfp.org/stellent/groups/public/documents/manual_guide_proced/wfp203513.pd
 UN OCHA/Norwegian Refugee Council, Guidance on Profiling Internally Displaced
Persons, April 2008 Edition,
http://procaponline.unocha.org/docs/library/Guidance%20on%20Profiling%20IDPs.April%2
008.pdf
 World Food Programme, Emergency Food Security Assessment Handbook, 2009,
http://www.wfp.org/content/emergency-food-security-assessment-handbook
 World Food Programme, Emergency Food Security Assessments, Technical Guidance
Sheet nr. 8, 2009,
http://home.wfp.org/stellent/groups/public/documents/manual_guide_proced/wfp197291.pd
f
 World Food Programme, Monitoring and Evaluation Guidelines, 2009,
http://www.wfp.org/content/monitoring-and-evalutation-guidelines

Page 19 of 19