Preparing an evaluation question
This section is structured as follows :
- Specify the nature of the expected use
- Ensure that the question concerns evaluation
- Specify the scope of the question
- Inferring a question from the Intervention logic
- Specifying the evaluation criterion
- Should question be open or closed?
Why is it important? |
The questions serve to concentrate the work on a limited number of points in order to ensure that the conclusions are useful and of a high quality. They therefore have to be carefully prepared and worded with precision. Ensure that the answer to the question will be useful As far as possible, the evaluation questions are proposed together with a comments on the following points:
If there is uncertainty on the usefulness of the question, it is better to exclude it and to concentrate the evaluation on other more useful questions. |
SPECIFY THE NATURE OF THE EXPECTED USE |
Question for knowing and/or understanding |
A question for knowing, understanding and/or estimating the effects of the intervention, for example:
Questions of this nature reveal new aspects of the intervention. They help to understand the effects and impact mechanisms, and raise the level of knowledge. |
Question for judging |
A question for formulating or helping to formulate a judgement on the evaluated intervention, for example:
Questions of this nature allow users to judge the merits of the intervention and to recognise good and unsatisfactory practices. They use the evaluation to communicate on the intervention, positively or negatively. |
Question for deciding |
A question for showing how the intervention can be improved, for example:
Questions of this nature lead to recommendations based on lessons from experience. The answers serve to prepare reforms or adjustments. |
Choosing one of three options |
The three types of question are not exclusive. On the contrary, there is a progression in the nature of the questions:
If all the questions of the same evaluation have no purpose other than furthering knowledge and understanding, the exercise is more a study or a piece of research than an evaluation. The nature of use has different levels: to decide, one has to have judged, and to judge one has to have understood. It is therefore enough to draft the question in relation to the highest level of use (question for decision-making or judging). Uses on a lower level can be considered as sub-questions. For example the question might be:
and a sub-question would be:
|
Recommendations |
|
ENSURE THAT THE QUESTION CONCERNS EVALUATION |
Before drafting a question, ensure that it does not concern audit or monitoring |
If a question concerns audit or monitoring, there are two options:
|
Avoid auditing and monitoring questions |
Evaluation, auditing and monitoring do not serve the same purpose. Consequently, the questions asked are not the same in each exercise. It is important to check whether a question is relevant to evaluation and, where necessary, to amend it so that it does not primarily concern auditing nor monitoring. Amending an auditing question
This question is limited to the verification of the legality and regularity of the implementation of a project, which is a matter of auditing. In an evaluation it would be relevant rather to ask whether the application of regulations was a particular factor of effectiveness or ineffectiveness.
Amending a monitoring question
This question concerns only the programme outputs, which is a matter of monitoring. In an evaluation it would be relevant to ask whether the quality of outputs is a particular factor of efficiency of inefficiency.
|
SPECIFY THE SCOPE OF THE QUESTION |
What is this about? |
Generally, an evaluation question concerns both the effects of the intervention and the intervention itself:
|
What is the purpose? |
|
Questions on the intervention design |
These questions may refer to the design of the intervention as a whole or to a particular step in the design process, such as:
Have the procedures of dialogue with the actors favoured ownership of the strategy and increased the chances of sustainable impacts? |
Questions on the implementation |
These questions concern the entire implementation of the intervention or a particular aspect of the implementation process such as:
To what extent do administrative funding and project management procedures facilitate or hinder the adaptation of aid to beneficiaries' needs? To what extent have phasing out procedures favoured the sustainability of impacts as regards food safety? |
Questions on the modalities of aid |
These questions concern modalities such as:
Has the funding modality opted for made it possible to obtain better effects in terms of food safety? To what extent have interventions in the field of transport been conducted in the form of sector-specific programmes, and what difference does this make from the point of view of effects generated? |
Questions on the intervention as a whole |
Finally, certain questions are drafted broadly and concern the intervention as a whole, including its design and implementation, for example: "To what extent has the design and implementation of the intervention helped to produce effect X?". If the question concerns the entire design and implementation, we choose to focus it on a precise, immediate effect. A question concerning the entire intervention and all its effects would probably be unevaluable. |
INFERING A QUESTION FROM THE INTERVENTION LOGIC |
What is this about? |
The intervention logic specifies the expected effects. Most evaluation questions concern one (or several) effects, which have to be specified. |
What is the purpose? |
To focus the question on the effects that are considered to be the most important or the least known.
|
Effect, need or problem |
Where possible and relevant, the question specifies the effect concerned. This is easy for effectiveness questions such as:
It is also preferable to specify the effect concerned in more complex cases, for example:
As shown above, many questions can be related to the intervention logic either directly (effectiveness) or indirectly (sustainability, efficiency, relevance). The only real exception is the question on unexpected effects. Effect
This question concerns an expected effect: better access to basic services. It is inferred directly from the intervention logic. Need
This question concerns a need that must be satisfied: better food for local communities dependent on fishing. It may be indirectly related to the intervention logic in so far as satisfaction of the need corresponds to an objective. Problem
This question concerns a problem. Is it realistic to want to improve the managerial capacity of local authorities in the current context of the country? The question is indirectly related to the intervention logic in that it concerns an objective. |
More or less extensive effects |
The question specifies whether it concerns:
Questions on sets of effects are of interest to policy-makers and strategic decision-makers but are generally more difficult to answer. All the effects
The scope of this question is very broad: all the impacts of European Aid in a country. Group of effects
This question concerns a group of logically related effects. A particular effect
This question concerns a precise effect, for a clearly defined public. |
Close or distant effects |
The wording of the question indicates whether one is interested in:
Questions on the most distant impacts are of interest to policy-makers and strategic decision-makers but are generally more difficult to answer. Reaching beneficiaries
This question concerns an initial result. Result
This question concerns a short-term result for direct beneficiaries. Intermediate impact
This question concerns an intermediate effect (improvement of access to services). A more direct effect would have been, for example: greater priority given to the territories neglected by local authorities. A more distant effect would have been, for example: reduction of regional disparities in access to basic services. Global impact
This question concerns a set of distant effects. A closer effect would have been, for example, local authorities' adoption of practices that involve actors in project design. . |
SPECIFYING THE EVALUATION CRITERION |
. |
What is this about? |
The questions are classified in different families that correspond to different "viewpoints" on what is being evaluated. Seven of these viewpoints, also called evaluation criteria, are to be considered: relevance, effectiveness, efficiency, sustainability, impact, coherence/complementarity, and Community value added. A question is drafted in relation to an evaluation criterion unless it is intended only for knowledge and understanding. |
What is the purpose? |
|
Focus on one evaluation criterion for each question |
Questions relating to several evaluation criteria are complex, for example: Has the intervention produced effect X satisfactorily compared to the objectives (effectiveness) and were the objectives in phase with the targeted beneficiaries' needs (relevance)? With this type of question it is probably necessary to implement two different evaluation methods: one to judge whether the expected effect was obtained and the other to identify the target group's main needs and judge whether the intervention satisfies them. The two methods use different survey and analysis techniques. Applying them both would probably entail doubling the costs or halving the quality of the answer. In this example it would be better to draft two different questions, to decide which one is more useful, and then to concentrate the resources on that question. |
Exception |
In some cases it is not possible to consider an evaluation criterion independently from another one, for example: Is effect X likely to continue after the end of financial support (sustainability)? In this example, we cannot answer the question without first having asked a sub-question: Has effect X been obtained (effectiveness)? In this case it is preferable to ask only one final question (sustainable effect). In preparing the answer, the evaluation team notes that a sub-question will be: Has effect X been obtained? |
The families of evaluation criteria |
This page proposes a typology of seven families of criteria. The first five correspond to the traditional practice of evaluation of development aid formalised by the OECD (DAC). The following two apply to all EC policies. Relevance The extent to which the objectives of the development intervention are consistent with beneficiaries' requirements, country needs, global priorities and partners' and donor's policies. Note: Retrospectively, the question of relevance often becomes a question as to whether the objectives of an intervention or its design are still appropriate given changed circumstances. Examples
Effectiveness The extent to which the development intervention's objectives were achieved, or are expected to be achieved, taking into account their relative importance. Note: Also used as an aggregate measure of (or judgment about) the merit or worth of an activity, i.e. the extent to which an intervention has attained, or is expected to attain, its major relevant objectives efficiently in a sustainable fashion and with a positive institutional development impact. Related term: efficacy. Examples:
Efficiency A measure of how economically resources/inputs (funds, expertise, time, etc.) are converted to results. Examples:
Sustainability The continuation of benefits from a development intervention after major development assistance has been completed. The probability of continued long-term benefits. The resilience to risk of the net benefit flows over time. Examples:
Impact Positive and negative, primary and secondary long-term effects produced by a development intervention, directly or indirectly, intended or unintended. Example:
Coherence/complementarity This criterion may have several dimensions: 1) Coherence within the Commission's development programme Example: Can it be said that the activities and outputs logically allow the objectives to be achieved? Are there contradictions between the different levels of objective? Are there duplications between the activities? 2) Coherence/complementarity with the partner country's policies and with other donors' interventions Example: Can it be said that there is no overlap between the intervention considered and other interventions in the partner country and/or other donors' interventions, particularly Member States? 3) Coherence/complementarity with the other Community policies Example: Is there convergence between the objectives of the intervention and those of the other Community policies (trade, agriculture, fishing, etc.)? Community value added The extent to which the development intervention adds benefits to what would have resulted from Member States' interventions only in the partner country. Examples
Use of the term criterion: A warning! This document concerns evaluation criteria, that is, the main ways of judging the intervention. In order to formulate fully transparent value judgements, the approach needs to be refined into evaluation questions, and then into judgement criteria. The word "criterion" is also used with a third meaning in the framework of quality assurance. In that case it concerns the quality assessment criteria of the evaluation. |
SHOULD QUESTIONS BE OPEN OR CLOSED? |
What does this mean? |
Generally questions requiring a 'yes or no' answer are avoided for two reasons:
|
Open questions requiring qualified answers |
Examples:
This type of wording is more appropriate if the question is intended to acquire knowledge or understanding, or to aid decision-making. In such cases the users expect qualified answers. |
Closed questions requiring "yes or no" answers |
Examples:
This type of question is more appropriate for accountability purposes, in a context where the objectives were set with precision. A closed question can also be useful if the intention is to validate a procedure or innovative instrument or to confirm a good practice. |