Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

Other Characteristics

The families included permanent servants, a term that refers to someone who is generally attached to a family for a prolonged period of time. This kind of service can occur in several ways; for example, either a young child from an extremely poor family may be "given" to another family or a man in debt may offer himself to a family in return for payment of his debt. The mean family size of the 40 families was 8.7 with permanent servants and 8.2 without permanent servants; family size ranged from 3 to 23, including permanent servants. The sex ratio was relatively even: 180 (51.6 per cent) males and 169 (48.4 per cent) females. Table 2 summarizes these characteristics.

TABLE 2. Selected characteristics of Dokur households

The age distribution of the study respondents was like that found in most developing countries; the high proportion of young people in the sample is illustrated in table 3.

In virtually any Indian village where there are employment possibilities, nearly all inhabitants work, including children and older adults. The principal occupations of the sample are listed in table 4.

Respondents were also asked if they had secondary occupations that they followed. Fewer than one-quarter of the sample responded positively. Twenty-seven labourers (7.7 per cent) listed agriculture as a subsidiary occupation, while 25 farmers (7.2 per cent) said they worked as agricultural labourers. The remaining 10 per cent indicated that they worked at a variety of tasks, for example as shepherds (five people).

The forty families in the study represented the major castes, eight of the fourteen to be found in the village.

TABLE 3. Age distribution of the study respondents

TABLE 4. Principal occupations of respondent household members

a. Twenty permanent servant, were employed by the study families and ate with them The remaining 12 were respondents who worked for other families. but lived and ate in their own homes. Permanent servants are common in Andhra Pradesh
b This group includes 98 children and 7 older adults

TABLE 5. Relative rank of respondent castes with traditional occupation^a

a. N=40.

TABLE 6. Family types represented by the study respondents

Some castes could not be selected-for example, the trade castes that include washermen (dhobies) or barbers, and, of course, the one village brahmin who serves as the patwari. (The brahmin's position, which is government appointed and involves such duties as keeping land records and recording sales, has recently been abolished.) Table 5 shows the relative rank of the castes with their traditional occupations.

The types of families in the study fell into five categories commonly recognized by demographers and described as follows:

1. Nuclear: one couple plus any children.
2. Supplemented nuclear: nuclear with the addition of a relative, e.g. mother.
3. Lineal joint: more than one couple of two generations.
4. Collateral joint: more than one couple of the same generation, e.g. married siblings.
5. Lineal-co/lateral joint: more than two couples of the same and different generations.

Table 6 presents the breakdown of the present families.

The total owned land and the operational land size of the study households are shown in table 7. These data were obtained only for the kharif season of the year before the study started. in general the operational land will be less than the owned land. Of the 38 landowning families, 34 were actively farming, and of those, 29 (85 per cent) had irrigated land, as shown in table 8.

TABLE 7. Land ownership among the families

a. Operational size was defined as the area of owned land, excluding the area leased or sharefarmed to someone else, but Including any land leased or shorefarmed from some other landholder.

TABLE 8. Irrigated land as a percentage of operational land

Data on land ownership collected in 1975 by the Economy Program of the International Crops Research Institute for the Semi-arid Tropics (ICRISAT) [3] demonstrated that one-third of all Dokur farm households owned less than 0.90 ha of land, while another one-third owned less than 2.11 ha. The remaining third had in excess of 2.11 ha of land. Using these data for comparative purposes, it may be seen that the study population is slightly over-represented at the lowest level, with 15 rather than 13 families, and under-represented at the middle level with 8 instead of 13 families. In the large farm category there are 17 families, or 4 more than expected according to distribution pattern determined by ICRISAT. However, land transaction records, as well as evidence from bank loan records supplied by the State Bank of India, would appear to indicate that our sample distribution reflects a more current picture of land-holding patterns, i.e. large farms are getting larger at the expense of small and medium-size farms.

Methodology

The study was designed to provide comprehensive information in a number of disciplines, although it was understood that the investigators lacked expertise in many of these fields. For this reason, advice was sought from professionals in some areas of study. There were, however, the inevitable omissions.

The two field investigators, B. Nag and P. Babu, both had previous field experience of a limited nature. Both lived in the village during the week for the entire period of study which was anticipated to last from early February 1982 until late April 1983. During this time, only one week was missed and information that could be collected retrospectively was obtained the following week.

The study team was fortunate in being offered the use of a cowshed by the sarpanch (elected village leader) and his family. A wall was constructed that left one-third of the shed for the team and two-thirds for the cows. Electricity was brought from the main house and a single light bulb and an outlet for a much-needed fan were provided. The team was able to move into its quarters in mid February. During the next few weeks, until the study officially began in the third week of April, the respondents were selected, a census taken, and data schedules were pre-tested. Each investigator was responsible for covering 20 families. Jointly the team took monthly measurements of height, weight, and arm circumference. The physical examinations were performed by a consultant physician with the assistance of the investigators.

In the following sections the methodology used is presented, along with a discussion of some of the difficulties encountered while trying to follow the original plan. The sections cover the following areas: census, dietary recall, morbidity, physical examinations, anthropometry, labour force participation, crop production, and assets.

Census

The investigators went as a team from house to house where they gathered information, as a general rule, from the head of the household and/or other older members of the family. To determine age, a difficult task in any developing country, several methods were used in combination. These included developmental age assessment and a village event calendar, as well as family records and horoscopes. When other members of the household were present, they contributed to the task of determining such questions as the number of years of education. Information was collected in the general areas of education, occupation, landholding size, farm inputs such as irrigation and access to wells, crops grown during each of the three seasons, and livestock owned. Much of this general information has been presented in the section on household characteristics. The landholding information was subsequently verified using the patwari's records, as all land transactions must be registered with him.

Food Consumption

Much has been written about the difficulties of assessing food intake in populations [20, 29, 32]. Each method has its strengths and weaknesses; no method is truly able to capture the usual variation found in individual diets. For this study, the 24-hour recall method was chosen as the basis for a weekly, one-day estimation of food intake. The reason for the selection of this technique was the low demand on the subjects interviewed. As one essential aspect of the entire project was to be its longitudinal nature, it seemed important to place as few demands on the respondents as possible, while at the same time collecting a maximum of useful information.

The female member(s) of the house who were responsible for cooking and serving food gave the necessary information to the investigators. If other members of the household were present, they may also have contributed to the recall procedure. At the same time, each person's diet was recorded as "normal" or "not normal"; if it was the latter, a reason was given. Diets which were not normal generally fell into several categories: solely breast-fed; breast-fed plus food; meals eaten away from home, e.g. went to market town; festival or marriage diet; illness responsible for abnormal pattern/amount; no available information, e.g. person away from village.

Data were also collected concerning whether or not each woman in the study sample was lactating or pregnant; if she was pregnant, what was the stage of pregnancy? The consumption of food outside the home was ascertained whenever possible, and included such items as tea, puri and dose, which were commonly eaten in local shops. To ascertain the quantity of these food items, shop owners were contacted and the ingredients and quantities recorded. These values were then used. In accordance with general procedure, the meal eaten most recently was recorded first and the investigator worked backwards. Food left over or fed to animals was also tabulated. A few items consistently recorded as part of the total day's consumption included sweets such as hard candies, toddy, and betel leaves used for making pan. If cooked food was borrowed from others, this was also taken into consideration.

Each investigator was supplied with a set of thirteen nested aluminium pans of the type commonly used by the villagers, ranging from 17 ml to 2,300 ml capacity. A set of measuring spoons was provided for measuring small amounts of food, such as oil or spices, and local measures such as the seer were also used to determine quantities consumed. These measures were used primarily for quantifying amounts of cereals and cooked food. For such foodstuffs as vegetables, the actual amount purchased in kilograms was noted. For fruit, if not purchased, the number and size was recorded and the nutrient value supplied by the National Institute of Nutrition (NIN) was used to calculate the value.

Before the actual study period began, the investigators repeatedly weighed and measured quantities of all basic foodstuffs commonly eaten by the villagers. Both cooked and raw quantities were checked. In addition, a local measure, the seer cup, was also standardized.

As it was impossible for us to measure milk consumption by breast-fed infants and children, data on mean intake of breast milk from extensive studies of south Indian children were used [19]. These data covered children from birth to 48 months of age. Data were averaged over two-month periods for children up to the age of one; three-month periods for those up to 18 months; and six-month periods for those up to four. The upper figure was used for the few children in the study who were still being breast-fed after 48 months of age.

A programme was written to calculate nutrient availability from the gram amounts of each food consumed by each respondent. The first step involved the calculation of the nutrient value of each diet record using nutritive values for Indian foods compiled by staff of the National Institute of Nutrition [8]. Nutrients selected for analysis included: protein, energy, fat, calcium, iron, B-carotene, thiamin, riboflavin, niacin, ascorbic acid, lysine, tryptophan, and sulphur-containing amino acids plus leucine and isoleucine. ß-carotene was selected for analysis for several reasons. First, there is little pre-formed vitamin A in villagers' diets in south India; what small amounts are present generally come from meat and milk products. Second, nearly all the foodstuffs recorded in the Nutritive Value of Indian Foods [8] list ß-carotene content only. As the recommended nutrient allowances for Indians are given as m g of b -carotene, it seemed far simpler to convert the IU of vitamin A given for meat and milk into b -carotene equivalents.

Programmes were subsequently written which provided the nutrient contribution made by each food in a day's diet as well as the nutrient value of the total day's diet. These values were then converted into a percentage of the daily allowances of nutrients for Indians according to Gopalan et al. [8], which take into consideration a range of ages and physiological states. The moderate activity level was chosen for both men and women.

Morbidity

Morbidity data were collected once weekly, but cover the entire week's illnesses. A preceded morbidity schedule was used and the investigators were instructed to proceed down the list asking if any of the recorded symptoms had been experienced during the past week. Codes were also given to individuals who could not be contacted or were out of the village as well as to those who had no complaints.

Table 9 specifies the morbid conditions recorded by the investigators, first as general categories and second as more specific complaints. The duration of each condition was also recorded.

continue

Contents - Previous - Next