NOTE ON MONITORING AND EVALUATION FRAMEWORKS

Mohsin Bashir, Anam Sethi, Anum Tariq

The purpose of preparing an M&E framework is to come up with a simple image that
shows how and why the project will work. It is the skeleton on which a project is built
and acts as a guide for both internal and external stakeholders. Most M&E frameworks
are pictorial in nature; they require systematic thinking and planning to better describe
projects. The visual representation of the predominant plan in an M&E framework is
flexible, points out areas of strength and flaws in the project, and allows stakeholders to
run through many different scenarios to find the best one.1

Since many development projects are donor funded, M&E frameworks provide an easy
understanding of the use of donor funds. From a donor’s perspective, they assist in
ensuring that funds are used for the intended purpose. They also limit the ambit and lock
the project in conceptually. This is particularly useful as most development projects
involve a large number of beneficiaries, which increases the chances of
digression.2Moreover, M&E frameworks can significantly enhance the participatory role
of stakeholders and the usefulness of evaluation as a management and learning tool.

There are several ways of designing and implementing an M&E framework. Different
programs and donor agencies may create frameworks that best suit their purpose and
requirements. Unlike most available guides that provide instructions for creating
frameworks specific to a donor agency, the generic instructions provided in this note can
be used to formulate a variety of different M&E frameworks. It can also be used by donor
independent organizations to develop indigenous performance management frameworks.
However, keeping in view the demand from training participants3, it includes guidance on
developing two of the most commonly used M&E frameworks in Pakistan, namely
Logical Framework Approach (LFA) and Results Framework (RF).

1Logical Model Development Guide. 2004. In, W.K Kellogg Foundation,

http://www.wkkf.org/knowledge-center/resources/2006/02/wk-kellogg-foundation-logic-model-
development-guide.aspx. (accessed September, 2013)
2Ibid.
3 These frameworks have been chosen based on feedback from focus groups with managers in Pakistani

organizations, most of whom have had to work with either of these frameworks.

1
DESIGNING AN M&E FRAMEWORK

This note adopts a five-step process for designing and preparing an M&E framework.
The steps are:
A. Analyzing the problem
B. Analyzing the solution
C. Selecting performance indicators
D. Identifying opportunities and risks
E. Presenting the framework

A. ANALYZING THE PROBLEM

The first step in creating an M&E framework involves an assessment of the existing
environment in order to identify the specific problem needs to be resolved.An
environment can be a complex web of sociopolitical, environmental, cultural and
economic conditions. A development project—that is planned and implemented within
the environment of the community—must be in line with these conditions and be
conducive to the project’s objectives. It is important to take time to carefully understand
the environment and clearly define the problem. Some useful questions in assessing the
project setting could be: What are the possible causes of the concerned problem/issue? At
what level and for whom is the problem/issue a concern? Who are the relevant
stakeholders?4 Addressing these questions and developing clear answers to them is the
key to creating an effective M&E framework.

While more than one approach can be found on creating M&E frameworks, this note
focuses on identifying a “core problem”5 and using problem trees to conduct a problem
analysis. The problem tree shows cause and effect relationships among problems in a
hierarchical manner. If a problem is a cause of the core problem, or contributes to its
ongoing existence, it is placed below the core problem. On the other hand, if it is an
effect of the core problem, or its existence depends on the existence of the core problem,
it is placed above. If a problem is neither a cause, nor an effect, it is placed at the same

4 Ellen Henert, and Ellen Taylor-Powell, Developing a Logic Model: Teaching and Training Guide. WI:
University of Wisconsin-Extension, Cooperative Extension, Program Development and Evaluation. 2008,
http://www.uwex.edu/ces/pdande/evaluation/pdf/lmguidecomplete.pdf, (accessed November 24, 2013 ).
5
Mdf Tool: Problem Tree Analysis. In,
http://www.toolkitsportdevelopment.org/html/resources/91/910EE48E-350A-47FB-953B-
374221B375CE/03%20Problem%20tree%20analysis.pdf. (accessed November 24, 2013).

2
level. In case a problem has multiple causes, they are placed side by side. Similarly, if
there are multiple effects, they are also placed side by side. Vertical links depict cause-
effect relationships, while horizontal links depict joint causes and effects. This
arrangement and ordering in a problem tree helps visualize which problems lead to or
result from the core problem.6

Figure 1 below is a pictorial representation of a problem tree with “High Infant Mortality
Rate in Bihar” as the core problem. This core problem has been adapted from a project
initiated by the Aga Khan Foundation (United Kingdom) and funded by the Department
for International Development (DFID). The aim of this project is to reduce both neonatal
and infant mortality in the Indian state of Bihar by means of improving breastfeeding and
weaning practices among mothers who live in the poorest and most marginalized
communities. For the convenience of the reader, this example will be used throughout the
note to demonstrate further steps for creating an effective M&E framework.

6
Project Identification and Design: Developing a Logic Model Part 1. In Global M&E Initiative,
http://cmsdata.iucn.org/downloads/module2_developing_logic_model_1_04.pdf. (accessed November 24,
2013).

3
 Figure 1
Problem Tree Analysis: Infant Mortality

Tarnished Image of Increased Reduced Human and

Overall Health Expenditure on Social Capital
Services Medical Care and
Counseling for
New/Expecting
Mothers

High Infant
Mortality Rate
in Bihar

Unhygienic Lack of Hospitals Lack of

Living Conditions Awareness

Limited Outreach
of Health
Facilities

Lack of Lack of Quality Lack of Skilled Limited

Transportation Medication Doctors Knowledge of
Infrastructure Health
Procedures

Source: Adapted from Aga Khan Foundation United Kingdom (2011). “Reducing Infant Mortality in
Bihar through Optimal Infant Feeding Practices.”

4
 B. ANALYZING THE SOLUTION

Once a problem tree is created, solutions to the problems are proposed through an
objectives analysis7, a process that systematically identifies, categorizes, specifies, and
balances out (if required) the objectives of all parties involved in a specific situation.
Negative statements from the problem tree are converted into positive statements
(objectives), and are presented in the form of an objectives tree, also known as a means-
end model. This diagrammatic representation of a project’s objectives shows the
hierarchal relationships between a set of solutions that are derived from the problem
tree8.The top of the tree depicts the desired end or goal of the project, whereas the bottom
levels represent the means to achieving that end. These means offer a base for the
development of activities that address the initial problems. The most appropriate and
feasible path to achieve the objective within the objectives tree is then highlighted.9

Structurally, an objectives tree can be seen as a positive mirror image of the problem tree.
However, it might be necessary to reorder the position of objectives as the tree is
developed. Also, not all objectives may be feasible or realistic within scope of the project
intervention and other solutions might need to be generated. Nonetheless, all objectives
are considered at this stage as they provide a sound foundation for developing
projects.10Figure 2 below shows the objectives tree for the example of infant mortality
illustrated earlier.

7
WHO. Logical Framework Analysis. In,
http://www.who.int/ncd/vision2020_actionplan/documents/LFAUNDP.pdf. (accessed November 25, 2013).
8
Project Identification and Design:Developing a Logic Model Part 1.
9
WHO. Logical Framework Analysis.
10
Ibid.

5
 Figure 2
Objectives Tree Analysis: Infant Mortality

Improved Image of Reduced Increased Human

Overall Health Expenditure on and Social Capital
Services Medical Care and
Counseling for
New/Expecting
Mothers

Reduced Infant
Mortality Rate
in Bihar

Better Standard Improved Increase in

of Living Availability of Awareness
Health Facilities

Improved
Outreach of
Health Facilities

Improved Increased Increased Education on

Transportation Availability of Availability of Health
Infrastructure Quality Skilled Doctors Procedures
Medication

Source: Adapted from Aga Khan Foundation United Kingdom (2011). “Reducing Infant Mortality in
Bihar through Optimal Infant Feeding Practices.”

6
While several guides to making M&E frameworks recommend going straight from the
objectives tree to the framework itself, the inclusion of another step, called “Developing a
Causal Chain”, can be very useful. A causal chain is a sequential diagram that maps out
the intended pathway to achieve the objectives of a project. It separates these objectives
into long-term impacts, overall purpose of the project, population level results (required
to achieve the overall purpose), project level activities (required to achieve the population
level results), and inputs (required for carrying out the project level activities).

For example, if your overall long-term goal is to improve the situation of education of
education in Pakistan and you want to set up a school, you first need funds, infrastructure,
buildings, instructors, stationary and so on. These are your inputs. Once you have your
essential inputs, you can carry out project-level activities, such as conducting classes,
contacting parents, creating the curriculum, and training teachers. From all these
activities, you may bring out some changes in the population you are serving. For
instance, by conducting classes, you may hope to enhance the subject knowledge and
reading and writing abilities of the students. These are your population-level results.

It is important to note that project level results are directly under your control whereas
population level results are not. So, you can ensure that classes take place, but you cannot
control whether the students will retain the learning imparted to them in class. That is
something you hope to be a consequence of the successful carrying out of the project-
level activities. Eventually, if the activities are persistent and effective in bringing about
the intended population level results, you may be able to fulfill the overall aim of the
project in the medium to long term i.e. if you opened the school to increase the attainment
of primary education in a certain village, it can be achieved once the number of children
have graduated from the school. Finally, the completion of the project’s objectives will
give some assurance that you are contributing to the larger and longer-term goal of
improving the situation of education in Pakistan.

Figure 3 below shows one such causal chain.

7
 Figure 3
Causal Chain

Activties and Solution to the

Population-
Project-Level Focal Problem
Inputs Level Results Ultimate Goal
Results (Medium-Long
(Medium Term)
(Short Term) Term)

There is no set terminology for the elements that form the causal chain; different
frameworks use their own terminology. Table 1, termed “M&E Framework
Terminology”, provides a comparison of the terminology used in this note with that
employed by other frameworks11.
Table 1
M&E Framework Terminology

M&E Framework Terminology

Logical Framework Results Framework (RF)

Step/Activity/Process
Approach (LFA) Terms Terms

Inputs Inputs Inputs

Intermediate Results
Project-Level
Outputs (sometimes referred to as
Results/Activities
Sub-IR)
Population-Level
Outcomes Intermediate Results (IR)
Results/Activities
Solutions to Focal
Purpose Strategic Objective (SO)
Problem

Ultimate Goal Impact Goal

11
Adapted from Developing Human Rights Based Ea2j Interventions. In Manual on HRBA2J, Asian
Consortium for human rights-based access to justice, http://www.hrbajustice.asia/wp-
content/uploads/2012/06/Module-3-pp.-197-254.pdf. (accessedNovemeber 24, 2013).

8
 C. SELECTING PERFORMANCE INDICATORS

Once the causal chain is established, the next step is to define the immediate and
intermediate results in terms of measurable performance indicators. These indicators give
a plan of how the M&E framework will be operated to monitor progress and assess the
effects of the intervention. This section develops an understanding of the significance,
levels and types of indicators.

Significance of Indicators
Formulating performance indicators is fundamental to monitoring and evaluation as the
indicators keep track of changes throughout the project and provide measures of expected
results. Particularly, indicators help to:

• Make informed decisions for ongoing projects

• Measure growth and success from the perspective of various stakeholders
• Ensure consistency among activities, outcomes, purpose and impacts
• Demonstrate progress achieved to all stakeholders and maintain accountability
• Assess performance of project and staff12

Several guidelines are available for gauging the quality of indicators and selecting the
best ones. It is generally recommended that the performance indicators follow the
“SMART” criteria, often used in the Management by Objective style of performance
management. That is, indicators should be specific to intended results; measurable, in
either quantitative or qualitative terms; achievable within the allotted time and resources;
relevant to the intended results; and time-bound13.

While these guidelines show what characteristics a “good” indicator can have, selecting
the optimal indicator is a subjective decision. Basically, the best indicator is the one that
is best aligned with the requirements of management. For any given component of the
causal chain, there can be several indicators. Choosing among these indicators can be
tricky, especially when each indicator has its strengths and weaknesses. Exhibit A
provides a more detailed analysis of what constitutes a ‘good’ or ‘bad’ indicator, based
on criteria developed by the World Bank.

12
UNDP, ‘Handbook on Monitoring and Evaluating for Results’,2009, p 61-73.
13
UNDP, ‘Handbook on Monitoring and Evaluating for Results’,2009, p 57-58.

9
Levels of Indicators
Usually indicators are associated with each level along the causal chain. Each level has
an objective statement that sets a specific target to be achieved and is measured by its
corresponding indicator(s). Matching objectives to their respective indicators ensures
availability of relevant sources of data for program evaluation. There can be several
levels of indicators, depending on the logic model used, but the three most frequently
used indicators are described below.14
• Indicators for “Long-Term Impact” provide a measure of change in the living and
development conditions of people at different levels. These levels are defined by
the nature of the initiative and hence can be at a community, subnational, national,
regional, or global level. Examples of Long-Term Impact indicators include
national literacy rate, number of countries with predominantly democratic
governments, and number of people living below poverty line in Province A etc.
• Indicators for “Immediate Outcomes at the Population-Level” measure progress
against specified results to ensure whether the expected change has taken place or
not. Such indicators include national literacy rate after implementation of Project
X, number of countries willing to accept democracy after conference, and number
of people living below poverty line after hunger drive etc
• Indicators for “Project-Level Activities” monitor the execution of a project’s
activities and provide a measure of the activities’ quantity, duration and program
efficiency 15 .Such indicators include number of citizens attending education
awareness sessions, number of countries represented in conference, and number
of social workers volunteering at hunger drive etc.

Types of Indicators16
There can be multiple indicators of one outcome and these indicators can be changed
during the implementation process as more data becomes available. The different types of
indicators that can be used in a project are listed as follows:

14
UNDP, ‘Handbook on Planning, Monitoring and Evaluating for Results’,2009.
15
USAID.Selecting Performance Indicators Performance Monitoring and Evaluation Tips. Washington
DC: Center for Development Information and Evaluation,
2010.http://pdf.usaid.gov/pdf_docs/pnadw106.pdf, (accessed November 27, 2013).
16
UNDP, ‘Handbook on Planning, Monitoring and Evaluating for Results’,2009.

10
 • Qualitative indicators
These reflect judgments, opinions, perceptions and attitudes of people towards a
particular matter. They assess changes in “sensitivity, satisfaction, influence,
awareness, understanding, attitudes, quality, perception, dialogue or sense of well-
being.”17Qualitative indicators give a measure of results in terms of compliance,
quality, extent or level, such as the existing quality of health services provided to
mothers in Bihar, which could be indicated in terms of poor, average, good, and
excellent.
• Quantitative indicators
These are statistical measures that assess changes or results in terms of numbers,
percentages, rates, or ratios. Examples of quantitative indicators include number
of infant death per 1000 population, percentage of social workers who are CPR
certified, ratio of men to women etc.
• Proxy indicators

These do not provide direct measures of progress and can be used to make probable
assumptions about the progress of the project. Proxy indicators are usually employed
when suitable data is unavailable or the system is incapable of monitoring results
captured by direct indicators. For example, development organizations such as the United
Nations Development Programme (UNDP) use ‘life expectancy’ as a proxy indicator to
measure the quality of health services, as it can be assumed that if people are living
longer, then health services have improved18.

Baselines, Targets and Data Collection19

After the selection of indicators, baselines and targets need to be established for the
intended level of change. Baseline is the initial value of the indicator (for an input,
activity, output, purpose and impact) obtained from the baseline data. The baseline study
is an early element in the monitoring and evaluation plan of the project. It systematically
assesses the situation before the project intervention and provides the basis for measuring
progress after the intervention. A baseline helps demonstrate whether the activities
translate into the desired outcomes. Any deviation would suggest that either there is a
need to change the activities to achieve the desired outcomes, or a need to change the
expected outcomes to relate better with the activities.

17
UNDP, ‘Handbook on Monitoring and Evaluating for Results’,2009.
18
UNDP, ‘Handbook on Monitoring and Evaluating for Results’,2009.
19Ibid.

11
With the baseline established, a target needs to be set. A target is the final value expected
to be achieved by the end of project. It usually depends on the duration of the
interventions, activities and overall project period. For the purpose of monitoring and
evaluation, it is important for the baseline and target to be aligned with the indicators,
using same unit of measurement.

Finally, identification of indicators requires identifying the data collection methods. For
example, data for one indicator might need to be collected through direct surveys, while
another indicators’ data might be available in the organizations audit reports, and so on.

D. IDENTIFYING ASSUMPTIONS OF OPPORTUNITIES AND RISKS

Most M&E frameworks require listing critical assumptions taken while making the
causal chain. These assumptions provide a list of risks and opportunities that can affect
the translation of a project’s activities into its intended outcomes and so on. Figure 4
provides a conceptualization of the Figure 3’s causal chain when assumptions are taken
into account.

12
 Figure 4
Relationship between Components and Assumptions

Activities and Population Solutions to Ultimate Goal

Project Level Level Results the Focal
Results (Medium Problem
(Short Term) Term) (Medium to
Long Term)
Assumption Assumption Assumption
s s s

Obvious assumptions, which are either not important or highly unlikely to occur, are
discarded. Assumptions that are quite likely but uncertain are kept and subsequently
monitored. If an assumption is both significant but unlikely to occur, it is called a killing
factor. In this case the project either has to be redesigned or decommissioned.20

It is important to note that assumptions must be limited to relatively likely circumstances

outside of the control of the project team. For example, if the activity is to run an
awareness campaign and the intended outcome is for people to be aware of a certain
cause, “the message might not reach all the people” will be a bad assumption. This is
because the project team should have made sure that the awareness campaign had a
widespread coverage.

Having discussed a general method of developing M&E frameworks, the next section
focuses on two commonly used formats for monitoring and evaluation; Results
Framework (RF) and Logical Framework Approach (LFA).

E. PRESENTING THE FRAMEWORK

The final step in preparing an M&E framework is choosing a presentation format. A

number of M&E frameworks have been developed by different organizations based on
their needs and requirements. Each of these frameworks is essentially a presentation of
information collected through the four steps identified earlier. For example, many

20
Ibid.

13
international development agencies, such as DFID, Swedish International Development
Cooperation Agency (SIDA), and United Nations Development Programme (UNDP),
prefer to use the LFA. RF is also becoming increasingly popular and has been adopted by
United States Agency for International Development (USAID), Canadian International
Development Agency (CIDA), and World Bank, among others, as their M&E framework
of choice. Some other agencies, such as Amnesty International and United National
Developlment Group (UNDG), have been known to use other M&E frameworks, such as
Theory of Change (Toc) and Results Based Management (RBM) respectively.

This note will focus on designing an LFA, based on DFID guidelines, and an RF, based
on USAID guidelines, as most development projects in Pakistan use one of these
frameworks for their monitoring, evaluation, and reporting21. Guidance on designing a
traditional LFA is also provided. However, as mentioned earlier, this can also be used as
a generic guide to design other frameworks based on the step-by-step direction and tools
provided.

1. DFID’S LOGICAL FRAMEWORK APPROACH (LFA)

Logical Framework Approach is one of the earliest M&E frameworks to be used in

program design. It forms part of a systematic planning procedure, which condenses the
planned project mechanism into a relatively simple and linear M&E framework.
Moreover, it forms the basis on which the implementing agencies plan their monitoring
and evaluation system. Used as a problem solving mechanism, it further takes into
account the views of all the relevant stakeholders so that they can unanimously identify
and agree on criteria for project success and list down all major assumptions.

This note focuses on designing a Logical Framework Matrix based on DFID guidelines.
In 2009, DFID revised the Logical Framework format in order to make it more
comprehensive and accessible in nature. Diagrammatically, the DFID LFA takes the form
of a table that is divided into a series of rows and columns to represent the impact,
outcome, and outputs of a particular project. A separate sheet is often attached to list the
activities associated with each output. Exhibit B and C contain an example of the DFID
logical framework matrix and a template of the activity log respectively.

21
Based on focus group discussions with six cohorts of M&E managers from all over Pakistan.

14
Setting the Goal, Purpose, Output and Activities
The first step in designing the DFID LFA is defining its key components. These include
ultimate goal or impact, purpose, outcomes, and outputs/activities. Each of these
components can be translated into the elements of the causal chain shown in Figure 3.
The LFA ‘impact’ translates into the ultimate goal; the ‘purpose’ translates into solutions
to the focal problem; the ‘outcomes’ translate to population level results; the ‘outputs’
translate into project-level activities; and finally, inputs are segregated into inputs specific
to outputs and inputs specific to activities.

Assigning Indicators (OVI)

After all the components are identified, indicators for each one of them are developed, as
discussed in Section C. Indicators are further divided into baseline and target indicators,
as well as milestones, in order to give a better representation of what needs to be
achieved at what point in the project. Milestones track the trajectory of the project and
should be set at different intervals depending on the project timeline.

Establishing Source
Once indicators are developed, it is important to consider how the monitoring and
evaluation procedures will be used to gather data. Sources of obtaining information and
means of data collection need to be established for each indicator, as described in Section
C. In the case of the DFID LFA, these means of verification are titled Source, and
represent the methods used to obtain information for each indicator. This is done to
ensure that indicators are based on data that is measurable and independently verifiable,
and to make certain that indicators measure exactly what they are designed for.
Impact Weighting
At each output level, impact weighting is shown together with risk rating for individual
outputs. Impact weighting assigns a percentage to the contribution an output makes
towards the eventual outcome of the project, whereas risk rating refers to the risk matrix
and is recorded as low, medium or high.
Inputs
Inputs are also stated for each output to record the quantity and quality of resources used
to achieve the final product. They are also beneficial in keeping track of budgets, and
help ensure that the benefits derived from the outputs of the project outweigh the costs of
implementing it.

Specifying Assumptions

15
Identifying assumptions is the last step in designing an LFA. Clear logical links between
components of an LFA can be broken down by external factors outside the control of the
project, in accordance with our discussion in Section D. Therefore, assumptions are
stated following an if-and-then logic at each level of the matrix, as shown in Figure 4
below. For instance, activities deliver outputs if the assumptions at this level hold true.
Assumptions are part of the vertical logic in LFA and provide the necessary and
sufficient conditions at each level for achieving the level above. They are stated as
positive conditions and described in operational detail.22 For DFID LFA, assumptions are
shown for impacts, outcomes, and outputs.

LOGICAL FRAMEWORK APPROACH (LFA)

The major difference between a traditional LFA and a DFID LFA is that a traditional
LFA is less complex in nature. A traditional LFA is presented in the form of a 4x4
matrix. The vertical axis contains Activities, Outputs, Purpose and Goal, which represent
the Project Level Activities, Population Level Results, Solutions to the Focal Problem
and Long Term Impacts/Ultimate Goal from Figure 3 respectively. Unlike DFID LFA, a
traditional LFA has all activities listed for all outputs. Also, the inputs are optional in this
matrix. On the horizontal axis, what the DFID LFA refers to as “Source” becomes
“Means of Verification (MOV)”; and indicators are termed “Objectively Verifiable
Indicators (OVI)”. The OVI section contains targets for each indicator, however baselines
and intermediate milestones are not required. A template of the traditional LFA matrix
(along with the optional “Inputs” section) is shown in the appendix as Exhibit D.

2. RESULTS FRAMEWORK (RF)

A Results Framework is a clear articulation of the causal chain (presented in Figure 3),
which shows how the program’s objective(s) can be achieved. The framework represents
cause and effect relationships between the objective and a series of intermediate results.
Critical assumptions, that are necessary for the development hypothesis to hold, are
noted. In addition, the framework establishes an evidence-based approach to monitoring
and evaluation by including specific indicators of impacts and outcomes, and identifying
baselines and targets.23 As in LFA, results at different levels are presented. An example

22NORAD. 1999. The Logical Framework Approach (Lfa) Handbook for Objectives-Oriented Planning.
http://www.norad.no/en/tools-and-publications/publications/publication?key=109408. (accessedNovemeber
25, 2013).
23 Roberts, Dawn, and Nidhi Khattri."Designing a Results Framework for Achieving Results: A How-to

Guide."World Bank (2012).

16
of representing a project through a Results Framework is given in Figure 5 below.
Exhibit E demonstrates how a Results Framework matrix can be made for the program
aimed at reducing infant mortality.
Figure 5
An Example of a Results Framework

Goal
To reduce neonatal and infant mortality (NMR &
IMR) in selected blocks of three districts of Bihar
(and through demonstration of and advocacy for an
appropriate model for statewide adoption)

Strategic Objective
To increase the adoption of one or more recommended IYCF
practices by 300,000+ mothers of children under two years of
age, among the poorest and most marginalised populations in
the three districts of Bihar.

IR 1: IR 2: IR 3:
Pregnant women, Community-based Contribute to
mothers of children front line workers appropriate policy
under two years of age (ANM, AWW and improvements and
and other care-givers ASHA), TBAs and programme
in the family have the facility based health implementation
appropriate knowledge service providers changes instituted
and develop requisite (staff nurses/ Mamtas) with focus on IYCF
skills to follow the have increased counselling,
recommended IYCF knowledge levels and including message
practices, through the perform better in the delivery and
messages and context of IYCF problem solving,
assistance imparted by counseling. following advocacy
peer educators in for the same.
selected blocks of
three districts of
Bihar.

Source: Adapted from Aga Khan Foundation United Kingdom (2011). “Reducing Infant Mortality in
Bihar through Optimal Infant Feeding Practices.”

17
Setting the Goal, Objective and Results
The Results Framework essentially uses the same components as those in Figure 3 but
terms them differently. The main components of a Results Framework are: goal (ultimate
goal/long term impact), strategic objectives (solutions to the focal problem), intermediate
results (population level results) and sub-intermediate results (project level activities).
Results Framework focuses more on these components as compared to project inputs and
implementation processes. Strategic objectives and intermediate results are discussed in
greater detail below to help understand their significance to RF.

i) Establishing a Strategic Objective (SO)24

The strategic objective (SO) is the focal point of the Results Framework, and should be in
accordance with our analysis of the causal chain. Setting the strategic objective is a
critical task because it sets a standard against which the project is to be assessed in terms
of its performance, and also forms the basis of the management contract. Therefore, the
SO must reflect a balance between the accountability and ambition of the implementing
agency in terms of its envisioned objective. Consequently, it is also a difficult task since
the decision of setting the strategic objective needs to be based on a realistic assessment
of what can be achieved by the project within the time frame and the given resources. It
also needs to be consistent with the agency’s strategic plan and be stated precisely,
clearly and in strong action words. Most importantly, it should emphasize the results of
the action and not the action itself.

ii) Identifying Intermediate Results (IRs)

Once the strategic objective is set, a set of ‘lower level’ results need to be identified
which may be defined in terms of intermediate results (IRs). These critical series of
results establish the causal links that bridge the gap between the current status of the
program and the desired higher level result, as depicted in the causal chain. It is important
to indicate which units of the implementing agency are involved in achieving which
outcome and how the outcome will be effectively measured.25

24USAID. Building a Results Framework Performance Monitoring and Evaluation Tips. Washington DC:
Center for Development Information and Evaluation, 2000.http://pdf.usaid.gov/pdf_docs/PNACA947.pdf,
(accessed November 25, 2013).
25Roberts, Dawn, and Nidhi Khattri."Designing a Results Framework for Achieving Results: A How-to

Guide."World Bank (2012).

18
After the identification of intermediate results, the links between results and objective
need to be clarified. For this purpose, linkages need to meet two criteria. First,
relationships between two results, or between a result and an objective, must be causal.
Causality implies that achievement of one result is not only necessary but also contributes
to the achievement of the other. Second, the causal linkages between two results, or a
result and an objective, must be direct. It is important to note that one intermediate result
can have multiple causal links as it can affect the achievement of other results too. Also,
linkages are usually directed upwards in a Results Framework; results with a lower
impact level typically help achieve results with a higher impact level. Nevertheless, the
links can move in any direction as long as they contribute to the achievement of the
strategic objective. It is also possible for links to exist across Results Frameworks for
different SOs.26 Hence, to make effective use of Results Framework it is critical to make
these connections clear.

iii) Selecting Action Statements

When constructing SOs and IRs, it is essential that they be worded in a clear and precise
manner in conjunction with the activities and goals of the project at hand. When it comes
to formulating these action statements, it is important ensure that the following criteria
are met:

• Action statements should be aligned with the outcomes of the project. Therefore,
they should express the results of an action, rather than expressing the action
itself.
• Action statements should be objectively measurable and open to only one
interpretation.
• Action statements should be one-dimensional and should focus on one
component.
• Action statements should be sensitive to the practical considerations of time and
space.

Table 2 provides guidelines for how a ‘good’ action statement should be constructed.

26
USAID. Building a Results Framework Performance Monitoring and Evaluation Tips. Washington DC:
Center for Development Information and Evaluation, 2000.http://pdf.usaid.gov/pdf_docs/PNACA947.pdf,
(accessed November 25, 2013).

19
 Table 2
Action Statement Construction

Statement Selection Good vs. Example Reasoning

Criteria Bad
Results Statements Good Increased economic Clearly expresses an
growth in target sectors outcome, or the result of
an action.
Bad Increased promotion of Does not express an
market-oriented policies outcome; is more
process-oriented
Clear and Measurable Good Increased level of Can be objectively
investment measured
Bad Increased ability of Unclear; open to more
entrepreneurs to respond than one interpretation
to an improved policy,
legal, or regulatory
environment
One-dimensional Good Families with an annual Clear and one-
income above dimensional
$40,000/year
Bad Healthier, better Contains diverse
educated, higher income components and is
families multidimensional
Specifies time frame Good Provide immunization to Helps determine if action
all families in district A is reasonable and feasible
by 2016
Bad Provide immunization to Open-ended; no clear
all families in district A time frame; cannot asses
whether action is
achievable or not

Source: Adapted from USAID (2010). “Performance Monitoring and Evaluation TIPS: Building a Results
Framework”. 2nd Edition, Page 4;
IEG, World Bank (2012). “Designing a Results Framework for Achieving Results: A How-To Guide”.
Page 37.

Assigning Indicators and Developing Performance Monitoring Plan

Once the operating team has carried out the steps above, it can then develop a set of
performance indicators and data sources for each level of results, as described in Section
C. It should be noted that multiple indicators can be required to measure the qualitative
and quantitative aspects of an objective or intermediate outcome. Also, only a sufficient
number of indicators should be used as a lot of information is not necessarily

20
helpful27.Finally, baselines and targets are set to operationalize the plan for monitoring
performance and evaluating effects of interventions.

After finalizing the Performance Monitoring Plan, the details should be captured in a
Performance Management Framework (PMF). PMF is a framework used to
systematically plan the various data collection tasks that need to be carried out for
monitoring, evaluation, and reporting. See Exhibit E for further details on how to
construct a PMF. Another good idea at this point will be to have a Indicator Performance
Tracking Table (IPTT), a tool used to track and document the progress being made to
address each indicator at the Impact, Outcome, and Output levels. Please refer to Exhibit
F for a template of an IPTT.

Finally, USAID requires that, for each indicator, a Performance Indicator Reference
Sheet (PIRS) be created. The PIRS acts as a summary tool that shows how the
performance monitoring plan is implemented on the ground. Each indicator in the
framework has its individual PIRS. There is no one defined format of PIRS; its form and
content can be customized to fit the project requirements. Exhibit G provides one sample
of what a PIRS can look like.28

Specifying Critical Assumptions

The next step is to specify the critical assumptions. As explained section D, assumptions
are related to external factors that carry a certain level of risk for each intermediate and
immediate result in the results chain. Stated assumptions should not have a low
probability of holding true otherwise the development strategy (devised to achieve the
objective) will not be effective.29 These assumptions determine whether the intermediate
result identified is a realistic means of achieving the end result or not.

27
UNDP, ‘Handbook on Planning, Monitoring and Evaluating for Results’,2009.
28
Microlinks-Microenterprise Learning Information & Knowledge Sharing, Performance Indicator
Reference Sheet. In, http://www.microlinks.org/good-practice-center/value-chain-wiki/performance-
indicator-reference-sheet. (accessedNovemeber 25, 2013).

29 Ibid.

21
CONCLUSION

M&E evaluation is often misunderstood as a compliance requirement by a donor.

However, through the steps identified in this note, it is evident that an M&E framework is
essentially a powerful performance management tool. The significance of such a
framework is even stronger in development projects because the long linkages between
activities and impacts form the basis for existence of such projects, in the absence of
financial bottom lines.

It is important to note that M&E frameworks are not static documents. They should be
dynamic and flexible enough to incorporate changes during the course of the project
intervention. An M&E framework for any project should be continually reviewed, re-
examined and revised in light of the results of the evaluations or other insights that reflect
a deeper understanding of the assumptions of the project, its external environment, or
how a project is operating.

22
 Exhibit A

Good Versus Bad Indicators

Indicator Selection Good vs. Example Reasoning

Criteria Bad
Measurable Good Number of human rights Can be measured
violations quantitatively; numerical
precision makes data
interpretation easier
Bad Types of human rights Cannot be measured with
violations numerical precision; may
be used as a supplement
to other indicators to
provide rich information
Practical Good % of targeted population Data can be collected on
that understand their a timely basis; reasonable
voting rights in cost as it will not
(representative sample) require more than 3 to 10
percent of total program
costs
Bad % of targeted population Data cannot be collected
that understand their on a timely basis;
voting rights (census) unreasonable in cost as it
will require more than 3
to 10 percent of total
program costs
Reliable Good Number of people who Data can be measured
were tested for HIV at repeatedly with precision
work (by visiting the by different people
doctor) in the last 12
months
Bad Number of people Data cannot be measured
receiving quality care repeatedly with precision
and support services by different people
through workplace
programs
Relevant Good HIV infection rates in Result of indicator, to
targeted communities some extent, is caused by
project-supported
activities; link between
project output and
indicator results is clear
Bad National HIV infection Result of indicator is not
rates caused by project-
supported activities; link
between project output

23
 and indicator results is
not clear
Useful to Management Good Amount by type of Information provided is
resources mobilized critical to support
program management
decisions
Bad Number of staff Information provided
meetings may not be critical to
support program
management decisions
Direct Good Number of teenagers Indicator closely tracks
who reported delaying the result it is intended to
sexual activity until at measure i.e. increased
least XX years of age HIV/AIDS prevention
practices among young
adults
Bad Number of youth Indicator does not closely
counselors trained in track the result it is
HIV prevention intended to measure i.e.
messages increased HIV/AIDS
prevention practices
among young adults;
youth counselors may be
trained but may not
deliver better services
Sensitive Good Number of sterilizing Serves as an early
equipment working at warning for changing
hospitals (early warning conditions
of the potential lack of
sterilization of
equipment)
Bad Number of new HIV Does not serve as an
infections at hospitals early warning for
due to lack of changing conditions; will
sterilization of take considerable time
equipment and resources to measure
Responsive Good Number of communities Indicator reflects change
who have implemented as a result of project
at least one BCC activities
methodology/activity
Bad Level of behaviour A more abstract indicator
change
Objective Good Number of PLWHA Operationally precise;
support groups one-dimensional; not
experiencing an annual ambiguous
increase in membership
of at least 5%

24
 Bad Number of expanding Interpretation of results is
and successful PLWHA required
support groups

Capable of being Good Gender, age, location, Data can be broken down
disaggregated ethic group by gender, age etc.; can
help track whether or not
specific groups
participate in and benefit
from activities targeted at
them
Bad Worldwide, national Gives a broad overview
but cannot be broken
down to specifics

Source: Adapted from World Bank (2008). “Planning for, Measuring and Achieving HIV Results: A
Handbook for Task Team Leaders of World Bank lending operations with HIV components”, Annex N,
Page 137.

25
 Exhibit B

Logical Framework Approach (LFA) Template

ORGANISATION Aga Khan Foundation (United Kingdom)

NAME

PROJECT Reducing Infant Mortality in Bihar through Optimal Infant Feeding Practices
NAME
IMPACT Impact Baseline Mileston Mileston Target Assumptions
Indicator 1 e1 e2 (date) What are some of the
assumptions you will
Insert the Add first Planned How much How How What is be making whilst
ultimate goal of indicator. work has much of much of the undertaking this
the project Further already been the work the work projected project? Answer may
indicators may accomplishe with with completion be given in list form.
be added as d with regards regards date of this
needed. regards to to this to this project?
this specific specific
indicator? indicator indicator
Answer may do you do you
be given as hope to hope to
a complete complete
percentage by this by this
e.g. 15%. date? date?
Answer Answer
may be may be
given as given as
a a
percenta percenta
ge e.g. ge e.g.
40%. 75%.
Achieved

Source

For any information derived from external sources,

please give a reference here

Impact Baseline Milestone Milestone Target

Indicator 2 1 2 (date)

Planned
Achieved

Source

OUTCOME Outcome Baseline Milestone Milestone Target Assumptions

Indicator 1 1 2 (date)

Insert population Planned

level
results/activities
Achieved

Source

26
 Outcome Baseline Milestone Milestone Target
Indicator 2 1 2 (date)

Planned

Achieved

Source

OUTPUT 1 Output Baseline Milestone Milestone Target Assumption

Indicator 1.1 1 2 (date)
Insert project Planned .
level
results/activities.
Further Outputs
Achieved
maybe added as
needed. Source

IMPACT Output Baseline Milestone Milestone Target

WEIGHTING (%) Indicator 1.2 1 2 (date)

Planned

Achieved

Source RISK RATING

INPUTS (£) DFID (£) Govt (£) Other (£) Total (£) DFID SHARE (%)

INPUTS (HR) DFID (FTEs)

Revised DFID (Post 2009) Format for LFA: Infant Mortality

Source: Aga Khan Foundation United Kingdom (2011). “Reducing Infant Mortality in Bihar through
Optimal Infant Feeding Practices.”

27
 Exhibit C

Revised DFID (Post 2009) Format for LFA: Optional Activity Log

OUTPUT 1 ACTIVITY 1.1 Milestone 1 Milestone 2 Milestone 3 Risks Monitoring

Officer

ACTIVITY 1.2 Milestone 1 Milestone 2 Milestone 3 Monitoring

Officer

ACTIVITY 1.3 Milestone 1 Milestone 2 Milestone 3 Monitoring

Officer

28
 Exhibit D

Traditional Format for Logical Framework Matrix

Project Objectively Verifiable Means of Assumptions

Objectives Indicators Verification (What else to be
(What does the (How to measure (Where/how to get aware of?)
project want to change?) information?)
achieve?)
Goal

Purpose

Outputs

Activities

Inputs

29
 Exhibit E
Performance Management Framework

Performance Data Source Data Collection Data Reporting Frequency Reporting

Indicator Method Disaggregatio Responsibility
n
Goal: What is the ultimate goal of the project?
G1 Insert first Where was How was further data How was data How often was Which
performance information collected? E.g., Third organized? E.g., information collected? department was
indicator at this regarding this Party Evaluation Gender or E.g., annually. responsible for
level. Further indicator Region. reporting
indicators may be collected? E.g., information
added as needed. Survey Results, collected? E.g.,
website. M&E
Department
G2
IR 1: Vocational Training Opportunities Created
1.1
1.2
Sub-IR 1.1: Potential Trainees Identified
1.1.1
Sub-IR 1.2: Training Opportunities Communicated
1.2.1
IR 2:
2.1
Sub-IR 2.1: Skills of Selected Youth Enhanced
2.1.1
Sub-IR 2.2:
2.2.1

30
 Exhibit F
Indicator Performance Tracking Table (IPTT)

Indicator Baseline Year 1 Year 2: Midterm Year 3 Year 4: Final Year 5

Evaluation Evaluation
Exp Act Ratio Exp Act Ratio Exp Act Ratio Exp Act Ratio Exp Act Ratio
IMPACT
Impact Targets Actual Percentage
Indicator the progress of
1 project towards achievement
s hopes targets in given
or year against
expects target
to
achieve
Impact
Indicator
2
OUTCOME
Outcome
Indicator
1
OUTPUT 1
Output
Indicator
1.1
Source: Adapted from USAID (2008). ‘IPTT Guidelines: Guidelines and Tools for the Preparation and Use of Indicator Performance Tracking Tables.’

31
 Exhibit G
Performance Indicator Reference Sheet

Source: Sharing, Microlinks-Microenterprise Learning Information & Knowledge. Performance Indicator Reference
Sheet. In, http://www.microlinks.org/good-practice-center/value-chain-wiki/performance-indicator-reference-sheet.
(accessedNovemeber 25, 2013).

32
 References

Aga Khan Foundation United Kingdom (2011). “Reducing Infant Mortality in Bihar through
Optimal Infant Feeding Practices.”
http://www.akdn.org/akf_dfid/AKFUK_DFID_GPAF_Revised_Logframe_India.pdf
(accessed 12 November 2014)

Developing Human Rights Based Ea2j Interventions. In Manual on HRBA2J, Asian

Consortium for human rights-based access to justice, http://www.hrbajustice.asia/wp-
content/uploads/2012/06/Module-3-pp.-197-254.pdf. (accessed September 7, 2013).

"Guidance on Using the Revised Format."InDFID Practice paper, edited by Department for
International Development: UKAID, 2009.

Henert, Ellen, and Ellen Taylor-Powell.Developing a Logic Model: Teaching and Training
Guide. WI: University of Wisconsin-Extension, Cooperative Extension, Program
Development and Evaluation. 2008.
http://www.uwex.edu/ces/pdande/evaluation/pdf/lmguidecomplete.pdf. (accessed
November 24, 2013).

IEG, World Bank (2012). “Designing a Results Framework for Achieving Results: A How-To
Guide”. Page 37.

Logical Model Development Guide. 2004. In, W.K Kellogg Foundation,

http://www.wkkf.org/knowledge-center/resources/2006/02/wk-kellogg-foundation-logic-
model-development-guide.aspx. (accessed September 7, 2013)

Mdf Tool: Problem Tree Analysis. In,

http://www.toolkitsportdevelopment.org/html/resources/91/910EE48E-350A-47FB-
953B-374221B375CE/03%20Problem%20tree%20analysis.pdf. (accessed November 24,
2013).

NORAD. 1999. The Logical Framework Approach (Lfa) Handbook for Objectives-Oriented
Planning. In, Fourth, http://www.norad.no/en/tools-and-
publications/publications/publication?key=109408. (accessedNovemeber 25, 2013).

Project Identification and Design:Developing a Logic Model Part 1. In Global M&E Initiative,
http://cmsdata.iucn.org/downloads/module2_developing_logic_model_1_04.pdf.
(accessed November 24, 2013).

Roberts, Dawn, and NidhiKhattri."Designing a Results Framework for Achieving Results: A

How-to Guide."World Bank (2012).

Sharing, Microlinks-Microenterprise Learning Information & Knowledge. Performance Indicator

Reference Sheet. In, http://www.microlinks.org/good-practice-center/value-chain-
wiki/performance-indicator-reference-sheet. (accessedNovemeber 25, 2013).

33
UNDP. 2009. Handbook on Planning, Monitoring and Evaluating for Results. In, Evaluation
Office, NY, http://web.undp.org/evaluation/handbook. (accessed November 24, 2013).

USAID. Building a Results Framework Performance Monitoring and Evaluation Tips.

Washington DC: Center for Development Information and Evaluation, 2000.

USAID (2008). ‘IPTT Guidelines: Guidelines and Tools for the Preparation and Use of Indicator
Performance Tracking Tables.’ http://pdf.usaid.gov/pdf_docs/Pnadq294.pdf

USAID (2010). “Performance Monitoring and Evaluation TIPS: Building a Results Framework”.
2nd Edition, Page 4.
USAID. Selecting Performance Indicators Performance Monitoring and Evaluation Tips.
Washington DC: Center for Development Information and Evaluation, 2010.
http://pdf.usaid.gov/pdf_docs/pnadw106.pdf, (accessed November 27, 2013).

WHO. Logical Framework Analysis. In,

http://www.who.int/ncd/vision2020_actionplan/documents/LFAUNDP.pdf. (accessed
November 25, 2013).

World Bank (2008). “Planning for, Measuring and Achieving HIV Results: A Handbook for
Task Team Leaders of World Bank lending operations with HIV components”, Annex N

34