Newsletter 2, July 2002

Year of publication: 2003

This is the second newsletter informing you about our multi-country research into the sustainability of changes in hygiene behaviour. You will first learn a bit more about the background of the research and how it is set up. Then we will give you an overview of some preliminary findings of the study. This overview was made by IRC-colleague Kathleen Shordt. Last but not least you can again read contributions of the research teams from NETWAS-Kenya (Beth Karanja and Vincent Njuguna), NEWAH-Nepal (Renuka Rai and Saraswati Khanal), WaterAid-Uganda (Joyce Mpalanyi Magala and Brenda Nahinu), VRCWS-Ghana (Emmanuel Nyavor and Joan Awunyo-Akaba), COSI-Sri Lanka (Palitha Jayaweera and Nashidi Somarathne and SEUF-India (Suma Zacharia and Beenakumari Govindan). The teams undertook a first round of data collection and in June we all met in South India to discuss the preliminary analysis and to plan for the next steps of the research. The overview of the preliminary findings is based on their field studies. Should you want to react or require further information about the project, please get in touch with one of us. You find the contact details on the last page of this newsletter. If you want to receive a copy of the previous issue of this newsletter, please mail us or write us.

A brief recap about the research and the research partners

It is now close to two years since the start of this project, which aims to study the sustainability of changes in hygiene behaviour. The project combines a so-called concerted action financed by the European Commission (EC) and actual field research, financed by the Dutch government (DGIS) and the project partners. It will last three years and has the following objectives:

• To develop an active network in the field of hygiene promotion;
• To assess the level of sustainability of behavioural change one to three years after a hygiene promotion intervention;
• To develop a methodology for simple/cost-effective longitudinal monitoring of behavioural changes;
• To gain insight into relationships between project approaches, external conditions and sustainability of changes in hygiene behaviour;
• To determine policy and programming implications and influence policy to increase the effectiveness of water and sanitation programmes.

Early 2001 the research teams came together to define the research framework, to develop country specific research hypotheses as well as the tools for data collection.
The entire research centers around the following behaviours: handwashing, latrine use and maintenance and water storage. Data are collected from households and schools, whereby data collection is done through questioning, observation and demonstration.
More details can be found in the previous newsletter (October 2001), which can be found on IRC's website (www.irc.nl).

The first round of data collection

In June all research teams met in Kerala, South India, to discuss the outcome of the first round of data collection. The raw data, that were sent to IRC on forehand, had been analysed by Kathleen Shordt and Sandy Cairncross of the London School of Hygiene and Tropical Medicine. Already at this stage of the research the outcome of the analysis helped us to address a number of questions:
What can be said about the behavioural outputs that were expected after the hygiene promotion intervention? What can we say about our hypotheses? For which hypotheses did we find significant linkages and for which did we not? Do we have sufficient background information to already draw conclusions from the first round of data collection?
Some interesting findings appear to be emerging from the studies at this early point:

• There was not as much difference as expected between the 1998 and the 2000 communities. In some cases (Ghana and Nepal) 1998 did better than 2000. Perhaps the amount of time after the intervention is not as crucial as other variables.
• Some variables that are usually thought to be important for successful programming showed only mixed results. These included:
• Location of water point, access to water point was obviously related to handwashing in only 1 out of 3 studies. This might imply that construction (and good site selection) are not enough.
• Location of soap and water in the household determined good handwashing in only 2 out of 6 studies.
• Knowledge of critical handwashing times was better than handwashing demonstrations. In Uganda knowledge was related to good demonstration; in Uganda it was not. This fits with the usual idea that knowledge does not determine behaviour. This might imply that hygiene promotion and education should go beyond transmitting information.

Other trends in the data:

• The length of the project intervention seemed to be related to continued behaviours in Nepal and India.
• Design of latrines (user and child friendly) was related to latrine use or maintenance in both studies where it was tested (Nepal and Kenya).
• Reported handwashing was significantly related to the demonstrations in Kenya and India. India used household pocket voting; Kenya asked an innovative question about what "people usually do".
• Ghana showed an interesting relation between price of water and storage at home; people who said they would buy more if water were cheaper tended to store less water at home.
• Some external variables that are usually considered to be determinants of good hygiene behaviours appeared here. Thus, education of women and socio-economic level were determinants of various behaviours for handwashing, latrine use/maintenance and covering food.

The extensive exchange of project experience and findings allowed us to learn a lot about doing research and interpreting data. In a number of cases this led to adaptations for the next round of survey:

1. To illustrate that significant linkages are equally interesting as not-significant linkages Sandy Cairncross quoted Sherlock Holmes, who said: it is interesting to find out about why a dog barks in the night, but it is even more interesting to find out why a dog does not bark in the night.
2. While wanting to find out about women's level of education, the research team in Kenya wanted to prevent women from feeling embarrassed about their lack of formal education. They solved this by asking it in the following way: 'I am impressed by the way you have responded to all of my questions, may I ask, do you happen to have been at school?'
3. In India the research team used pocket-voting at the household level to cross-check information obtained through interviewing. Given the less positive outcome through pocket voting about latrine use and hand washing, it is assumed that pocket voting is more stimulating to bring out behavioural truth than interviewing.
4. Sometimes samples were too small for conclusions in terms of significance.
5. We realized that we need more background information about the communities and the hygiene promotion interventions to give a meaning to the data and to put them in perspective.
6. Still, even though we only have done one round of survey, data is already being used. When reading the contribution from the Ghanean team this becomes very obvious. It was found out that schools did not fulfil their project obligation. The hand washing facilities they were assumed to provide were absent. This made project management take up this issue more seriously with the schools.
7. Some of the data collection tools were adapted. It appeared that some of the data collected through the collection tools was not needed to address the hypotheses. Questions leading to superfluous data could therefore be deleted.
8. In some cases the sampling was adapted. The team in Sri Lanka for example, was advised to include some control villages since it is felt that that will allow for a more meaningful analysis of data.

Later this year the teams will do a next round of survey. Below you will find impressions of what the research teams did and found.