I arrived in Chennai on July 18th, intent on doing some work for Sangam India (an NGO I had become acquainted with during my option). I had spoken to Sriram a couple of times during my time in Delhi and he convinced me that there was a significant task which needed to be carried out: a survey for the new slum that they had “adopted.” I arrived and chatted with Nivedita and Sriram (the founders and co-chairs of Sangam India) regarding the work that would need to be done and when it would need to be organized for. I got to work immediately, with help from Katie Bush, a PhD student from the University of Michigan.

There are three areas to the survey in general: the consent form, the census sheet and the survey itself. I’ll deal with them in that order.

The consent form was designed by Katie but edited by myself.  We wrote some very basic text in English and Tamil, so that if the occupant was literate in one or the other, they could sign in that language. However, many of the interviewees were not literate in any language, and therefore had to have the consent form read out to them. In a few cases, they marked their acceptance with a thumbprint. Some others have agreed to answer the questions (and then done so) but have not filled in the consent form. This may have been an oversight on the part of the interviewers or due to a lack of anything to form a thumbprint with for illiterate interviewees. The consent form basically states that there is no immediate monetary benefit to conducting this survey, that all questions are answered voluntarily, that data will be published anonymously and will otherwise remain confidential and that consent is required before the first question may be asked.

The census form is based on a tried and tested format which was used for the pilot census which Sangam India used in Ramavaram, the slum they worked with previously. It asks for the name, age, position (in relation to head-of-family), caste, occupation and education status for each member of the household. Prior to this is the name of the interviewee, their age and gender, the house number and any phone number they may have and their monthly household income.

Katie and I designed the questions, aiming at the areas we deemed most important. It was partly adapted from the National Family Health Survey 1992-1993. This included 6 broad areas:  hygiene, environment, household, nutrition, health and education. A significant deficiency in any of these could have long term effects for the slum inhabitants in the future. Below, I’ll elaborate on each of these sections.

Hygiene focused on personal habits including bathing, washing hair, trimming nails, dental hygiene, hand washing, wound treatment and the use of footwear and underwear. It also included one question on menstrual health, which we aim to build on in future visits. We had originally intended to have an extended segment on women’s health, but we felt that this would have been too forward. Since this community has very few ties with the slum Sangam India previously worked with, we are aiming to build trust over a period of time. With this in mind, we felt that this information would be better gathered at a later date. The relatively poor response rate for this question seem to back up this decision.

The section on the environment was a brief one, designed to educate us on the source of water (separately for washing and drinking), lighting and to illuminate how the slum dwellers disposed of their garbage. An important inclusion in this was of what kind of toilet facility the houses had. The answer, in most cases, was none. This section brought us the most information, since we didn’t know that there was no drainage system and no means to move rubbish from the river and behind the houses (where most of it is dumped) to anywhere else.

The household section covered whether the house was structurally sound, the number of rooms and people per room, as well as other basics details about the house. Included in these were bathroom and kitchen status, drainage, pets (and whether they stayed inside or outside the house), water storage and treatment and asked whether there was a basic cleaning routine. These questions were important so we can see what the slum is like in terms of houses (whether the abodes are more temporary or permanent) and the conditions of the residents within. The results showed that there was a variety in terms of building stability as well as the population per room. The latter, in particular, is a problem which we must address as soon as possible, as the fires occurring in MGR Nagar are consuming more and more houses, leaving people with little option but to live outdoors or depend on the charity of their neighbours and share dwellings. In the slum, however, this means cramming an unreasonable number of people into a space which is simply too small to be healthy. Other striking findings are that the bathroom and toilet facilities are all but non-existent. The facilities were often marked as “shared”, only for the elaboration to come back that people performed their ablutions freely in a field or lake, polluting the resources nearby which some other residents relied on for fresh water. As a final, related note, there were very few houses which treated their water in any way before consuming it. Though the local government has provided some metro-tanks, my understanding is that this water should be boiled or at least filtered, as its source is bore-hole wells, before drinking it.

Health covered three areas: status, access and vices. Status was included to assess the overall wellbeing of the community. We put together a few very basic screening questions to find out the prevalence of blindness, TB, malaria, diarrhoea and physical disability. The majority of health was dedicated to access, as this is the area in which we can help out the most. We asked which health centres they had used in the past and which they would consider returning to. The reason for this two-part question was to eliminate any centres which were nearby but which were not being used by the residents for whatever reason. Finally, “vices” covered two questions which covered whether anyone in the house smoked cigarettes or drank alcohol. There were many interviewees who freely admitted drug use outside of this realm, including some admissions of betel nut use. One problem with this section is that it changes depending on who was being asked. One interviewer observed that whenever the male head of the family was asked, neither alcohol nor tobacco was a problem, but when his wife or children were questioned, a different story emerged.

The education section was designed (by a separate education committee) with a few thoughts in mind. We wanted to know the number of children attending school and if there was any reason why they weren’t attending. The idea of providing vocational training for adults and children alike was voiced in one meeting, and we wanted to see if there were enough people who would commit to such a programme to consider running this. Other than this, we wanted to know what type of training they would consider. Finally, we were interested in the children who were not attending school and what they were doing instead. If they were working as labourers, for example, we were interested in giving them vocational training too, or helping them back into school. My interpretation of the results was a little disappointing, as a large proportion of interviewees gave their family’s highest educational qualification as ‘nil’, meaning that none of them had had any formal schooling. More worrying were the number of surveys which were simply left blank, indicating a complete lack of interest.

Finally, there is nutrition. We appreciated that this would be an important section to cover, but we did not anticipate how difficult it would be to get the data in a usable way. I had to abandon using the nutrition data since it was not specific enough. I had not made this section multiple choice, like many other aspects of the survey, as the sheer frequency that one eats meat, vegetables, paruppu or fruits could vary significantly. I realize that by doing this, I was being a little overambitious. While the question said, ‘How many times per week will you eat X?’, the answers often came back as ‘Whenever possible’ or ‘Daily.’ The latter was the more frequent answer, and the more annoying. Obviously, the inhabitants were commonly consuming X, but whilst ‘Daily’ for us clearly means ‘once per day,’ the nutrition habits of the slum dwellers seemed to be based around cooking as few times as safely possible, which might mean that ‘daily’ means one, two, three or more meals per day. Problems of this nature meant that I had to scrap even trying to recover data from this section.

One minor setback was that we designed the survey to be answered by one member of each household, though we soon realized that more than one family sometimes occupied one house. On the day, we gave spoken instructions to the volunteers to use one survey per family. The survey itself was carried out very well. Our volunteers were efficient and made a connection with the inhabitants. I was very aware that this survey was only partially a data collection exercise and was just as much to do with forging the kind of bond which can be relied on in later excursions when we are attempting to prove our intentions using action.

Download the Survey Materials (pdf):

• Survey Form
• Surveyors’ Handout
• Consent Form