What is an indicator?
It’s vital to have a unit or variable “in which” or “by which” a measurement is made, i.e., an indicator, in order to measure something. Finding performance indicators that will allow us to “objectively verify” whether or not our interventions have accomplished the planned activities, outputs, outcomes, and impacts is necessary while putting together a log frame or impact chain for a project.
Creating relevant performance indicators to gauge project performance is the program manager’s primary challenge. These metrics track what projects do, what they produce, the changes they bring about, and the outcomes of those changes. Decisions regarding what to measure must be made before indicators can be chosen. The foundation of efficient project execution and sound M&E procedures is having the appropriate indicators. As a result, their identification, choice, and use should be given careful consideration.
We’ll now look at some common challenges to choosing indicators.
- Selecting an indicator that is unaffected by the program’s activity
Imagine, for instance, a program that aimed to increase access to AIDS preventive and treatment services by educating healthcare professionals in these areas. The proportion of healthcare facilities with adequate conditions to offer care was the UNAIDS indicator used by the M&E plan’s designers. However, a variety of factors, including monitoring, the accessibility of supplies and equipment, and the development of suitable treatment procedures, can influence this indicator. The intended training program would not address any of these factors. The planners failed to take into account the fact that this global indicator did not fairly represent the program activities when using it. The number of clinicians who have received training or the number of facilities that have a trained provider would be better indicators.
- Choosing an ambiguous indicator
Consider a radio campaign to debunk particular myths about the spread of HIV/AIDS, for instance. Although raising awareness of HIV/AIDS is the campaign’s ultimate objective, the indicator proportion of the population that has this information does not make clear what specific knowledge is being raised. A more accurate statistic would be one that quantified the campaign’s goal, such as the percentage of people who reject myths X and Y concerning HIV/AIDS transmission.
- Choosing an indicator that depends on insufficient data
For instance, a program addressing challenges with the drug supply chose an indicator that showed the percentage of days per quarter that service delivery locations have drug stock outs. To offer this information, stock out data may not, however, be gathered frequently enough. A more accurate measure would be the percentage of service delivery locations where there was a medicine shortage at some point in the previous quarter. Additionally, population-level statistics may be unavailable or challenging to get. For instance, the baseline immunization coverage rates for a certain population might not be known.
- Choosing an indicator that is inaccurately indicative of the desired result
What would a suitable indicator be, for instance, if an IR said greater access to antiretroviral (ARV) therapy for expectant women to prevent mother-to-child transmission (PMTCT) of HIV?
Question: Would the indicator proportion of pregnant women taking ARVs be suitable?
Answer: As opposed to telling us how many pregnant women with HIV are using ARVs, this indicator tells us how many women overall who are taking ARVs are pregnant.
In other words, the number of pregnant women using ARVs is the numerator of this indicator, while the number of women taking ARVs is the denominator. Assume there were 400 women using ARVs in total and 100 of them were pregnant. The ratio would be 100/400, which can be expressed as 1/4 or 25%. The indicator would decline if the denominator grew, that is, if more non-pregnant women received HIV therapy but the number of pregnant women receiving treatment remained constant.
For instance, the percentage would be 100/1000, which is equal to 1/10 or 10%, if 1000 women were using ARVs. The indicator would show this shift, but it has no bearing on the program’s intended result, which is an increase in the proportion of pregnant women taking ARVs. Similar to this, if the indicator increased, such as if the proportion of pregnant women on ARVs increased from 25% to 50%, it might be because more pregnant women received ARV treatment, which would be the desired outcome, but it might also be because fewer non-pregnant women were taking ARVs, which would not be related to the program’s desired outcome. This would not be an appropriate indicator to use because it is unclear which modification took place.
Question: Let’s attempt another illustration. Would the indicator proportion of pregnant women taking ARVs be suitable?
Answer: No, this would not be a suitable indicator either.
Here, the denominator is the total number of persons taking ARVs, including all men, women, and children receiving treatment (let’s assume it is 5,000), and the numerator is the number of pregnant women taking ARVs (let’s say it is 100 again). In other words, this signal would inform us that 100/5000, or 1/50, or 2%, of all those using ARVs, are women.
If this indicator rose over time, say from 2% to 20%, it might be due to an increase in the number of pregnant women receiving ARV treatment (1000/5000, which is the program’s intended outcome), but it might also be due to a decrease in the number of people receiving this treatment overall (100/500), meaning that the proportion of pregnant women receiving treatment remained constant. The indicator may have decreased if more people were receiving treatment overall, fewer women were HIV-positive, or there were fewer pregnant women. Therefore, it would be difficult or impossible to appropriately evaluate the data presented by this indicator.
Let’s try another illustration:
Question: Would it be appropriate to use the indicator percentage of pregnant women who are HIV-positive and taking ARVs?
Answer: Yes, this indicator would give the necessary information, in response.
Here, the denominator is the total number of HIV-positive pregnant women, and the numerator is the number of pregnant HIV-positive women who are taking ARVs.
With this measure, factors unrelated to the IR, include a drop in the prevalence of HIV among pregnant women or a rise in the number of non-pregnant women getting ARVs.
Guidelines for choosing indicators
The following are some general principles for choosing indicators:
- Pick data-required indicators that can be gathered using the available resources.
- For each important activity or result, choose at least one or two indicators (preferably from various data sources).
- Pick at least one metric for every core activity (e.g., training event, social marketing message, etc.).
- Pick no more than 8–10 indicators per important program emphasis area.
- Whenever possible, use a variety of data collection sources.
To learn more with regard to indicators in MEAL, register for a Monitoring, Evaluation, Accountability & Learning (MEAL) programme today.