- Email: [email protected]
Matrix approach to project planning design and management
Matrix approach to project planning design and management Colin F Palmer design and evaluation, evaluation showing the result of comparing plans with actual accomplishment. Too many projects are based on the design shown in Figure 1, which is limited in its effectiveness. It is insufficient to treat the design as a set of inputs, activities and outputs, with some policy statement and one objective. The first stage of any design is to identify the problem to be solved and its causes before considering what are the most effective means of solving it. At the appraisal and preparation stage, there should be a consistency between the donor’s policy objectives and the recipient’s needs. The design should commence with a project document, which provides a solid foundation for subsequent implementation. At the beginning of the project document, there should be a statement presenting the problem or situation, briefly at this stage, for which a solution is sought. It should not attempt to suggest means or methods of solving the problem and whether the proposed project is justified; these will be examined later in the design. A series of objectives needs to be set so that answers to the following questions can be given.
The failure of many projects is due, in many cases, to poor design. The design is the formal framework that establishes, in advance, what the intention of the project is; a logical plan of activities, with the means to measure progress, taking into consideration the environmental factors that could affect the assumptions on which the project is based. The matrix method is primarily a project planning device, which de~nes project inputs, outputs and the intermediate and higher objectives in measurable or objectively verifiable terms. It hypothesizes the conservative (means-ends) link between inputs, outputs and objectives, and articulates the environmental factors that could affect the causative links. Targets are established which permit subsequent measurement or verification of achievement of the defined outputs and objective. The paper proposes a methodology to the design of a successful project, and an example of the matrix method applied to a development project. Keyword: management project management
techniques,
matr~
method,
l
The failure of many projects is due in most cases to poor design. The design is the formal framework that establishes the intention of the project. It is a logical plan of activities, with the means to measure progress, taking into consideration the environmental factors that could affect the assumptions on which the project is based. An integral part of project planning is both the
Devel~~pm~nt Administration Group, Institute of Local Government Studies, University of Birmingham, PO Box 363, Birmingham 815 2T-T. UK
162
0263-7863/87/030162-05
l
Why is the project being designed? What are the reasons for the project?
Policy 1 Objective
Manipulation of objectives
*Output1
Figure I. Typical project design
$03.00 @ 1987 Butterworth
& Co (Publishers)
Ltd
Project
Management
e What is the project expected to solve if completed successfully and to time? l Who will benefit from the project? There should be several levels of objectives depending on the type of project under consideration and, in particular, a definition of the target group. The target group definition should state who will be the beneficiaries if the project is completed successfully on time. In many instances, a governmental agency or organization will receive direct assistance, and then, through them, the target group will benefit. Descriptions such as ‘the urban poor’ or ‘unemployed craftsmen’, while perhaps appropriate for statements of policy, are clearly inadequate for individual project documents. A precise description is therefore necessary to help ensure that the benefits of a given project do flow to the target group. OBJECTIVES An objective is a simple expression of a desired end. In formulating an objective, the designer should not confuse ends with means. The statements ‘to assist’ and ‘to strengthen’ denote means rather than ends and should not be used when stating objectives. At the objective level, one is not so much interested in the approaches being used, as these are described under ‘outputs’ and ‘activities’, as in the effect the project is expected to have. ‘To assist’, etc., are means to those ends, not ends in themselves. As a general rule, the project should be formulated in terms of a single national, regional or local objective and a single intermediate objective.
National,
regional or local objectives
Depending on which one of the three levels is appropriate, the national/regional/local objective should give the reason for the project, the broader objective towards which the efforts of the project are directed. It should answer the question ‘Why is the project being undertaken?’ Once the objective level has been chosen and formulated. it should be verified that: it gives the reason for the project, which is the level beyond the intermediate objective, o it states a single objective or a group of compatible objectives; it does not contain two or more objectives that are on different levels, l it is so presented that progress towards it can be assessed.
l
name the group to whom the project is directed effect sought on completion of the project. Once the intermediate objective has been ated, it should be verified that:
and the formul-
the statement specifies what effect will have been achieved if the project is successfully completed on time, e a single intermediate objective preferably is stated, l the intermediate objective is stated in terms of ends, not of means. l
For example, ‘by 1991, the Sudan Agricultural College will be capable of graduating 150 qualified agricultural specialists annually without external assistance’, and ‘an irrigation network scheme in the Sudanese Gezeira agricultural project is intended to produce increased per feddan yield’. There can be further objectives at the local or regional level, which describe objectives beyond the intermediate stage. Progress towards these levels will depend often on a number of related projects, each meeting its own intermediate objective. INPUTS,
OUTPUTS
AND
ACTIVITIES
The inputs are the means, including money, equipment, materials, technical advice, training, etc., which are later mobilized to produce the planned outputs. Activities are where time and resources are used to transform the inputs into outputs. Here, specific questions can be answered, such as: o what will the machine do? e what will the staff actually do? The outputs are the specifically intended results, as opposed to the magnitude of results that can be expected from good management of the inputs provided. For example, manpower, machinery and training (inputs) allow th e t raining of staff, ordering of equipment and materials (activities), for trained operational staff with material and equipment in a specific place (outputs) when managed efficiently to produce an irrigation network. At this stage of the design, the system shown in Figure 2 will exist. In this manner, there is produced a logical flow through the project and a ‘means-ends’ or ‘cause-effect’ relationship connecting the various levels. Clarification of this relationship or linkage is the basis of the design concept. PROJECT
LINKAGES
For example, ‘the modernization of the Sudan agricultural sector’, and ‘increased yield per feddan is intended to result in expanded exports of agricultural crops. ’
There is a causative link between project outputs, such as (the irrigation network)l(trained operational staff) and the intermediate project objective, such as (im-
Intermediate
I-
objective
The intermediate objective gives the primary reason for the project, i.e. the development that is expected to be achieved, or the problem to be solved if the project is comnleted successfullv to time. The statement of the intermediate objective should
Vol 5 No 3 August
1987
Figure 2. Design sequence
163
proved crop yieids)/(qualified agricultural specialists). The outputs must be made available before the objective can be achieved. The presence of the outputs does not, however, guarantee that the objective will be reached. Environmental factors, such as farmer attitudes, lack of credit or political interference, may prevent the objective being achieved. Thus, the causative relationship between the project outputs and the intermediate objective must be carefully stated and used as a hypothesis, with the environmental factors being identified. The evaluation will then conclude whether or not the hypothesis was realized. There is also a causative link between the intermediate objective, (improved crop yields}/(qualified agricultura1 specialists), and progress towards the national objective, (expanded export of agricultural crop)/(modernization of Sudan’s agricultural sector). Achievement of the intermediate objective does not guarantee that the national objective will be reached. Environmental factors could again affect planned progress, such as (price and marketing conditions)/ (emigration of trained staff), towards the highest level objectives. Thus, the causative relationship between the intermediate objective and the national objective must be clearly stated as a hypothesis, with the environmental factors identified and the hypothesis validated (see Figures 3 and 4). These linkages should also be proportional as well as causative. Thus, the improved crop yields will mean very little for export if only carried out in a small area. Also, if the number of trained agricultural specialists necessary to make any significant impact on the Sudan agricultural sector is estimated to be 2 000, then the figure of 150 trained by the proposed project must be considered a modest figure towards the highest level objective. By analysing a project in terms of meansends linkages, it may highlight a lack of realism in the original plan. It is important to determine if means are sufficient in quality and quantity to produce the desired end product. The number of levels between inputs and the ultimate objective may be greater than the five shown in Figure 5. What is important is the train of thought linking one level to the next. Thus, the project planner may find if necessary to expand the number of levels to
Environmental
Figure 5. The levels between input and zdtimnte objective
adequately reflect the realities of the particular project. Further objective levels, such as local and regional, may be appropriate, as well as the separating of outputs or activities into two or more distinct phases.
TARGET
SETTING
There are a number of problems associated with target setting and quantification. Many projects set targets at the output, intermediate and higher objective levels only, but if the project is broken down into distinct phases, then targets at the input stage through to ultimate objective can be extremely effective. An obvious problem is that, in many cases, it will only be possible to quantify part of an objective, and because only this part of the project is able to be checked, the remainder may suffer from lack of attention. An obsession with the quantitative target may result in the neglect of the qualitative aspect of the project, and there is a need to stress that both are necessary. In the example of training agricultural specialists, the target of 150 may become more important than the quality of those graduating. Yet a qualitative target could be incorporated in the project, such as the number of contact hours for specific subjects accepted internationally as the norm for this discipline. What is important is that measures or indicators will demonstrate at each level that the set targets are being achieved and that the project is effective and having the desired impact on the target group.
National objective
MONITORING
Figure 3. Example
of relationship
between
objectives
Figure 4. Example
of rel~t~o~~hi~ between
objectively
164
AND
EVALUATION
An important benefit of setting targets at the various levels is that it focuses the attention of the project analyst on the data requirements for monitoring and evaluation. It may become obvious that a baseline survey will be necessary, particularly when peoplecentred projects are being designed. A knowledge of statistical sampling becomes a great advantage at this stage: knowing what size of sample will incur a particular sampling error can avoid a very time consuming and expensive survey. The advice of a statistician should be sought if knowledge of surveys and sampling is limited. The important point is that a well designed project should be able to highlight the need for a data collecting system. A we11 designed project should incorporate those factors, objectives and systems as indicated in Figure 6.
Project
Management
Environmental
I
factors
l
National objectives l
Regional objectives Local objectives Intermediate objectives Monitor Measure/evaluate
Monitor Measure/evaluate
T-y&T-=
Monitor/measure
Figure
6. Proposed
project
CPM/PERT?
design
MATRIX METHODOLOGY A key element in project planning and evaluation working part of a logical matrix, which l
l
is the
defines project inputs, outputs, intermediate, and higher objectives in measurable or objectively verifiable terms, hypothesizes the causative (means-end) linkage between inputs, outputs, intermediate, and higher objectives,
Summary
The matrix is primarily a project planning device. It is also used for re-examination of the original design of ongoing projects, as a necessary prelude to evaluation, i.e. it sets the standards against which the project will be evaluated. Evaluation then consists of determining and validating whether or not the project outputs are being produced; whether these outputs are serving to achieve the project purpose; and whether this achievement is making a significant contribution, as planned, to the higher objectives. As shown in Figure 7, the matrix has both vertical and horizontal logic. Its vertical logic defines the series of causative linkages, which is intended to transform project inputs into development changes at the various levels, and permits the planner/evaluator to judge whether these linkages are viable. Its horizontal logic encourages the evaluator to measure progress and verify planning assumptions at each level (input, activities, output, objectives and targets) separately and independently from other levels.
Targets
National objective: skilled craftsmen
increase
national
pool of
Intermediate objectives: training centres and supporting services capable of graduating skilled craftsmen
Environmental
1 Government building programmes will employ 90% at least of trained craftsmen 2 80% of centres opened after year 1 still working after year 4 1 20 centres established by 1990 2 85% of graduatess employed in own 3 Each centre capable of training 100 people in 4 crafts a year: welding, bricklaying, plumbing and electrician
outputs
1 2 3 4 5 6
Building with equipment installed 20 trained leaders (1 per centre) Trainee selection methods established Instructors for centres trained in 4 crafts 500 skilled craftsmen graduate in 4 crafts Teaching material completed for 4 centres
Activities
1 2 3 4 5
Prepare selection methods for trainees Select buildings, install equipment and supplies Train instructors for centres Training programme for 1st course (4 crafts) Develop teaching material and curricula for 4 crafts
Inputs
articulates environmental factors, which will affect the causative linkages, establishes the targets, which will permit subsequent measurement or verification of achievement of the defined outputs, and objectives.
I months 6 9 9-12 12-18 12
l-3 l-3 9-12 9-18 entire period
Donor 1 6 craft experts
(welding, electrician, bricklaying, plumbing, organizer, deputy) 2 Equipment (workshop, vehicles, expendable materials) 3 Finance (partial)
Sudan 4 Centres,
counterparts, 5 20 crafts, instructors, 6 Islamic Bank finance
Figure Sudan)
Target,
7. Matrix
Vol 5 No 3 August
outline
1987
for
admin. support staff accountants, drivers, secretaries (partial)
‘Craftsman development programme,
6 ;
factors
Intermediate to natural target: 1 Government building programme will materialize 2 National economy will gradually improve 3 Continuation of national policy favouring labour-intensive work programme 4 Craftsman do not emigrate Output to intermediate target: 1 Sufficient trainees available 2 Trained craftsman willing to be mobile 3 Trainees able to complete programme of training 4 Finance policy formulated and implemented 5 Qualified centre staff turnover doe5 not exceed 1 O-l 5% a year Activities to output: 1 Sufficient number of suitable trainees available 2 Trainee drop out rate does not exceed 20% 3 Remuneration sufficient to retain staff 4 Suitable instructors within Sudan available Inputs to activities: 1 Finance available when required 2 Physical facilities (electricity, tools, transport, etc.) available for installation of equipment 3 Staff and equipment given priority through immigration and customs
; 6
eastern region
(3 year project,
1988-1990,
16.5
CHARACTERISTICS AND LIMITATIONS THE MATRIX METHOD
OF
All aspects of project planning (i.e. the formulation of targets, causative linkages, targets, and environmental factors) are defined by the project planner and the sector manager and are project-specific. Similarly, the degree of rigour and the level of effort in collecting and analysing data for both planning and evaluation are determined by management and are project-specific. The matrix method is ethically neutral. It gives no guidance on socio-economic equity or benefit incidence, such as equitable income distribution, employment opportunities, access to resources, popular participation in decision-making and in the fruits of development projects, unless such aspects have been explicitly included in the statements of goal or purpose. The matrix method is programmatically and technically neutral. It gives no guidance on proven strategies and techniques, cost and feasibility of replication, effects on ecology, concentration on key problem areas, reliance on the private sector, etc. It does not assure that the project is optimal, i.e. that the project directly addresses the most critical constraint to highest level objectives, or that it is the most effective means for overcoming that critical constraint unless the planners/evaluators choose to explore alternative approaches. The methodology permits, but does not require, cost/ benefit and cost/effectiveness analysis. It introduces order and discipline into the intellectual processes of the planner. The matrix method is merely a convenient planner’s tool, which simulates and visually displays the project designs so that they can be manipulated, assayed and communicated. The matrix method is objective-oriented, it does not describe the actions, activities or processes which transform means into ends.
BIBLIOGRAPHY Anon AID evaluation handbook USAID, USA (1980) Anon Design and evaluation for UNIDO executed projects and programmes
166
UNIDO,
USA (1984)
Anon General guidelines for preparing project completion reports IBRD,
UK (1983)
Anon Guidelines for data collection, monitoring and evaluation plans Asia Bureau, USAID, USA (1984) for the appraisal of projects for possible DANIDA assistance Danish International Development Agency (April 1983) Anon Guidelines on logical framework planning and project benefit monitoring and evaluation Agriculture and Rural Development Department, Asian Development Bank, Indonesia (1981) Anon Principles for the evaluation of programmes by federal departments and agencies and Guide on the programme evaluation function Treasury Board of Canada (1981) Casley, D J and Lurry, D A Monitoring and evaluation of agricultural development projects IBRD, UK (1982) Chambers, R ‘Rapid rural uppraisal’ IDS Discussion Paper No 155, UK (September 1980) Forss, K ‘Planning and evaluation in aid organizations’ (unpublished PhD thesis) Stockholm School of Economics, Sweden (1985)
Anon Guidelines
Colin Palmer is a lecturer in quantitative methodsloperational research in the development administration group at the University of Birmingham, UK. After many years in industry. he was appointed Visiting Professor to the Federal University, Rio de Janeiro, Brazil, in operational research with the British Council. He was seconded to the University of Mauritius for two years from Birmingham University and has undertaken many consultancy assignments, mainly in Africa and Latin America,
Project
Management
The failure of many projects is due, in many cases, to poor design. The design is the formal framework that establishes, in advance, what the intention of the project is; a logical plan of activities, with the means to measure progress, taking into consideration the environmental factors that could affect the assumptions on which the project is based. The matrix method is primarily a project planning device, which de~nes project inputs, outputs and the intermediate and higher objectives in measurable or objectively verifiable terms. It hypothesizes the conservative (means-ends) link between inputs, outputs and objectives, and articulates the environmental factors that could affect the causative links. Targets are established which permit subsequent measurement or verification of achievement of the defined outputs and objective. The paper proposes a methodology to the design of a successful project, and an example of the matrix method applied to a development project. Keyword: management project management
techniques,
matr~
method,
l
The failure of many projects is due in most cases to poor design. The design is the formal framework that establishes the intention of the project. It is a logical plan of activities, with the means to measure progress, taking into consideration the environmental factors that could affect the assumptions on which the project is based. An integral part of project planning is both the
Devel~~pm~nt Administration Group, Institute of Local Government Studies, University of Birmingham, PO Box 363, Birmingham 815 2T-T. UK
162
0263-7863/87/030162-05
l
Why is the project being designed? What are the reasons for the project?
Policy 1 Objective
Manipulation of objectives
*Output1
Figure I. Typical project design
$03.00 @ 1987 Butterworth
& Co (Publishers)
Ltd
Project
Management
e What is the project expected to solve if completed successfully and to time? l Who will benefit from the project? There should be several levels of objectives depending on the type of project under consideration and, in particular, a definition of the target group. The target group definition should state who will be the beneficiaries if the project is completed successfully on time. In many instances, a governmental agency or organization will receive direct assistance, and then, through them, the target group will benefit. Descriptions such as ‘the urban poor’ or ‘unemployed craftsmen’, while perhaps appropriate for statements of policy, are clearly inadequate for individual project documents. A precise description is therefore necessary to help ensure that the benefits of a given project do flow to the target group. OBJECTIVES An objective is a simple expression of a desired end. In formulating an objective, the designer should not confuse ends with means. The statements ‘to assist’ and ‘to strengthen’ denote means rather than ends and should not be used when stating objectives. At the objective level, one is not so much interested in the approaches being used, as these are described under ‘outputs’ and ‘activities’, as in the effect the project is expected to have. ‘To assist’, etc., are means to those ends, not ends in themselves. As a general rule, the project should be formulated in terms of a single national, regional or local objective and a single intermediate objective.
National,
regional or local objectives
Depending on which one of the three levels is appropriate, the national/regional/local objective should give the reason for the project, the broader objective towards which the efforts of the project are directed. It should answer the question ‘Why is the project being undertaken?’ Once the objective level has been chosen and formulated. it should be verified that: it gives the reason for the project, which is the level beyond the intermediate objective, o it states a single objective or a group of compatible objectives; it does not contain two or more objectives that are on different levels, l it is so presented that progress towards it can be assessed.
l
name the group to whom the project is directed effect sought on completion of the project. Once the intermediate objective has been ated, it should be verified that:
and the formul-
the statement specifies what effect will have been achieved if the project is successfully completed on time, e a single intermediate objective preferably is stated, l the intermediate objective is stated in terms of ends, not of means. l
For example, ‘by 1991, the Sudan Agricultural College will be capable of graduating 150 qualified agricultural specialists annually without external assistance’, and ‘an irrigation network scheme in the Sudanese Gezeira agricultural project is intended to produce increased per feddan yield’. There can be further objectives at the local or regional level, which describe objectives beyond the intermediate stage. Progress towards these levels will depend often on a number of related projects, each meeting its own intermediate objective. INPUTS,
OUTPUTS
AND
ACTIVITIES
The inputs are the means, including money, equipment, materials, technical advice, training, etc., which are later mobilized to produce the planned outputs. Activities are where time and resources are used to transform the inputs into outputs. Here, specific questions can be answered, such as: o what will the machine do? e what will the staff actually do? The outputs are the specifically intended results, as opposed to the magnitude of results that can be expected from good management of the inputs provided. For example, manpower, machinery and training (inputs) allow th e t raining of staff, ordering of equipment and materials (activities), for trained operational staff with material and equipment in a specific place (outputs) when managed efficiently to produce an irrigation network. At this stage of the design, the system shown in Figure 2 will exist. In this manner, there is produced a logical flow through the project and a ‘means-ends’ or ‘cause-effect’ relationship connecting the various levels. Clarification of this relationship or linkage is the basis of the design concept. PROJECT
LINKAGES
For example, ‘the modernization of the Sudan agricultural sector’, and ‘increased yield per feddan is intended to result in expanded exports of agricultural crops. ’
There is a causative link between project outputs, such as (the irrigation network)l(trained operational staff) and the intermediate project objective, such as (im-
Intermediate
I-
objective
The intermediate objective gives the primary reason for the project, i.e. the development that is expected to be achieved, or the problem to be solved if the project is comnleted successfullv to time. The statement of the intermediate objective should
Vol 5 No 3 August
1987
Figure 2. Design sequence
163
proved crop yieids)/(qualified agricultural specialists). The outputs must be made available before the objective can be achieved. The presence of the outputs does not, however, guarantee that the objective will be reached. Environmental factors, such as farmer attitudes, lack of credit or political interference, may prevent the objective being achieved. Thus, the causative relationship between the project outputs and the intermediate objective must be carefully stated and used as a hypothesis, with the environmental factors being identified. The evaluation will then conclude whether or not the hypothesis was realized. There is also a causative link between the intermediate objective, (improved crop yields}/(qualified agricultura1 specialists), and progress towards the national objective, (expanded export of agricultural crop)/(modernization of Sudan’s agricultural sector). Achievement of the intermediate objective does not guarantee that the national objective will be reached. Environmental factors could again affect planned progress, such as (price and marketing conditions)/ (emigration of trained staff), towards the highest level objectives. Thus, the causative relationship between the intermediate objective and the national objective must be clearly stated as a hypothesis, with the environmental factors identified and the hypothesis validated (see Figures 3 and 4). These linkages should also be proportional as well as causative. Thus, the improved crop yields will mean very little for export if only carried out in a small area. Also, if the number of trained agricultural specialists necessary to make any significant impact on the Sudan agricultural sector is estimated to be 2 000, then the figure of 150 trained by the proposed project must be considered a modest figure towards the highest level objective. By analysing a project in terms of meansends linkages, it may highlight a lack of realism in the original plan. It is important to determine if means are sufficient in quality and quantity to produce the desired end product. The number of levels between inputs and the ultimate objective may be greater than the five shown in Figure 5. What is important is the train of thought linking one level to the next. Thus, the project planner may find if necessary to expand the number of levels to
Environmental
Figure 5. The levels between input and zdtimnte objective
adequately reflect the realities of the particular project. Further objective levels, such as local and regional, may be appropriate, as well as the separating of outputs or activities into two or more distinct phases.
TARGET
SETTING
There are a number of problems associated with target setting and quantification. Many projects set targets at the output, intermediate and higher objective levels only, but if the project is broken down into distinct phases, then targets at the input stage through to ultimate objective can be extremely effective. An obvious problem is that, in many cases, it will only be possible to quantify part of an objective, and because only this part of the project is able to be checked, the remainder may suffer from lack of attention. An obsession with the quantitative target may result in the neglect of the qualitative aspect of the project, and there is a need to stress that both are necessary. In the example of training agricultural specialists, the target of 150 may become more important than the quality of those graduating. Yet a qualitative target could be incorporated in the project, such as the number of contact hours for specific subjects accepted internationally as the norm for this discipline. What is important is that measures or indicators will demonstrate at each level that the set targets are being achieved and that the project is effective and having the desired impact on the target group.
National objective
MONITORING
Figure 3. Example
of relationship
between
objectives
Figure 4. Example
of rel~t~o~~hi~ between
objectively
164
AND
EVALUATION
An important benefit of setting targets at the various levels is that it focuses the attention of the project analyst on the data requirements for monitoring and evaluation. It may become obvious that a baseline survey will be necessary, particularly when peoplecentred projects are being designed. A knowledge of statistical sampling becomes a great advantage at this stage: knowing what size of sample will incur a particular sampling error can avoid a very time consuming and expensive survey. The advice of a statistician should be sought if knowledge of surveys and sampling is limited. The important point is that a well designed project should be able to highlight the need for a data collecting system. A we11 designed project should incorporate those factors, objectives and systems as indicated in Figure 6.
Project
Management
Environmental
I
factors
l
National objectives l
Regional objectives Local objectives Intermediate objectives Monitor Measure/evaluate
Monitor Measure/evaluate
T-y&T-=
Monitor/measure
Figure
6. Proposed
project
CPM/PERT?
design
MATRIX METHODOLOGY A key element in project planning and evaluation working part of a logical matrix, which l
l
is the
defines project inputs, outputs, intermediate, and higher objectives in measurable or objectively verifiable terms, hypothesizes the causative (means-end) linkage between inputs, outputs, intermediate, and higher objectives,
Summary
The matrix is primarily a project planning device. It is also used for re-examination of the original design of ongoing projects, as a necessary prelude to evaluation, i.e. it sets the standards against which the project will be evaluated. Evaluation then consists of determining and validating whether or not the project outputs are being produced; whether these outputs are serving to achieve the project purpose; and whether this achievement is making a significant contribution, as planned, to the higher objectives. As shown in Figure 7, the matrix has both vertical and horizontal logic. Its vertical logic defines the series of causative linkages, which is intended to transform project inputs into development changes at the various levels, and permits the planner/evaluator to judge whether these linkages are viable. Its horizontal logic encourages the evaluator to measure progress and verify planning assumptions at each level (input, activities, output, objectives and targets) separately and independently from other levels.
Targets
National objective: skilled craftsmen
increase
national
pool of
Intermediate objectives: training centres and supporting services capable of graduating skilled craftsmen
Environmental
1 Government building programmes will employ 90% at least of trained craftsmen 2 80% of centres opened after year 1 still working after year 4 1 20 centres established by 1990 2 85% of graduatess employed in own 3 Each centre capable of training 100 people in 4 crafts a year: welding, bricklaying, plumbing and electrician
outputs
1 2 3 4 5 6
Building with equipment installed 20 trained leaders (1 per centre) Trainee selection methods established Instructors for centres trained in 4 crafts 500 skilled craftsmen graduate in 4 crafts Teaching material completed for 4 centres
Activities
1 2 3 4 5
Prepare selection methods for trainees Select buildings, install equipment and supplies Train instructors for centres Training programme for 1st course (4 crafts) Develop teaching material and curricula for 4 crafts
Inputs
articulates environmental factors, which will affect the causative linkages, establishes the targets, which will permit subsequent measurement or verification of achievement of the defined outputs, and objectives.
I months 6 9 9-12 12-18 12
l-3 l-3 9-12 9-18 entire period
Donor 1 6 craft experts
(welding, electrician, bricklaying, plumbing, organizer, deputy) 2 Equipment (workshop, vehicles, expendable materials) 3 Finance (partial)
Sudan 4 Centres,
counterparts, 5 20 crafts, instructors, 6 Islamic Bank finance
Figure Sudan)
Target,
7. Matrix
Vol 5 No 3 August
outline
1987
for
admin. support staff accountants, drivers, secretaries (partial)
‘Craftsman development programme,
6 ;
factors
Intermediate to natural target: 1 Government building programme will materialize 2 National economy will gradually improve 3 Continuation of national policy favouring labour-intensive work programme 4 Craftsman do not emigrate Output to intermediate target: 1 Sufficient trainees available 2 Trained craftsman willing to be mobile 3 Trainees able to complete programme of training 4 Finance policy formulated and implemented 5 Qualified centre staff turnover doe5 not exceed 1 O-l 5% a year Activities to output: 1 Sufficient number of suitable trainees available 2 Trainee drop out rate does not exceed 20% 3 Remuneration sufficient to retain staff 4 Suitable instructors within Sudan available Inputs to activities: 1 Finance available when required 2 Physical facilities (electricity, tools, transport, etc.) available for installation of equipment 3 Staff and equipment given priority through immigration and customs
; 6
eastern region
(3 year project,
1988-1990,
16.5
CHARACTERISTICS AND LIMITATIONS THE MATRIX METHOD
OF
All aspects of project planning (i.e. the formulation of targets, causative linkages, targets, and environmental factors) are defined by the project planner and the sector manager and are project-specific. Similarly, the degree of rigour and the level of effort in collecting and analysing data for both planning and evaluation are determined by management and are project-specific. The matrix method is ethically neutral. It gives no guidance on socio-economic equity or benefit incidence, such as equitable income distribution, employment opportunities, access to resources, popular participation in decision-making and in the fruits of development projects, unless such aspects have been explicitly included in the statements of goal or purpose. The matrix method is programmatically and technically neutral. It gives no guidance on proven strategies and techniques, cost and feasibility of replication, effects on ecology, concentration on key problem areas, reliance on the private sector, etc. It does not assure that the project is optimal, i.e. that the project directly addresses the most critical constraint to highest level objectives, or that it is the most effective means for overcoming that critical constraint unless the planners/evaluators choose to explore alternative approaches. The methodology permits, but does not require, cost/ benefit and cost/effectiveness analysis. It introduces order and discipline into the intellectual processes of the planner. The matrix method is merely a convenient planner’s tool, which simulates and visually displays the project designs so that they can be manipulated, assayed and communicated. The matrix method is objective-oriented, it does not describe the actions, activities or processes which transform means into ends.
BIBLIOGRAPHY Anon AID evaluation handbook USAID, USA (1980) Anon Design and evaluation for UNIDO executed projects and programmes
166
UNIDO,
USA (1984)
Anon General guidelines for preparing project completion reports IBRD,
UK (1983)
Anon Guidelines for data collection, monitoring and evaluation plans Asia Bureau, USAID, USA (1984) for the appraisal of projects for possible DANIDA assistance Danish International Development Agency (April 1983) Anon Guidelines on logical framework planning and project benefit monitoring and evaluation Agriculture and Rural Development Department, Asian Development Bank, Indonesia (1981) Anon Principles for the evaluation of programmes by federal departments and agencies and Guide on the programme evaluation function Treasury Board of Canada (1981) Casley, D J and Lurry, D A Monitoring and evaluation of agricultural development projects IBRD, UK (1982) Chambers, R ‘Rapid rural uppraisal’ IDS Discussion Paper No 155, UK (September 1980) Forss, K ‘Planning and evaluation in aid organizations’ (unpublished PhD thesis) Stockholm School of Economics, Sweden (1985)
Anon Guidelines
Colin Palmer is a lecturer in quantitative methodsloperational research in the development administration group at the University of Birmingham, UK. After many years in industry. he was appointed Visiting Professor to the Federal University, Rio de Janeiro, Brazil, in operational research with the British Council. He was seconded to the University of Mauritius for two years from Birmingham University and has undertaken many consultancy assignments, mainly in Africa and Latin America,
Project
Management