Measuring handwashing behaviour change and use(fulness) of M&E data

Printer-friendly versionPrinter-friendly version

The last official day of the symposium brought forth many interesting things to think about.  Some of the key things discussed were various ways of measuring hand washing behaviour and whether the information we collect is actually used. 

The discussion around measuring hand washing behaviour change outlined clearly the difficulties of measuring behaviours as opposed to infrastructure and offered promising examples of ways to measure change in a cost effective manner.  Direct observation (structured) is thought by the panelists to still be the best way to measure behaviours, however it does require both time, good training, and a significant amount of funds (depending on the size of the study).  Self-reported rates, though most often used by practitioners, provide the weakest set of indicators for actual hand washing practice.  Proxy indicators such as the presence of hand washing facilities, and water and soap at the hand washing facility, were found to be better measures of hand washing practice and are cost effective and relatively easy to measure through rapid assessments.  Another additional benefit of using proxy indicators is that they are often already measured through large national surveys such as the Demographic Health Survey (DHS) and the Multiple Indicator Cluster Survey (MICS).  These provide large data sets which allow us to infer behaviour change from the proxy indicator of the presence of a hand washing facility and soap and water.  It also allows for comparisons of those with hand washing facilities against key socioeconomic indicators, such as income and education level (the latest DHS data from multiple countries show a correlation between the presence of a hand washing facility in the home and higher levels of income, and higher levels of education).  Therefore, through the use of proxy indicators and large surveys such as the DHS and the MICS, we can have a better understanding of hand washing practice on the ground and potentially what influence these practices which can help improve our programming.  These indicators are promising in that although hand washing is an inherently difficult thing to measure, we are learning more and more about better ways to do this.

Another very interesting presentation was that of PhD student Rachel Norman from Cranfield University.  Rachel is currently conducting a study on how much WASH sector stakeholders are investing in M&E and how the data generated is being used.  Her study is being conducted in Uganda and Kenya. 

Her intermediate findings are that over time, there are an increasing number of indicators being monitored.  The general trend is that coverage is the most commonly reported on, while functionality less so.  The most common use (or purpose) of this data is for reporting, or as she put it, for description and explanation.  In fact, though we have more and more data and spend increasing amounts of money on M&E, we use very little of this information and there is no accountability for its use (the accountability mainly lies on collecting data for reporting and financial expenditure).  This brings forth interesting reflection questions for the sector: 

  • It appears that the trend is going towards more and more complex information being collected, but are we ready to make sense and make use of this information? 
  • What are the main reasons behind the lack of use of the information generated? Is it a question of lack of capacity to analyze and use the data, a question of usefulness of the indicators collected for practical action, or something else?
  • Linked to this is really the question of who needs to make which decisions, how often, and at which level, in order to improve WASH service delivery?

Although this may have been discussed at other sessions, I felt that this could have been discussed even more in the plenary sessions (though there were good side conversations held about this with many practitioners).   It would be great if the next symposium on WASH M&E can dive deeper into these types of questions so we can really take practical actions which can improve service delivery overall.  

Overall, it was a great learning space and good way to connect with many people thinking about issues not only on M&E, but on equity, sustainability, and many of the other components needed for service delivery.

Take a look at Andrés Hueso González's blog on his experience of the third day of the symposium.

Jolly Ann Maulit

Date: 15 April 2013


Submitted by franck flachenb... (not verified) on

Dear all,

I fully agree that self-reporting is not sufficient for measuring handwashing with soap, interviews must be coupled with direct observation.

Also, - this is my point - the questions should be phrased in a proper way: no more of "do you wash your hands with soap?" but rather "what are you using the soap for?"

So for minimal standards I would propose something like:

a.Have you used soap today or yesterday?

b. When you used soap today or yesterday, what did you use it for?

c. Can you show me / tell me where members of your household most often wash their hands?

Please if disagree, feel free to react and make other suggestions

Franck Flachenberg – Concern Worldwide

Submitted by Jolly Ann Maulit (not verified) on

Thanks for your input and ideas Franck. I agree that the way the questions are phrased in self-reports is key. In Malawi we are testing some questions to see which ones best correspond to actual practice (we also include some indicators for observation). Some of the questions we are trying are as follows:

1. Please list the times when you usually wash your hands. (Don't prompt answers.)

2. During the times mentioned in the question above, what do you use to wash your hands with?  

I agree with the way the soap questions are phrased; we are also including these in the test to verify the responses for question 2 (above), the hypothesis being that if they use soap habitually for hand washing, they will mention having used it today or yesterday. For soap we are also trying: When you have soap in the household, what are you using it for?  We'll have to see which ones are most effective as we continue to test the questions.  

We're finding a challenge with where members of households most often wash hands because this can vary depending on the critical time under question. For example, if they have a hand washing facility by the toilet they will normally use this location after toilet use, but a different location than this is often used for before eating or cooking, etc. We are realizing that these locations can vary depending on the activity, and they may not remain the same over time. Although the observation of a specific location for hand washing may reflect practice at certain critical times, it doesn't necessarily work for all critical times. It's a challenge that we're working on trying to solve. We'll share an update on our findings once we conclude the trials.