GUIDELINES

FOR DEVELOPING
BASELINE STUDIES
EXPERIENCES FROM CARE NEPAL

CARE Nepal

May 1998
Jopie Duijnhouwer

F\Archive\Baselines\Guidelines

TABLE OF CONTENTS
1. Introduction.........................................................................................4
2. Overall framework ...............................................................................4
3. Why are we collecting baseline information? ........................................5
3.1. Baseline design principles..............................................................6
4. What kind of information has to be collected? ......................................8
4.1. Review of the logframe ..................................................................9
4.2. From indicators to information needs ..........................................10
4.3. Collection of general information.................................................11
5. How to collect this information?.........................................................12
5.1. Information collection methods ...................................................12
5.1.1. Review of secondary information ...........................................12
5.1.2. Participatory methods............................................................13
5.1.3. Household survey ..................................................................14
5.1.4. In-depth interviews ...............................................................16
5.2. Linking information needs with methods .....................................16
5.2.1. Putting the pieces together....................................................18
5.2.2. Developing formats ...............................................................21
6. How to analyze this information? .......................................................21
6.1. Consolidation of the data.............................................................21
6.2. Analysis of data from PRA exercises ............................................22
6.3. Analysis of household survey data ...............................................23

Design of a baseline study

1.

Introduction

In the last two years, CARE Nepal has significantly increased its
experience in designing and conducting project-wide baseline studies.
This was not the first experience of CARE Nepal in this field, as baseline
studies were done as early as 1991 in Bajura. One baseline study focused
on health1, the other on home gardening2. After these studies other
baseline studies were conducted, but the data were not always analyzed
and reports on these studies were not available. In 1995 the Primary
Health Care / Population section conducted a comprehensive study in
Bajura, Solukhumbu and Mahottari districts3.
The recent series of baseline studies started in 1997. The first baseline
study was conducted in Bardia (May / June 1997). Since then three other
project-wide baseline studies have been conducted in Mahottari (August
- September 1997), Bajura (December 1997 - February 1998) and Jajarkot
(March - April 1998). Another baseline study for Syangja district is
planned for the coming months.
These studies have increased CARE Nepals knowledge and experience in
doing baselines. With this report CARE Nepal wants to consolidate its
knowledge and experience on designing, conducting and analyzing
baseline studies. Furthermore, this report can serve as guideline to
individuals developing new studies. However, it should not be considered
as a recipe and the ideas suggested here should be used in a flexible way.

2.

Overall framework

During the design of a base ine study a number of questions should be

answered in a systematic way:
1. Why are we collecting baseline information?
Discusses the reasons why baseline information is collected and to a
certain extent which type of information will be collected and the most
appropriate method to do this;

R. Rajlawat, 1991. PHC Base Line Survey Report. RABNP No. 19.
A.R. Giuliano et al., 1992. Home Gardening Baseline Survey Bajura and Mahottari
Districts Nepal.
3 Valley Research Group, 1995. CARE / Nepal Family planning Survey in Bajura,
Mahottari and Solukhumbu.
1
2

2. What kind of information has to be collected?

This section elaborates the method used by CARE Nepal of
transforming project logframes into defined information needs;
3. How to collect this information?
Discussion on the choice of the methodology to be used for
information collection, and also the design of the formats to be used
during the field work;
4. How to analyze this information?
Consolidation and analysis of the data is complex, although it can be
simplified if sufficient preparations are taken during the study design.
These issues are closely related. While identifying information needs it is
essential to think about the possible methods of data collection and
about the inputs required for the proposed methods. A number of these
questions (#2 and # 3) can be answered during the design process (either
during a workshop with Country Office and project staff or by an
individual or small group), but only after some other issues are dealt with
(#1).

3.

Why are we collecting baseline information?

There are good reasons to do a baseline study: most importantly, it

allows organizations to measure their progress in achieving their
objectives and most donors will demand it. This also makes it possible to
learn from both successes and failures.
Currently, CARE Nepal is collecting baseline information before new
projects are started or enter a new phase and before projects
substantially extend their working area. In practice it is not always
possible to start information gathering before the project starts, in which
case the baseline study is done as soon as possible after the start of
activities.
The objectives of the baseline studies done so far can be summarized as
follows:
1. To gather baseline information related to the objectives and indicators
of the project document, so as to be able to measure progress and
also to verify targets and indicators.
2. To deepen the understanding of the problems and opportunities of
communities living in the study area;
3. To gain a better understanding of the working area which can lead to
refining the planned interventions;
4. To establish working relations with the stakeholders in the working
area.

3.1. Baseline design principles

After the decision has been made to do a baseline study, certain issues
have to be dealt with before the actual design starts. These issues follow:
> the scope of the study.
In the past many organizations have conducted large scale baseline
studies, which covered a large number of variables. Many of these studies
are large scale household surveys with huge questionnaires. This type of
study can produce a wealth of information, but their scope makes the
design, implementation and analysis much more complex and external
support will often be necessary. A disadvantage of a very broad study is
that a substantial chunk of the information collected may not be relevant
for the project.
> the focus of the study.
This issue is related to the previous one. It is important to remember that
the quantity of interesting information that could be collected is almost
unlimited, but the quantity that can be collected and analyzed definitely
has its limits. Therefore it is necessary to define prior to the study what
kind of information should be collected. One of the guidelines that can be
used is the logframe which shows which indicators the project wants to
achieve. Information directly related to these indicators is necessary to
assess whether our project interventions are successful or not (e.g.
during mid-term or final evaluations). However additional information
can improve the quality of project interventions: it is important to know
that most of the adult men in Bajura migrate during four months in
winter. It can also produce information to test assumptions made in the
logframe, e.g. on womens workload.
> study team composition
Another important consideration which has be dealt with beforehand is
who will design conduct the study and analyze the information. This can
done either by external consultants or by CARE staff (or a combination of
consultants and staff).
The recent series of baseline studies was done by project and CO staff
and with minimal external support. Involving project and CO staff in the
process of designing, implementing and analyzing has advantages. It
offers an excellent opportunity to increase the insight of staff in the
project objectives as well as in the local situation. This is especially the
case when the studies include a large number of PRA exercises, which
usually yield much qualitative information about the working area and its

population. If the baseline is limited to a household survey, hiring

external enumerators would probably be a better option.
The study will also benefit from the participation of staff, because they
may understand well the project area, both in knowledge of the area and
in a better understanding of the local customs and language1.
Doing the baseline study with project and CO staff has also certain
disadvantages: people who have lived for a long time in the study area
may be biased, although this is unlikely at the moment of the baseline.
Furthermore the studies are time-consuming and may interfere with
regular project implementation. However, the financial possibilities will
often decide whether a study can by done by external consultants or
should be done as much as possible by CARE staff.
> study methodology
The choice for a specific type of study methodology (basically
Participatory Rural Appraisal (PRA) tools versus household questionnaire)
is linked with the objectives of the study, the nature of the logframe
(quantitative versus qualitative indicators) and with the study team
composition. During the Bardia baseline development workshop the
participants compared the characteristics of the two methodologies (see
Table 1).
The design principles used for the series of studies done so far in CARE
Nepal can be summarized as follows:
1. the baseline study is based on information needs derived from
indicators in the logframe;
2. additional information is collected if directly relevant to project
interventions;
3. a mix of participatory methods and a household questionnaire is
used to obtain both quantitative and qualitative data2;
4. the scope of the study is such that data collection analysis and
reporting can be done by CARE Nepal staff;
5. obtain reliable data using a method that can be easily replicated;
6. the costs of the baseline study should be reasonable.

This is not always the case; because field staff is regularly transfered, they will not all
understand local languages or dialects in the beginning.
2 There are arguments to split the baseline study in two parts and to start with a
household survey. The results from this survey could then be used to design the
qualitative / participatory component, with mostly participatory tools, more focused on
why issues, problems and opportunities identified during the household survey.
1

Table 1.

Comparison of participative methods and household

questionnaire.
participatory methods
structured survey questionnaire
static; gives detailed information
about one particular point in time

flexible / dynamic ; can give

information on trends and in-depth
information on the causes of the
observed trends
listen to the population

top-down

high level of participation from

community during analysis
produces qualitative information

community is less involved; they

are producers of information
produces quantitative information

does not produce statistically

generalizable data; can infer only

can be solidly documented;

reliability of data can be assessed;
generalizable
rapport is less important; people
may feel uncomfortable being
asked many questions
can be intimidating for women,
especially in very traditional areas

more emphasis on building rapport

with population
can involve both men and women if
well-done

4.

What kind of information has to be collected?

After choices have been made on the design principles, the logframe
indicators have to be translated to information needs. Most of the
indicators taken in consideration will be effect and impact level

indicators. Output indicators are considered to be services delivered by

the project and prior to project interventions these will be zero. There
may be some exceptions, e.g. when a project aims at improving existing
services. In this case it is necessary to collect information about the
current level of services and the perception of the population on the
quantity and quality of these services. In addition to this, assumptions at
all levels should also be included in the baseline study1.

Including information needs regarding assumptions is limited to those which can be

measured / assessed at district level, e.g. the availability of market channels for the
distribution of improved seeds. Whether the macro-economic situation is stable or not,
will not be a part of the baseline study.
1

Note that this step will be started by CARE after a choice has been made
to involve CARE staff as much as possible in design and implementation.
When it is decided that the baseline study will be designed, implemented
and analyzed by external consultants, the initiative for the rest of the
design process will be on the shoulders of these consultants, who can use
any of the next steps or propose their own methodology under guidance
from CARE and using the parameters of the Terms of Reference.
For Mahottari and Jajarkot the design of the baseline study was done
during a workshop with the study team members.This workshop covered
the following topics:
1. analysis of the logframe and identification of terms in the logframe
that have to be clarified or defined;
2. identifying information needs based on the project logframe;
3. orientation on the use of different study methods;
4. linking information needs with the most appropriate study methods;
5. sampling methods and sample size;
6. develop work plans for the field study based on the study methods
identified under point 4;
7. develop formats for data collection.

4.1. Review of the logframe

The first step of the workshop is a thorough review of the project

logframe. All objectives and indicators have to be clear to all the
participants of the workshop. Terms, which are ambiguous, unclear and
which have not yet been defined in the project document, have to be
defined first1.
The quality of the effect and output indicators is important in completing
this step. If the terms used are well-defined and specific, it is in most
cases not difficult to identify the information needed.

see for example J.J. Duijnhouwer, 1997. Baseline study for Bardia Buffer Zone
Development Project, where a large number of terms used in the Bardia logframe are
discussed.
1

This box illustrates the importance of well-defined terms when

defining information needs. In the Bardia logframe indicator 3a is:

3a) 25% of the households owning land and engaged in farming

activities within the core project area perform any or a combination of
the farm practices promoted by the project by the end of phase I.
However, farm practices promoted by the project were not defined,
which made it impossible to identify the information needs. So, as a
first step, the intended practices were specified and only then the
information needs could be defined.
In the Bajura logframe the indicator on kitchen gardening is more
precise:

1a) 40% HHs produce at least 6 different types of vegetables in their

kitchen garden over a year and some vegetables exist for at least 6
months per year.
Based on this indicator it is relatively easy to identify the information
needs:
how many households have kitchen gardens;
what is the variety of vegetables cultivated (in households with
kitchen gardens;
during which months do they actually produce vegetables.

4.2. From indicators to information needs

If the indicators are formulated in a clear and unambiguous way and if the
terms used are clear for everybody, then the transformation of indicators
in information needs should be rather straightforward (an example of this
Table 2.
Indicators and information needs for Intermediate Goal # 4 of
Jajarkot Poverty Reduction Project.
4) Diversified and increased sources of income.
indicators
information needs
4a) increased number & value
of (active) sources of income
in the project area.

> existing sources of income

> the range of income generated through
these sources of income
> market opportunities for (forest) products
and traditional skills

4b) changes in participants

self- perceptions in income.

> perceived change in household socioeconomic status

4c) changed consumption

patterns (as a proxy for
income)

> purchase of cooking oil, masala, soap, hair

oil, toothpaste during the last three months
and year

exercise is given in Table 2).

It is important to realize that there are is a lot of information that could
be collected. To collect information for the first indicator (4a) a fullfledged socio-economic household survey could be done, which tries to
measure all the sources of income, expenditure etc. However, this is not
in line the design principles which state that design, field work, analysis
and report writing will be done by the available CARE staff.
Sometimes it is not easy to come up with information needs, either
caused by the phrasing of the indicators (see e.g. indicator 4b) or by
technical problems related to collect data. However, indicator 4c gives
more problems (see 5.2 for a discussion on proxy indicators).
It is, for example, difficult to measure the productivity of crops in many
areas, because the correct measurement of the production requires a
huge study with complex sampling techniques which measures the
production in farmers fields during harvest over an extended period of
time. Because of the complexity of this type of research the productivity
of crops is estimated by asking households about the production of their
crops during the last season. An additional question which compares the
production of last year with the average production of the five previous
years can give a good idea about the production of most of the crops
(cross-checking with data collected on food self-sufficiency may give
further indications about the reliability of the data).

4.3. Collection of general information

Data on the indicators mentioned in the logframe provide a basis for the
comparison of the situation at the start of the project and the end of
project. However, there is information of a more general nature which is
important to gather during a baseline study. Some of this information
may be available from secondary sources or from the situational analysis
in which case there is no need to gather additional information.
Other important general information may not be available and it may be
necessary to include these issues in the baseline study. Land tenure is not
a variable that CARE projects are trying or able to influence, but it is very
important to know what percentage of the population owns land and what
percentage is landless. Project interventions which require land (kitchen
gardening, agroforestry, tree plantation on private land, improved
varieties, etc.) will not be profitable for landless families.
Other information like seasonal migration is relevant for project
interventions. If this information has not been collected during the
situational analysis, or if more in-depth information is required, it can be

included as part of the baseline study. Village histories and well-being

ranking increase our knowledge about the population living in the project
area.
But again, many things are interesting and more-or-less relevant for the
project. It is important to limit the collection of information to a
minimum, without omitting essential information. In the Jajarkot baseline
study the area planted with and the production of improved varieties was
initially included in the study design. However, the logframe does not
refer to it and as other questions regarding productivity were already
included, it was left out from the final version.

5.

How to collect this information?

After completion the list of the information needs based on the logframe
the next step is to identify the most appropriate way to collect this
information. For our purposes three broad methods of information
collection should be considered:
1. review of available documents and literature;
2. participatory tools;
3. household survey;
4. in-depth interviews with stakeholders.

5.1. Information collection methods

5.1.1. Review of secondary information
The first option is of course the easiest one, but quantitative and
qualitative studies of most of the project areas are lacking. Information
can be obtained from the district headquarters, line agencies and Central
Bureau of Statistics. Most often the data obtained from these sources do
not give specific enough information on the working area of the projects.
Furthermore the quality of the information from secondary sources is
sometimes questionable.
However, it is always worthwhile to investigate if other organizations have
been active in the working area and if they did proper baseline studies or
other research oriented studies. In the case of Bardia several studies were
available for the project working area and although these studies were
geographically most often only partly overlapping, they provided a lot of
useful information. The information obtained from the available
documents was analyzed and compared in a literature review study which
took place before the design of the baseline study. No further information
was collected during the baseline study when the information from
different sources on a specific topics was consistent. However, if data

were inconsistent or unlikely, additional information was collected to

cross-check.
5.1.2. Participatory methods
Currently, participatory methods are very popular in data collection.
During the baseline studies extensive use was made of these methods
such as village histories, social and natural resource mapping, well-being
ranking, seasonal calendars, trend lines and Venn diagrams1. For certain
types of information it is the only method available to collect information.
This is the case for village histories or Venn diagrams. Other data can be
collected in different ways, but using participatory methods is far easier
and cheaper than other methods. However, the method most suitable is
in most cases determined by the type of information required.
For example, a seasonal calendar of workload can be done in
participatory way, but could also be done by daily or weekly structured
interviews with a large number of women. The latter method is much
more cumbersome and probably not much more reliable than simply
asking people about the comparative workload in different months. On
the other hand, asking information about family planning in focus groups
may be much less reliable than information obtained during household
interviews.
Participatory tools are easy to handle for most of the CARE staff, who all
had training on PRA techniques, and many of them have extensive
practical experience. It is also much more interesting for the participants,
who can at the same time visualize and observe and reflect on their own
situation. It gives field staff also an excellent opportunity to learn about
the peoples perception of situations and changes over time. It also offers
the opportunity to establish rapport between staff and the community
and it demonstrates to the participants that project staff respect and
value their perspective and knowledge.
However, the problem with some (not all) participatory methods is the
reliability of the data. The data obtained through participatory methods
are more subjective and based on the perception of people, living in a
specific area in a specific period. Most of the information is relative and
can not be compared across different areas. For example, our informants
in Bardia, who are living in clearings in the forest, described a strongly
negative trend for the availability of firewood, which was comparable to
the trend line produced in Mahottari. For study team members the
situation was not comparable, because in Bardia piles of dry wood were
1

For a description of the experiences with these tools in baseline studies, see Annex 1.

used to protect the fields against wildlife, whereas people in certain

clusters in Mahottari district used no wood at all and relied completely on
dungcakes. Although the trends in these two districts may be
comparable, the situation can not be compared in absolute terms.
There is a general exception to the rule that data from participatory
methods are more subjective. A lot of hard information can be obtained
from social mapping, like the exact composition of the village, the
number of landless households or households which are involved in
seasonal migration, the number of water points, toilets, kitchen gardens,
etc.

5.1.2.1.

Sampling issues for participatory methods

Sampling for participatory methods has specific problems, related to the

type of information generated by the method. In the case of quantitative
information like the number of households with kitchen gardens or
toilets, each of the households can be treated as a separate yes/no
answer. If a field check shows that the information provided by the
participants in the social mapping exercise is accurate, then each of the
households can be treated as a separate datapoint.
However, in other cases (Venn diagams or seasonal calendars) the
information from the participatory exercises will provide quantative
information, but will produce only one or two datapoints per cluster.
Although the information obtained through these methods is not hard,
it is necessary to cover a reasonable number of clusters to get some idea
about patterns. Information from earlier clusters should be checked later
during the discussion other groups. There is no scientific rule about how
many clusters or wards to sample for these participatory exercises.
During the recent series of baseline studies at least 12-18 clusters were
covered representing 20-25 % of the total number of clusters in the
project area. The larger the number of clusters the better the information,
but in remote districts in Nepal the logistical consequences of increasing
the number of clusters are considerable.
A third group of exercises produces basically qualitative information. It is
not possible to give fixed rules for the number of samples for this type of
execises. It is again important to cover a reasonable number of clusters
to be able to distinguish patterns and ensure representation of different
(caste) groups. However, if cropping patterns are found to be identical in
most of the clusters, it is not necessary to repeat these over and over and
it may be more interesting to discuss other topics related to agriculture,
like changes in cropping pattern over the years or eventual differences

between different socio-economic groups in the community. If new

information keeps coming up it is important to continue
For these qualitative exercises it is even more important to assess the
variation between different castes, socio-economic strata, men and
women. Different groups make different use of natural resources, men
may be involved in seasonal migration with cattle, whereas women and
children are herding cattle around the village or cut fodder and hay in the
dry season. Low caste blacksmiths use the forest to produce charcoal,
etc. It is worthwhile to discuss certain topics with other groups instead of
repeating the questions with the same type of focus groups (unless there
are interesting differences between the different focus groups).
Another issue in using participatory methods is the composition of the
groups of participants. In general separate exercises with men and
women should be done if the outcome of these exercises is expected to
be different. In the case of village histories, trend lines and social maps
can be done with mixed groups of men and women.
For seasonal calendars of workload, natural resource mapping and most
of the semi- structured interviews it is necessary to work with separate
groups of men and women. Other exercises, like Venn diagrams and
well-being ranking can be done with separate groups, but it depends on
the time available whether this can be realized or not.
5.1.3. Household survey
A household survey, if properly designed and implemented, is the best
method to collect statistically valid information that can be generalized to
the whole project area. The data obtained in this way can be compared
with other periods and is therefore a good way to compare the situation
at the start of a project with end of project situation.
Household surveys have also certain disadvantages: they require qualified
staff to design, conduct and analyze the findings and the jargon and
sophisticated software used for data analysis give these surveys a
superior appearance which is not always justified.
The reliability of the data collected with a household survey can be
determined beforehand and it is possible to determine the sample size
depending on the desired reliability. For a detailed discussion of some of
the statistical issues of sample surveys, see Annex 2.

5.1.4. In-depth interviews

Sometimes in-depth interviews with key informants can give valuable
qualitative information, which link the perception of change of the
interviewees with the causes for these changes. In-depth interviews
asking about changes in socio-economic situation can reveal the causes
behind the changes that take place in the life of ordinary people.
Sometimes, when a specific person is very important for the project (like
the Park Warden in Bardia) only in-depth interviews with the Park Warden
and/or other officials will reveal changes in attitude towards communities
close to the Park.

5.2. Linking information needs with methods

The comparison of the different methods in section Table 1 shows that

none of the above-mentioned methods is superior in all aspects. All have
their advantages and disadvantages and for each of the indicators
mentioned it will be necessary to reflect on which will be the most
appropriate method(s) for data collection. Sometimes it may even be
necessary to use two different methods for one indicator: e.g. use a
household survey to gather data on the production of foodgrains
combined with a participatory technique to investigate the change over
time in foodgrain production.
In general the choice for a specific information collecting method will
depend on:
the type of information required;
the time and human resources allotted to the study team;
the desired level of precision;
the time requirements of the different methods;
the possibilities of combining data gathering of different variables in
one tool.
Table 3 gives an elaboration of Table 2 and includes columns for the data
collection method and a plan of analysis. This table shows that for each
of the indicators different methods are used to collect the information.
The existing sources of income were measured in different ways:
focus group interviews with men were used to get an idea about the
variety of sources of income, the importance of these sources of
income and changes over time (trends). These interviews also indicate
whether sources of income are open to everybody or limited to specific
groups or castes;
during natural resource mapping information is obtained on the
importance of Non Timber Forest Products (NTFPs) in a cluster;

quantitative information was obtained through a household survey,

which asked every sampled household whether they earn money
through different livelihood strategies. Through this survey it is
possible to assess how many people are involved in which activities
and what is the approximate amount of money earned through these
activities.

In this case the focus group interviews and natural resource mapping give
qualitative information on the sources of income, and the household
survey complements this. It allows comparison or triangulation between
the different methods. If a certain source of income is mentioned during
Table 3.
Indicators, information needs and information collection methods
for Intermediate Goal #4 of Jajarkot Poverty Reduction Project.
4) Diversified and increased sources of income.
indicator
information need data collection
plan of analysis
method
4a) increased
number & value
of (active) sources
of income in the
project area.

4a) > existing

sources of income
in the district
> the estimated
income generated
through these
sources of income

4a)> natural
resource mapping
> semi structured
interviews

4a) > maps;

description of FUGs;
> synthesis of
management of
natural resources
> household survey > tabulation of
sources of income
per category, cluster
or caste
4b) > focus group
4b) changes in
interviews with men 4b) >inventory of
participants self- 4b) > market
and women
opportunities and
perceptions in
opportunities for
problems
income.
(forest) products
> household survey
and traditional skills
> average income
> reported income
per household per
source of income
4c) > household
4c) changed
survey
4c) > tabulation of
consumption
4c) > purchase of
expenditure
patterns (as a
cooking oil, masala,
patterns for
proxy for
soap, hair oil,
households in
income).
toothpaste during
different sociothe last three
economic categories
months and year
> household survey
> school attendance
> literacy rates
of children
(boys/ girls) in
relation to caste and
socio-economic
status

the focus group interviews but which does not come back during
household surveys additional attention should be given to this subject by
cross-checking with other key informants.
Another method which could have been used to add information is
ranking of the importance of the different sources of income. However,
the household survey already provides this information and adding extra
methods does not seem necessary in this case.
Indicator 4c is an interesting example of the use of a proxy indicator.
Objective 4 aims at increasing both the diversity as well as the value of
the sources of income. It is for several reasons very difficult to measure
income, and to avoid all the technical problems related to measuring
income a proxy is chosen. A proxy measures the change indirectly, so the
change in income is not measured, but rather the change in household
expenditures. Buying more things more frequently indicates that the
family income has gone up and that is felt to be sufficient for our
purpose. The selection of the items mentioned in Table 3 is based on the
importance of these products and the socio-economical situation of the
project area. Most products are luxury products, which means that
people will buy these products after they have bought the necessary food.
It is assumed that an increase in income will lead to an increase in the use
of these luxury products. The list used in this example is based on
previous experiences with measuring changes in household expenditures
in remote areas of Nepal.
5.2.1. Putting the pieces together
After the list of necessary information is completed and the most
appropriate data collection methods are selected it is time to put the
pieces together. The individual methods have to be combined in a overall
study methodology, which may include a review of secondary data, PRA
tools, as well as a household survey.
The most important issues to be dealt with are:
the sequence of the different methods to be used

It is best to start with exercises (after introducing the study team

members, the objectives of the project and the study) which create
good rapport and which involve a large group of people. Village
histories and social mapping are most suitable for these purposes. An
example of a typical workplan for a cluster or ward is presented in
Annex 3.

the time and staff available for the study

Project staff have their own on-going activities according to their

workplan. A baseline study should not take too much of their time. In
the recent baseline studies project staff were involved in field work for
four to five weeks.
the number, location and variation of the population in the wards or
clusters covered during the study
In many districts the variation between clusters and wards is large,
especially if their is a distinct cluster-wise segregation of castes or
ethnic groups.
the time spent in each cluster or ward
This issue is linked to the previous two: increasing the number of days
per cluster means a reduction in the number of clusters that can be
covered. Reducing the number of days per cluster means that less
information can be gathered! The number of clusters covered until
now has varied from 12 (Bajura and Mahottari) to 18 (Bardia);
size and composition of the study team
All studies were done with two teams of three to four persons (with
the exception of Mahottari). Doing the baseline study with a smaller
number of people has definite advantages, because the team members
have the opportunity to learn from their experiences and because it is
much easier to coordinate the study and check the formats. Preferably,
each of the study teams should include at least one woman (although
two is better) and each team should include people with different
professional backgrounds so that they can adequately explore the
varied themes covered in the baseline study.
During this part of the design it may be necessary to make changes in the
information needs as defined in earlier stages, because it may not be
possible to use all the different methods. Not only is there a limit to what
the study team members can do, but the villagers also get tired if they
have to participate in many different exercises. Overburdening villagers
or team members will reduce the quality of the information. Examples of
a cluster-wise study plan with an overview of methods plus the expected
outcome is added as Annex 4.
The timing of the field work is also important, although the experience
has shown that external factors (e.g. pressure from donors or start of

project activities) often play a decisive role in the timing of the studies.
Planting and harvesting time of rice should be avoided at all times,
because this is the season with the highest labor peak for farmer families
and it will be difficult to meet people. Although the timing of the studies
in Bajura (December / January) and Bardia (May / June) was not very good
(seasonal migration in Bajura and a heat wave in Bardia), it did not have a
negative impact on the participation of the population. The best season is
between November, after Tihar until April, except that at this time it may
be difficult to meet migrating men.

5.2.1.1.Timetable
After the completion of the study methodology for each of the clusters, a
timetable can be made for the baseline study.
This should include the following elements:
1. training / orientation of study team members
2.

3.

4.

5.

Until now three-day trainings have been given in the project area
field testing of the methodology
Although it is very important to test the methodology before starting
the baseline study, it was in the past not possible to do field testing
because of time constraints. This lack of testing was overcome by a
very intensive support to both of the teams during the field work in
the first clusters. After the completion of all the methods the
experiences were reviewed by all team members and adaptations were
made if necessary.
travel
In remote areas the distances between VDCs and even between
clusters within one VDC can be considerable and it is therefore
important to include sufficient time for traveling.
cleaning, arranging and correcting notes and formats
Often notes are made during discussions with individuals or groups; at
times it is not always possible to make notes. Therefore the formats
have to be reviewed, completed or cleaned by the team members if
necessary during the field work. The team coordinator should check if
all formats are completed and if no exercises have been forgotten.
This also takes time and has to be included somewhere.
rest and recreation
The circumstances during the field work are often demanding and the

team members need time to rest, wash their clothes, and get charged
again.
The time table for the baseline study in Bajura is added as Annex 5 and
the estimated time requirement for different levels of staff is presented in
Annex 6.
5.2.2. Developing formats
The process often described as collecting information is in fact often a
process of losing information. The loss of information will be greatest for
participatory methods as the participants may discuss a large number of
topics in a semi-structured way. While noting down the contents of these
discussions a lot of information is lost and this becomes only worse when
the data from all wards or clusters are compiled. It is therefore essential
to develop strong formats to prevent the loss of data which are
considered most relevant.
For household surveys good formats are also essential, to avoid
ambiguities in the questions and to make later analysis easier. A set of
formats used during the Bajura baseline study is added as Annex 7.
During the baseline studies done so far, Access software (a database
program developed by Microsoft) was used to design formats for the
household survey. This program is programmer- and especially userfriendly. The formats used for data entry in the computer can be identical
to those used by the study team members used in the field, which
simplifies data entry in the computer.

6.

How to analyze this information?

6.1. Consolidation of the data

After completion of the field work the data have to be collected and
entered in the computer. This is often a tedious job which includes
checking all the formats and entering the data.
Data can be entered in Word (e.g. focus group interviews, village histories
and most other PRA tools), Excel or Access. Access is better for data
management, but few staff are used to working with this software. It has
the advantage that it is very flexible and can store large chunks of text as

well as figures and pre-defined categories. Especially the latter

characteristic is very valuable for large surveys1.
A certain level of quality control in data entry is possible by comparing
some ( e.g. 5 %) of the original forms with the data as entered in the
computer. If mistakes are often observed in a few specific variables it may
be worthwhile to check them on a larger scale and correct all entries.
Access has advantages in data entry, because the formats used in the
field can be identical to those used for data entry in the computer.
Information entered in Access can transferred with one push on the
(right) button to transfer it to Excel for (statistical) analysis.

6.2. Analysis of data from PRA exercises

The analysis of the results from many of PRA methods is tricky.

In the case of village histories it will be limited to a qualitative analysis, in
which the common characteristics of the village histories are brought
together. Some of the village histories contain many distinct historical
events, like migration streams, floods, epidemics, etc. (e.g. Bardia),
whereas others almost mythical in nature (Mahottari and Bajura).
Social maps and natural resource maps contain often a large variety of
information. Some of this information can easily be analyzed, like the
number of households, the number of toilets, water points, etc. Other
information is much more qualitative, like the use of the natural
resources and the quality of the water sources. The analysis of the first
type of information can be done with the use of a spreadsheet (Excel),
while common sense is the most essential ingredient for the analysis of
the qualitative information.
The results of seasonal calendars can be transformed in quantitative data.
However, the number of observations for these participatory methods is
much smaller: one per cluster compared to 25 observations per cluster
for the household survey. The results from the seasonal calendars can be
averaged over a number of clusters, because the timeframe used
(months) is the same for all clusters (analysis stops at making averages
and comparing the averages from different months).

Excel treats Chettri, Chetri and Chhetri as three distinct categories. This kind of
different spellings and typing mistakes can be avoided by creating pre-defined lists.
1

This is not the case for trendlines. The participants in trend lines often
use time references which are different in many clusters. This makes it
much more difficult to calculate averages for several clusters. In practice
it is often possible to find a number of clusters with a comparable time
scale and these can be used to calculate averages.
The same holds for the results of wealth ranking. Participants in certain
clusters distinguish four categories, whereas others may distinguish three
or five categories. In these cases it is necessary to elaborate the
characteristics of the categories mentioned for the most common
classification (often in four categories). This standard list can then be
compared with the characteristics of the categories of the other
classifications.
The analysis of the data from the Venn diagrams is more complicated. In
certain wards the participants in this exercise may mention only four
organizations, whereas other wards may mention nine or ten
organizations. To make the data from different wards comparable a
formula has been developed to normalize the results.
The process used is as follows: the ranking of an organization as given by
the participants during the Venn diagrams was transformed in a score.
The formula used for this transformation is 10/n * (n+1-r), with n the
total number of organizations mentioned in a specific ward and r the rank
of the organization in the ranking1. If the participants in one ward
mention four organizations and rank these as 1, 2, 3 and 4, these
organizations get a rating of 10, 7.5, 5 and 2.5 points respectively. In
the analysis of the Venn diagrams both the number of wards which
mention a specific organization as well as the average normalized score
of this organization can be mentioned.

6.3. Analysis of household survey data

Many different methods are available for the analysis of the data from
household surveys. Interested people can use e.g. G.W. Snedecor and
G.W. Cochran, Statistical Methods as a reference. For a baseline the
analytical part of the study is limited. The report should include averages
of the most relevant variables plus their standard deviation or confidence
limits.
The pivot table feature in Excel is a great help in the analysis of data and
produces cross tabulations of averages, sums, standard deviations, etc.
In this ranking, one is the most important organization, two the second most
important organization, etc.
1

The average number of trees per cluster can be replaced the total number
of trees for each category in the well-being ranking within ten seconds
(as long as the data are in one table of course).

LIST OF ANNEXES
Annex 1. Description of the PRA tools used during the baseline studies.

26

Annex 2. Sampling for a household questionnaire.

29

Annex 3. Cluster-wise work plan for a baseline study in one cluster.

32

Annex 4. The expected output of the tools used during the baseline study.

33

Annex 5. Time frame used for the baseline study in Bajura.

36

Annex 6. Estimated time requirements for different levels of staff involved in a baseline
study.
37
Annex 7. The formats used for the Jajarkot baseline study.

38

Annex 1. Description of the PRA tools used during the

baseline studies.
This overview gives a list of the different participatory methods used in
the baseline studies. It does not go into the details of the tools, which can
be found in many documents, but focuses on the possibilities and
experiences with these tools during the baseline studies.
For more information on the tools see for example:
Theis and H.M Grady, 1991. Participatory Rural Appraisal for
Community Development.
CARE Nepal, 1997. Participatory Rural Appraisal for Community
Development. A training manual based on experiences in Nepal.

Village histories

After an introduction in the village during which the objectives of the

project as well as the baseline study are explained, the study teams often
started with a village history. During these village histories a group of
people (including at least a few old men and women) are asked to relate
the story of their village, the most important events which took place. If
important events are known from other places (e.g. drought, flooding,
migration) probing can be used to assess the influence of these events on
this village.

Mapping

After village histories social maps or resource maps from the area are the
most logical steps. Initially social maps were made first, followed by
resource maps, but in many cases the team prefer to do these exercises
simultaneously, using the outlines of the natural or social resources map
to construct the second one.
Natural resource maps give an overview of a larger area, indicating
neighboring clusters, government forests, park boundaries etc. Certain
clusters (Bardia and Mahottari) were big (up to 400 households) and
composed of several sub-clusters. It is not possible to work with such
large numbers in participatory exercises, and therefore a choice was
made to select a part of the cluster or a sub-cluster for detailed study.
This can be done based on a social map.
In general the number of households in the samples varied from 35 to
120. The natural resource map was used to delineate the part of the
cluster that was studied more in detail during the later exercises.
Social mapping was used to identify all the households in the (sub)
cluster, to identify the major characteristics of the clusters, drinking
water, schools, etc. The social maps can also used to indicate landless
and/or migrating households, etc. This information combined with the

names of the household heads gave a complete overview of the

demographic situation of the cluster.
Participants during the social mapping showed a strong tendency to
expand the area covered in the map, because those who were living in
areas not covered by the map were afraid they would also be excluded in
the future from project activities. During social mapping people were also
asked about the existing traditional organizations within the communities
and the NGOs active in these clusters.

Well-being ranking

Small groups of people (mostly composed of two women and two men)
are asked to rank the households in their cluster. The list of households
was generated during the social mapping. In general well-being gives
very interesting results. The informants in most clusters distinguished
four to five different categories of well-being. After the ranking the
people were asked to analyze the choices they had made by trying to
describe the characteristics of the different categories.
The results from the well-being ranking were often used in the analysis
of the quantitative data. This proved to be very fruitful and in general
there was a strong relation between the results from the well-being
ranking and the household survey. This shows that well-being ranking is
an effective method to get an indication about the socio-economic wellbeing in an area.

Trend lines

Trend lines describe the changes over time in the availability of resources
(firewood, fodder or cattle) or production. This exercise can be combined
with e.g. natural resource mapping or semi-structured interviews.Trend
lines showed often negative trends, with the exception of trend lines on
food production in a few Terai clusters. There seems to be a tendency to
glorify the past and it is questionable whether these negative trends are
real or perceived.

Seasonal calendars

Seasonal calendars are used to get information on changes in the

availability of food, fodder and firewood during the year. As mentioned
under trend lines, it was easy to combine these exercises with related
exercises like natural resources mapping or semi- structured interviews.
Contrary to the trendlines, the seasonal calendars seem to be quite
accurate and are strongly related with the cropping pattern. Seasonal
calendars were made by separate groups of men and women.

Semi Structured Interviews

Semi Structured Interviews (SSIs) are interview during which the

interviewer limits the questions to a specific topic. Normally the
interviewer has a list of key questions, but follows the stream of issues
brought up during the discussion. SSIs can be combined with any of the

other tools described above and are often complementary. The more
visual exercises describe a given situation, whereas the SSIs explains the
reasons behind it.
Where trendlines or seasonal calendars clarify changes over time, the SSIs
can help to understand why these changes take place. Venn diagrams
produce a graphic overview of the important institutions in a cluster; the
complementary SSI can clarify the reasons behind the assessment made
by the population. Often SSIs are done with a specific group of people,
who are supposed to be in comparable situation and face the same
problems (focus group interviews).

Interviews with government organizations and NGOs

Apart from the fieldwork in the villages, interviews were held with key
stakeholders who are active in the buffer zone. These include park
authorities, iNGOs and local NGOs as well as governmental organizations.

Annex 2.

Sampling for a household questionnaire.

Sampling method and sample size are complex issues, and in this annex I
want to touch only a few aspects of this field. Those who are interested in
the details should refer to handbooks.
Sampling methods
Normally samples are used for household surveys, as it is in most cases
impossible and unnecessary to cover the whole population during a
survey. However, there are many ways to select a sample. The basic
method is simple randomized sampling, in which a given number of
households are selected from the total population from the study area.
This means that a list of al households from the study area (which may
included 10-16 wards) is needed plus a method for random selection. In
this way no sampling bias occurs, but the disadvantage is that the
selected households are not clustered and that visiting all households
may take a lot of time, especially in remote areas.
A second option is stratified random sampling. If it is known that large
differences exist between certain groups or areas, stratification may be
used. The strata are chosen in such a way that the variance within the
groups is as small as possible. For example, VDCs can be used as
different strata or ethnic groups. The sample size for each of the strata
can best be chosen in such a way that it is proportionate to the actual
shares of the strata in the population. This allows easy calculations of
population averages without complex weighing factors.

Multistage random sampling uses the same techniques as mentioned

above. In this case two different levels of sampling units are defined. In
the baseline studies done so far the primary sampling unit was in most
cases the cluster or ward. To avoid extensive travelling, two clusters or
wards were randomly selected from the nine wards in the studied VDCs.
The second sampling unit are the households within the clusters / wards
selected in the first step. Random selection can again be used for the
selection of the households. This is a case of two-stage random
sampling.
Sample size
The size of the sample i a complex topic. It depends among others on:
1 the type of variables measured;
The standard deviation of a specific variable has a large influence on the
sample size. The formula used to calculate the sample size

t 0.95 * SE 2
n=
(for variables with a normal distribution and 95 %
L2
confidence limits).
with
SE standard error or standard deviation
n
sample size
L
allowable error
t0.95
a constant, to be used forsamples larger than 60 and 95 %
confidence limits. This constant is approximately
4.0 for 95 %, 2.7
at 90 %, and 6.6 for 99 % confidence limits
There is a large differences between the SE of different variables. The
standard error of the weight of paperclips is small compared to the
weight of CARE staff. Four measurements of small size paperclips will
probably enough come up with a good estimate of the average small
size paperclip with an allowable error of 10 %. However, it may be
necessary to weigh 40 staff members to have the same estimate for
the average weight of the CARE staff.
Most often the SE of variables are unknown (this is not always the case,
for health related variables SE are often known), and the best
alternative is to use the SE of variables used in previous studies.
The situation is slightly different for the binominal distributed
variables, or in simple terms the yes/no answers. In this case the
formula to measure the sample size is:
t 0.95 * p * q
n=
L2
with:
p the probability that a certain event occurs
q the probability that a certain event does not occur (p+q =1)
In this case no standard deviation is necessary to establish the sample
size and if we take the worst-case scenario of p=0.5 (which implies
4 * 0.5 * 0.5 2
that q=0.5), the formula becomes: n =
= 2.
L2
L
This means that if the allowable error is put at 10 %, the sample size is
100. With an allowable error of 5 %, the sample size becomes 400.
Most household surveys measure both quantative as well as binominal
variables. Binominal variables require the largest sample size, and
therefore the desired sample size depends basically on the binominal
distributed variables.

2 the desired level of reliabillity;

One of the most important factors in deciding on sample size is the
desired level of reliability. The formula to calculate sample size
contains the t0.95.However, if the desired level of reliability is put at 99
%, the t0.99.value has to be looked up in a table (this alue is 6.6). At 90
% reliability we find t0.90 or 2.7. Substituting both values in the formula
will show the consequences of these changes.
3 sampling method
All these formulas are based on simple randomized samples. If other
sampling methods are used correction factors have to be used to
estimate the sample size. These correction factors are mostly based
on experience and there are no statistical calculations available to
determine these correction factors.
In the CARE M&E workshop the following correction factors were used:
stratified sampling
0.5-0.95
cluster sampling
1.5-3.0
4 resources available for the research.
This adds of course the more pragmatic considerations to the discussion
about the size of the sample.

Annex 3.

Cluster-wise work plan for a baseline study in one cluster.

day 1

day 2

village history

resource m apping
com bined
w ith SSI

SSIon sources of
incom e

socialm apping

venn diagram and

SSIon local
organizations

focus group
discussion on
agriculture

w ellbeing ranking

day 3

household questionnaire

Annex 4.

The expected output of the tools used during the baseline study.

ACTIVITY

EXPECTED OUTPUT

HOUSEHOLD

household questionnaire covering household

composition, literacy, farming systems (300
households: 100 high caste; 100 Chetrri; 100 low
aste households

QUESTIONNAIRE

ACTIVITY

EXPECTED OUTPUT

a map, indicating important characteristics of a

ward including clusters, forest, grazing land, up
hill agricultural land, etc.

RESOURCE
MAPPING

SSI on natural resources, focusing on land use

and user rights of forests and grazing lands
indicated on the resource map. Indicate also
when these are used and who controls the use of
these lands

a combination of SSI and trend lines on

the availability of fodder, firewood and
other forest products (if relevant)
a combination of SSI and seasonal
calendars on the availability of fodder,
firewood and other forest products (if
relevant)

a Venn diagram with all relevant

organizations and institutions

VENN

discussion on the importance of different types of

organizations and institutions, with special
emphasis on :
* (l)NGOs

Annex 5.

Time frame used for the baseline study in Bajura.

DATE
15th Dec
16 th Dec
17th Dec
18th to 20th Dec
21st Dec
22nd-24th Dec
25 th Dec
26th Dec
27th Dec
28th Dec
29th Dec - 3rd
Jan
4th Jan
5th Jan
5th - 11th
12th Jan
13th Jan
14th-16st Jan
17nd Jan
18rd- 21th
22th Jan
23th Jan

ACTIVITY
Leave for Nepalgunj
Leave for Kolti
Leave for Martadi
Orientation of study team members
Leave for Budhiganga VDC
Field work of Budhiganga VDC in two wards by 2 teams
Return to Martadi
Completion of forms and feed back
Leave for Kolti
Leave for Sapala and Rugin VDC
Field work of 2 wards in each VDC( Sapala and Rugin) by
two teams
Return to Kolti and completion of all forms
Leave for Wai and Jukot VDC
Field work of 2 ward in each VDC (Wai and Juku) by two
teams
Return to Kolti
Completion of forms and rest
Field work of Bandhu VDC by two teams taking a ward each
Completion of forms and evaluation of the overall work by
both the teams
See to all the entry of data in computer
Return to Nepalgunj
Return to Kathmandu

Kolti Area:

Rugin ----6 hrs from Kolti

Wai-------3 hrs from Kolti
Jukot------31/2 hrs from Kolti
Sappata---- 7 hrs from Kolti
Bandhu---- 2 hrs from Kolti

Martadi Area:

Budhiganga-- 4hrs from Martadi

Annex 6.
Estimated time requirements for different levels of staff
involved in a baseline study.

phase

staff level
coordinator

time
requirements per
person in days
7

total staff
requirement in
person days
7

design
logistics

assistant

training

coordinator

5 (incl. travel)

assistant

5 (incl. travel)

project staff
(7)
coordinator

28

assistant

30

30

30

280

data entry

project staff
(7)
secretary (3)

12

data analysis

coordinator

10

10

assistant

10

10

coordinator

25

25

fieldwork

report writing

Totals:
coordinator:
assistant
secretary
project staff

50 person days
49 person days
12 person days
308 person days

6.4. Annex 7.
study.

The formats used for the Jajarkot baseline

Annex 7.

Overview of forms and interview guides used for this study.

Form 1a.

Village history.

Form 1b.
Raja).

Information on village and local leaders (Mukhiya and/or

Form 2.

Compilation of households and well being ranking.

Form 3.

Traditional forms of community organizations.

Form 3a.

Description of the traditional organization of the community.

Form 3b.

Description of the traditional irrigation systems.

Form 4a.

Overview of community organizations present in cluster.

Form 4b.

Group description.

Form 5a.

Drinking water situation in the cluster.

Form 5b.

Availability of services (time needed for one way trip).

Form 6.

Characteristics of categories of well being ranking.

Form 7a.

Trend lines of livestock.

Form 7b.

Trend line of agricultural production.

Form 8a.

Seasonal calendar of the availability of fodder.

Form 8b.

Seasonal calendar of the availability of food.

Form 9.

Cropping patterns.

Form 10a.

Availability of different natural resources.

Form 10b. Use of natural resources.

Form 11.

Seasonal calendar of agricultural activities.

Form 12.

Seasonal calendar of workload for men and women.

Guidelines for semi structured interviews

Household questionnaires for the Bajura baseline survey.

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 1a
time

format
1b

form 1

number of participants
men:
women:

ward #:

village history

Raja)

event

information on village and local leaders (Mukhiya and/or

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 1

number of participants
ward #:

men:
women:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 2
SN

form 2

number of participants
ward #:

men:
women:

compilation of households and well being ranking

name

number of persons category

in the household

ethnic group or
caste

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 3

number of participants
ward #:

men:
women:

form 3
traditional forms of community organizations
Please describe the traditional forms of community organization in the
cluster.
> Is there a Mukhiya or other village leader? Is leadership changing?
> Are there any community activities organized in recent times? Who takes
the initiative?
> How are these organized, who are the members of these organizations? Try
to find out to what extent women and lower castes are included in different
steps of activities (e.g. decision making; implemenetation of activities)?
>What kind of activities do they undertake? What did they do last year?
> Ask specifically about irrigation committees. How are they organized, how
are caretakers selected? Do users have to pay for these services, etc. What is
the total area irrigated under these committees? Who are members of these
irrigation groups?
> If there are any other constructions that indicate involvement of the
community (traditional bridges; foot trails) try to find out about community
organization involvement.

Describe this kind of community organizations below. feel free to use Nepali if
you feel more comfortable about this.

format 3a

description of the traditional organization of the community

format 3b

description of the traditional irrigation systems

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 3

number of participants
ward #:

men:
women:

continue on the backside

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 4

number of participants
ward #:

men:
women:

How many community organizations are active in this cluster?

form 4a
overview of community organizations present in cluster
name of
number of
founded
activities / remarks
community
members
when by
organization
men women

How many members do they have? How is the composition (men/women;

castes)? (form 4b)
Did they start spontaneously or were they encoouraged by outsiders to form
community organizations?
How often do they meet? Do they have a special executive committee?
What are their most important activities?
Do they have saving and or credit activities?
What is their financial situation? How did they collect money? (fill in in format)
How are their relations with other organizations (line agencies, DDC, VDC,
NGOs?)
What are their plans for future activities?
form 4b
group description
started in:
name:
type of group:
group fund: NRs
used for: loans / group activities / not
used / other
activities undertaken until now: (use
extra paper if necessary)

group composition:
men
wome
n

Brahmin
Thakuri
Chettri
lower caste
other
(...........)

total

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 4

number of participants
ward #:

men:
women:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 4

number of participants
ward #:

form 4b
group description
started in:
name:
type of group:
group fund: NRs
used for: loans / group activities / not
used / other
activities undertaken until now: (use
extra paper if necessary)

form 4b
group description
started in:
name:
type of group:
group fund: NRs
used for: loans / group activities / not
used / other
activities undertaken until now: (use
extra paper if necessary)

men:
women:
group composition:
men
wome
n

total

group composition:
men
wome
n

total

Brahmin
Thakuri
Chettri
lower caste
other
(...........)

Brahmin
Thakuri
Chettri
lower caste
other
(...........)

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 4

number of participants
ward #:

men:
women:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 5

number of participants
ward #:

men:
women:

form 5a
drinking water situation in the cluster
1
source of
number of
distance in
drinking water
households using
time for most
this source
users

form 5b
services
primary school
secondary
school
health post
agricultural
office
veterinarian
bank

availability of services (time needed for one way trip)

distance
which services
remarks
from cluster
delivered?

cluster

all season? / accessible

for all castes?

sanitation
number of
toilets

if several sources, are used please mention all

remarks

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 5

number of participants
ward #:

men:
women:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 6

form 6

number of participants
ward #:

men:
women:

characteristics of categories of well being ranking

category A

characteristics

category B

characteristics

category C

characteristics

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 6

number of participants
ward #:

men:
women:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 6

number of participants
ward #:

men:
women:

category D

characteristics

category E

characteristics

category F

characteristics

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 7a

form 7

number of participants
ward #:

men:
women:

trend lines of livestock

100
80
60
40
20
0

explanation:

form 7b

trend line of agricultural production

100
80
60
40
20
0
explanation:

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 8a

form 8

number of participants
men:
women:

ward #:

seasonal calendar of the availability of fodder

100
80
60
40

fagu

chai

fagu

chai

magh

push

mang

kart

asoj

bhad

shra

asha

jest

bais

20

explanation:

form 8b

seasonal calendar of the availability of food

100
80
60
40

magh

push

mang

kart

asoj

bhad

shra

asha

jest

bais

20

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

Baisak Jest

Baisak Jest

Baisak Jest

form 9

number of participants
men:
women:
CROPPING PATTERNS
Khet Land

ward #:

Asar

Asar

Asar

Shraw

Shraw

Shraw

Bhadr

Bhadr

Bhadr

Asoj

Kartik

Mangs Push

Magh

Fagun

Chaitr

Percentag
e

Asoj

Bari Land
Kartik Mangs Push

Magh

Fagun

Chaitr

Percentag
e

Asoj

other
Kartik Mangs Push

Magh

Fagun

Chaitr

Percentag
e

Bajura Baseline study

date:
filled in by:
VDC:
cluster:

form 9

number of participants
ward #:

men:
women:

Bajura Baseline study

form: 10

date:
filled in by:
number of participants: men:
women:

resource mapping
cluster:

ward #:

VDC:
form 10 a

type of natural
resource

availability of different natural resources

availability: abundant,
moderate, scarce?

location (in same

cluster, ward, VDC?)

khet
bari
forest
grazing land
community land
irigation water

form 10 b

drinking water
firewood
fodder
herding animals

use of natural resources

who? Men,
women,
children?

where? cluster,
ward, VDC?

time required
for collection

xxxxxxxxx

ownership (private,
community,
government?)

who else uses

the same
resources?

who controls
these resources?

Bajura Baseline study

date:
filled in by:
number of participants: men:
women:

form: 10

date:
number of participants:

Bajura Baseline study

form: 11

filled in by:
men:
women:

form 11. seasonal calendar of agricultural activities

ward #:
custer:
VDC:
crops
/activities

bais
jest asha shra bhad asoj kart mang push magh fagu chai
may
jun
jul
aug
sep
oct
nov
dec
jan
feb
mar apr

date:
number of participants:

VDC:

form 12

Bajura Baseline study

form: 12

filled in by:
men:
women:

seasonal calendar of workload of men and women

ward #:
cluster:

10
9
8
7
6
5
4
3
2
1
M W M W M W M W M W M W M W M W M W M W M W M W
bais
jest asha shra bhad asoj kart mang push magh fagu chai
may
jun
jul
aug
sep
oct
nov
dec
jan
feb
mar apr

date:
number of participants:

Bajura Baseline study

filled in by:
men:
women:

GUIDELINES FOR SEMI STRUCTURED INTERVIEWS

farming systems
agricultural production
!What are the different types of ownership? (Owned, rented,
sharecropped, kamiya)
! What are the different types of land used by the population of the
cluster? Khet, bari, pakho bari, patan, lekh, etc.
!Is the land irrigated? All seasons/ reliability/ quality/ regularity of the
irrigation system?
!Cropping pattern: which crops are grown during the year; which crops
are intercropped?
!Is crop rotation practiced?
!What are the major crops?
!Is agricultural production consumed in the household or sold? Which
products are sold?
!Is the labor force from the own household large enough or do they hire
laborers from elsewhere?
livestock
!Types of livestock.
!Ownership of livestock
!Stall feeding/ free grazing/ other forms of livestock husbandry;
seasonal migration with the animals?
!Products of livestock. Milk, ghiu, dung, meat, sales of animals.
!Type of fodder used/origin of fodder/quantity and quality available
throughout the year.

GUIDELINES FOR SEMI STRUCTURED INTERVIEW

NATURAL RESOURCES
!What are the different types of natural resources/ forests/ gazing land/
pakho / pakho bari / lekh / patan/ etc.?
(use information from resource mapping)

date:
number of participants:

Bajura Baseline study

filled in by:
men:
women:

! Where do they use these natural resources for?

agriculture?
livestock?
construction materials?
collection of other products (honey, hunting, fish, fruits, medicinal
herbs, edible roots)
! Who collects which products and when is this done?
does it involve seasoanl migration?
!Who controls the forest?
government?
private?
community?
! Are any of these products sold? Which ones and where?
! Is there any (traditional) management system of the abovementioned
natural resources? Is there any sign of over exploitation of certain
products? Do people think that management is necessary (to avoid
overexploitation)?
INCOME GENERATION
! What are the most important sources of income (in cash as well as in
kind)?
Can they rank these sources of income according to importance? (Think
about sale of livestock, agricultural labor, skilled labor in Jajarkot,
migration to India, sale of agricultural production, etc.)
! Who is doing this type of IGA? Men/ women/ lower caste?
! Do women have their own sources of income?

design baseline studies

6.5.