3.1. Hierarchy of Activities and Objectives
3.2. Assumptions
3.3. Indicators
3.4. Milestones
The CGIAR and its collaborating partners implement agricultural research through projects. Therefore, a planning tool is needed which is
· output and impact oriented,
· increases consistency, accountability and transparency,
· facilitates co-operation and
· can be used by planners and decision makers at different levels.
The logframe approach if used in a flexible and participatory manner will fulfil these requirements.
The essentials of a logframe planning are documented in a matrix form either for a particular project, larger programme, individual centre, or the CGIAR system as a whole in a very condensed form. This diagram does not replace the process of elaborating strategies and targets but gives a consistent overview of the results of such a planning process.
Different institutions use logframe matrixes that include either four or five rows and normally four columns. Depending on the level of the logframe within the CGIAR system (system, centre, research project) the elements (cells) contained in the matrix vary slightly. In this chapter, the generic model of a logframe for a research project of a CGIAR centre is used to describe the general structure and concepts. Differences and the link between the various logframes are dealt with in chapter 4.
A logframe matrix basically covers three dimensions of the project design in a coherent and transparent form:
· A hierarchy of objectives which represents the underlying development hypothesis of the research project/programme (often called the "vertical logic");· A summary of external conditions crucial for the successful implementation of the project and for the achievement of its intended impact;
· Targets for each of the objectives which are precise and practical enough to allow monitoring of achievements (sometimes called the "horizontal logic").
From the perspective of research management, a logframe matrix:
· Consists of a part which is under the direct managerial control - activities and outputs plus the corresponding indicators and milestones;· Shows the intended impact of the research project as specified through the purpose, intermediate goal, and goal plus the respective indicators;
· Lists conditions (assumptions) which are neither part of the project strategy nor under the direct influence of project management, but of importance to the implementation. They are specified by threshold values for monitoring the development of these factors (indicators for assumptions).
Format of Logframe Matrix (Project level)
|
Hierarchy of Activities/Objectives |
Indicators for Achievements |
Assumptions |
Indicators for Assumptions |
|
Overall Goal |
Indications |
Assumption for sustainability of benefits |
Indicators |
|
Intermediate Goal |
Indicators |
Assumptions for achievement of overall goal |
Indicators |
|
Purpose |
Indicators |
Assumptions for achievement of intermediate goal |
Indicators |
|
Outputs |
Indicators |
Assumptions for achievement of purpose |
Indicators |
|
Activities |
Milestones |
Preconditions for implementation of activities |
|
The logframe process is not prescriptive with regard to a particular starting point. Planning may start by listing potential activities for an only roughly defined objective and then deduce a feasible objective. It may also start with defining desirable objectives and then elaborate the corresponding set of activities. In either case, a tight logic and realistic standards between the various levels are to be ensured. This can only be achieved through a number of iterations and adaptations.
In the left hand column of the logframe the intervention strategy of the project is summarised. This includes those levels which are under the direct control of management as well as those objectives which describe the intended reactions to and benefits resulting from the project.
Activities
In the lowest row the main tasks are listed which will be executed during the phase(s) of project implementation using project resources to produce a particular output.
The methodology does not specify how detailed activities should be spelled out in the logframe. One rule of thumb is to list major groups of activities which are of a different nature - e.g. conducting experiments, setting up an infrastructure, executing case studies, disseminating knowledge, etc.
Usually, activities will be implemented jointly between one or more CGIAR centres and their collaborative research partners. If the actors involved in the execution of activities vary it can be helpful to indicate for each activity who will be responsible and who participates.
|
Examples for activities: 1 · "Conduct an empirical study on seed distribution system in selected developing countries that went through a process of trade liberalisation with special emphasis on role of public sector in development of new varieties" · "Assess feasibility of microbial control agent of maize stem-borers." · "Disseminate an integrated disease management package for the control of bacterial wilt diseases in Ethiopia and Uganda" 1 The examples used throughout this manual are based on real statements of project descriptions in the MTPs of different Centres. However, for illustrative purposes some of them have been slightly modified, combined or rearranged. |
Outputs
In the second row the products (good and services) are listed which are delivered by the project to its direct clients and for which the project manager is directly responsible. These products can be additional physical goods (e.g., a new gene bank, improved varieties, etc.) or an increased body of knowledge (e.g., models for policy responses, improved management techniques, increased competence of national researchers, etc.). Many projects will include both types of outputs. Wherever possible the recipient of the respective output (s) should be specified within the output formulation itself.
|
Examples for outputs: · "Management options for NARS in agroindustry and environment are available" · "New cowpea and soybeans lines with improved resistance to key insect pests and diseases are available for further testing" · "Decision support systems available that allow evaluation of potential biophysical, economic, social and equity effects of prototype agroforestry practices, compared with existing practices" |
Purpose(s)
The purpose of the project describes the intended utilisation of the project outputs by the recipients resulting in some new and innovative behaviour of these direct target group(s). The purpose refers to the first level of project impact and is by definition outside the direct control of project management.
The respective group of actors (e.g., researchers, politicians, extension personnel) and the type of intended response (e.g., translation of research results into extension messages, formulation of new policies) have to be spelled out. In case a project aims at different groups of actors or tries to initiate different reactions, more then one purpose formulations will be required in order to comprehensively describe the utilisation of project outputs.
|
Examples for purposes: · "NARS have adopted a recognised approach for action research which facilitates self-study and comparative analysis" · "Plant breeders and genetic conservation specialists use information and tools from biotechnology for the characterisation and utilisation of genetic resources at molecular level" · "Policymakers and researchers in Indonesia, Viet Nam and Chile elaborate appropriate water allocation policies" |
Intermediate Goal(s)
The second level of impact covers the expected direct benefits resulting from the uptake of innovations. Also, they are the justification for a particular project.
For projects within the CGIAR frame the intermediate goals of a project will have to relate either to the increase in productivity of agricultural, fisheries, and forestry resources and/or to improved management of natural resources as these form the intermediate goals for the CGIAR system as a whole. Although a single project may make only a small and often indirect contribution to such objectives, a plausible relation between project outputs, purpose, and intermediate goal must be established. Generally, a modest but specific formulation is preferable to a simple repetition of system intermediate goals.
|
Examples for intermediate goals: · "Use of disease-resistant livestock with improved production potential will contribute to increased productivity where trypanosomosis and helminthosis constrain production" · "Sustainable supplies of water especially for people dependent upon its use for livelihood from aquatic resources" · "Increased efficiency in the use of chemical fertilisers, labour and water, reduced need for pesticide application" |
Goal
Here, the ultimate benefits for the target population are defined which specify the overarching orientation of the CGIAR system as a whole and, therefore, also define the context of particular projects. The hypothesis that an increase in productivity contributes to increases in income, the reduction of poverty, and improvement in food security is well established and forms the basic paradigm for the CGIAR. However, it is beyond the scope of a particular research project to establish this relation in greater detail. Given that, projects must only establish linkages to the level of System's intermediate goals.
The logframe approach acknowledges the fact that projects are not implemented in isolation but that they depend to a large extent on developments within the scientific, socio-economic, political and natural environment. Through the elaboration of assumptions in the logframe matrix, these crucial factors are included into the planning process. Assumptions are specific to a particular logframe but are always outside the respective management responsibility of a particular project or centre. In case of a research project logframe they may thus refer to complementary outputs of other research projects. At the centre level they may specify expected contributions by partner organisations or other centres.
Unfortunately, the general acknowledgement of the importance of external conditions (or their development into a specific direction) for project planning and monitoring sharply contrasts with the quality of formulated assumptions in many planning documents. Assumptions should not provide a general alibi for unrealistic or incomplete project planning. Instead, they have to be carefully specified and worded to allow continuous monitoring and to distinguish the development of external conditions from the realisation of the project strategy. In order to reduce the unlimited number of potentially relevant conditions to a manageable number of assumptions to be included in the logframe matrix, a sequence of guiding questions has proven to be helpful.
First question: Is the particular condition external to the project, i.e., not part of the project strategy and not included in the first column?
Assumptions can refer only to complementary developments (influenced by other actors) or natural conditions which can not be forecasted with certainty. They must not include elements which are beyond direct control but which the project nevertheless intends to influence - e.g., reaction of recipients to products provided by the project.
|
Example: "Low input management practices are not counteracted through distribution of subsidised fertiliser" Rather than: "Farmers are interested in ecological cultivation techniques" |
Second question: Is the respective external condition really crucial for the success of the project, i.e., can outputs not be produced without it, are recipients in a position to adopt the innovations, or will their changed behaviour not lead to positive impact unless the assumption holds true?
It is essential to formulate assumption as specific as possible and refrain from general statements which are true under all circumstances.
|
Example: "Legal framework which gives legal status to water user groups is introduced" Rather than: "Conducive political environment" |
Third question: How likely is it that the respective assumption holds true?
Obviously there is no need to include assumptions when it is certain that they will not pose a risk to the success of the project. On the other hand it would be dubious to include assumptions which are unlikely to become reality and which imply an unrealistic project approach. Precisely by including a particular assumption into the logframe matrix, planners document that, at this moment, chances for its realisation success are sufficiently high that it will not preclude project success.
From the management point of view indicators are the core of the logframe approach. Sometimes the term "Objectively Verifiable Indicators (OVI)" is used to stress that they are not merely subjective judgements but should be formulated in such a way that different observers will come to the same result when measuring the achievements. Indicators are performance standards which translate the narrative summary of the objectives for a particular research project into empirically observable characteristics. For each of these objectives (outputs, purposes, and intermediate goals) indicators provide information on the minimum targets in terms of target groups or target organisations, quality, quantity, area and time. A logical framework can be considered consistent and realistic if planners can reasonably expect to achieve the objectives at different levels as specified through the corresponding indicators.
This requires sound cause-effect relationships between indicators and the respective output, purpose, and goal. In particular, reasonable expectations about the up-take and utilisation rates of outputs by clients as well as the time lag involved in this process are required when formulating the corresponding indicators and specifying targets at output, purpose and goal level.
Assumptions should be specified through indicators in the same way in order to define threshold values which specify crucial levels beyond which planners see project success increasingly endangered. This requires careful monitoring of the development of external conditions and adequate reaction.
In order to fulfil the expected functions and to be useful for appraisal of project planning as well as management of project implementation and monitoring of achievements, indicators must have a number of characteristics.
Indicators must be substantial and relevant
An indicator must capture the main aspect of a particular objective (output, purpose, or goal). This means that it specifies by itself or in combination with other indicators those aspects of the conditions to be achieved by the project which are essential from the point of view of the direct clients and target groups of the project and of the project concept as a whole. This also implies that for complex objectives more then one indicator will be required in order to cover different dimensions. At the same time, the number of indicators formulated should always be limited to the minimum number required to describe the corresponding objective in order to avoid unnecessary work of data collection and processing.
Indicators must be independent
An indicator should be formulated as a precise description of an objective to be measured. This implies that an indicator is logically at the same level as the corresponding objective and neither the cause nor the effect of achieving this particular objective. For instance, the achievement of an output can not be indicated by simply listing the last activity required to produce this output. In such a case there would be no cause-effect relation between activity and output level of the logframe. The indicator for an output must also not refer to its expected utilisation as described at the purpose level.
Indicators must be precise and realistic
Indicators have to describe the corresponding output, purpose, or goal as precisely as possible and meaningful in terms of:
(1) Target group Who receives (output) or utilises (purpose) the goods and services, who gains what benefits (intermediate goal)?
In the case of centre research projects the direct clients will normally include other researchers and research institutions, as well as policy makers, extension services, and the development community. At the intermediate goal level, target groups will include producers and consumers.
(2) Quality What are the criteria which allow a particular case to be judged successful or not?
The quality of an indicator may be expressed in descriptive terms or as a quantitative value; For CGIAR research projects a quality of an indicator at output level with a descriptive quality could read "... new production technologies, which are tested under farm conditions..." or "... NARS which have a comprehensive and consistent medium term research strategy elaborated and updated annually..." An example for a quality expressed in numeric terms might be "...new variety, which show an increased yield potential of at least 10%..."
(3) Quantity How much/many does the planning (realistically) target for?
Wherever possible this should include a quantitative target in order to be able to decide whether a change actually observed during project implementation is within the planned target or falls short of it.
(4) Time By when should a particular objective be achieved?
For management purposes it is often advisable to define target values for different points in time in order to allow periodic monitoring of achievements and to decide on necessary corrective actions.
(5) Location In which area or region (geographical, agro-climatical, etc.) should the objectives be realised?
Indicators must be verifiable
An indicator can only be used if the data required to verify the indicator can be collected at reasonable cost and effort. If the verification is impossible or too costly, the indicator has to be replaced by one which is easier and/or cheaper to verify.
In the traditional format of a logframe matrix "Means of Verification" are listed in the third column. However, experiences show that they often include very general sources of data (such as "Internal project reports", "statistics"). A detailed description of primary data collection (including type of data, sampling method, intervals, etc.) does not fit into the logframe matrix and requires a separate document. Therefore, the column "Means of verification" has been omitted, the column "Assumptions" moved to the left and an additional column "Indicators for assumptions" was introduced. If however planners feel that "Means of Verification" should be maintained as a separate column, the logframe matrix can easily be expanded to include six instead of four columns.
If assumptions are limited to the most crucial ones and are formulated carefully, indicators should be specified for them as well. Only with a specification through an indicator will it be possible to systematically monitor assumptions. The qualitative and quantitative dimension of the indicator in this case does not represent a target to be achieved but signifies a critical level after which the success of a project is endangered.
Within the CGIAR logframe the indictor(s) at the highest goal level have a special character and are therefore termed "indications".
|
Example for an output indicator: "At least five alternative management options developed and made available to local institutions for maize, each providing more than 20% increase in marginal rate of return over farmers' practices, all resource-conserving, and some showing significant degrees of adoption by the year 2000" Example for a purpose indicator: "Some 20 maize cultivars identified and released by NARSs by the year 2000 in the region which have high and stable yield, as well as excellent resistance to streak, grey leaf spot and increased and more stable production under drought and low nitrogen, yielding 5-10% more than the best local checks in target environments" Example for an intermediate goal indicator "Increased productivity through two or more improved fallow systems validated in benchmark areas in at least four EPHTA countries by 2006" |
Milestones are already one element of the project description within the Medium Term Plans They define key intermediate targets which have to be fulfilled in order to produce a particular output. Therefore, milestones have the same function for activities as indicators have for the different objectives of a logframe matrix.
In practice, milestones are often easier to formulate and more straightforward than indicators because activities are normally less complex and more concrete then outputs, purposes, and goals. Nevertheless, all formal criteria for indicators apply for milestones as well.
|
Example for a milestone: "Two options for improved on-farm storage that have attractive marginal rates of return and acceptable levels of environmental impact identified by 2000" |
