Means of Verification: What it is and how to use it

In the last article we anticipated that the means of verification would help us to measure the effective fulfillment of the indicators. If you remember, we wanted to know whether the teachers who had participated in the training had indeed increased their knowledge.

What are the means of verification used for?

Each indicator must have an associated means of verification to indicate where and in what form information on the achievement of the overall objective, the specific objective, and the results can be found. Indicators for which no suitable means of verification can be identified shall be replaced by other verifiable indicators.

It is important to bear in mind that:

• Practical and cost-effective means of verification have to be chosen. It is advisable to use realistic sources that are easy to collect and organize.
• It is not enough to indicate the source of the information. When you choose a method, be specific. For example, stating that your means of verification is the national statistics department leaves many questions open. Is the information up to date? Is it accessible? Is it updated on the basis of the required periodicity?
• When there is no information available for the indicator, you need to design a specific activity that consists of collecting the information. But if we are still in the design phase of the proposal, we also have the option of changing the indicator. What we cannot do is to leave the indicator without its means of verification.

Here are a few examples

Let us look at the example

Means of verification n. 1

In this example, the only practical and economical means of verification is the first one. It costs nothing for the technician who will deliver the job placement kits to have the women who will receive the kit sign a list. The distribution list is a good, easy and cheap way to show that women have received the kit.

Means of verification n. 2

Let’s move to the second means of verification of the example.

However, the second one already creates some difficulties. Women may be working but in the informal sector, for example selling sandwiches on the street. Or they may be entrepreneurs in their own business. In both cases, the indicator has been achieved, the women are working, but the means of verification is not good as it asks for something that does not exist: a signed contract. In this case it is necessary to find another means of verification to help us support that 40% of the women who have received a job placement kit are working.

One means of verification that can work is a written report following a direct observation: a monitoring visit could be carried out where the women who have been in the training are interviewed to see what their status is. This means of verification is time-consuming, as distances may be long and contacts with the beneficiaries may have been lost. An ad hoc activity with allocated funds would undoubtedly have to be designed in order to be able to carry out the data collection.

Another plausible means of verification for indicator 2 could be the business’ accounting books. Whether informal or self-employed, traders keep clear accounts of their businesses however small they may be. This would help us to verify our indicator.

Certainly, if this analysis had been done at the design stage of the intervention, we would probably have come up with a different type of indicator. Instead of asking how many people are working, we could have asked how the income of the people who have received the insertion kit has changed. For example, the indicator could be transformed into:

In this case, obtaining the information would require some analytical and tabulation work, but it would give us concrete and objective information.

Means of verification n. 3

Regarding the third proposed example, what we should be sure of before indicating this as a means of verification is the frequency with which the information is updated. If the frequency of updates does not allow us to have the data for the year we are interested in knowing in time, we would not be able to use this source of information as a means of verification of the indicator. We could investigate whether other institutions, more local and linked to the community, have some kind of record on unemployment, even if it is unofficial information. Otherwise, the only alternative would be to change the indicator, as a census would be too costly in terms of both time and money.