CHAPTER 1 INTRODUCTION

Transcription

1 CHAPTER 1 INTRODUCTION 1.1 Background o the Survey India s irst National Family Health Survey (NFHS-1) was conducted in (International Institute or Population Sciences, 1995). The Ministry o Health and Family Welare (MOHFW) subsequently designated the International Institute or Population Sciences (IIPS), Mumbai, as the nodal agency to initiate a second survey (NFHS-), which was conducted in An important objective o NFHS- is to provide state-level and national-level inormation on ertility, amily planning, inant and child mortality, reproductive health, child health, nutrition o women and children, and the quality o health and amily welare services. Another important objective is to examine this inormation in the context o related socioeconomic and cultural actors. The survey also provides estimates at the regional level or ive states (Bihar, Jammu and Kashmir, Madhya Pradesh, Rajasthan, and Uttar Pradesh) and or three metro cities (Chennai, Kolkata, and Mumbai), as well as slum areas in Mumbai. This inormation will assist policymakers and programme administrators in planning and implementing strategies or improving population, health, and nutrition programmes. Comparative state results rom NFHS- have already been published (International Institute or Population Sciences and ORC Macro, 000). The current report provides a more comprehensive picture o the indings or Maharashtra. The NFHS- national sample covers more than 99 percent o India s population living in all 6 states. It does not cover the union territories. NFHS- is a household survey with an overall sample size o 90,0 ever-married women in the age group living in 9,486 households. NFHS- was conducted with inancial support rom the United States Agency or International Development (USAID), with additional unding rom UNICEF. Technical assistance was provided by ORC Macro, Calverton, Maryland, USA, and the East-West Center, Honolulu, Hawaii, USA. Thirteen ield organizations were selected to collect the data. Eight o the ield organizations are private sector organizations and ive are Population Research Centres (PRCs) established by the Government o India in various states. Each ield organization had a responsibility or collecting data in one or more states. The Centre or Operations Research and Training (CORT), Vadodara, was selected as the ield organization or NFHS- in Maharashtra. 1. Basic Socioeconomic and Demographic Features o Maharashtra Situated in the western part o India, the present state o Maharashtra was created on 1 May 1960 out o the territories o ormer Bombay state, Madhya Pradesh, and Andhra Pradesh. It has a land area o 07,71 km which is about one-tenth o the total land area o the country. As per the 001 Census o India, the total population o the state is 96.8 million, which is 9.4 percent o the total population o India. In terms o population size, it is the second largest state in India, next to Uttar Pradesh. The population density o Maharashtra has increased rom 57 persons per km in 1991 to 14 persons per km in 001. The state has six divisions Konkan, Pune, Nasik, Aurangabad, Nagpur, and Amravati which are urther divided into 5 districts. Forty-six percent o the state population is concentrated in two divisions Konkan and Pune which orm

2 the most industrialized part o the state including Mumbai. According to the 001 Census, the population density within the state ranges rom 08 persons per km in Nagpur division to 807 persons per km in Konkan division (Director o Census Operations, Maharashtra, 001a). Maharashtra is one o the most economically developed states in India. The per capita state income at current prices was Rs.,98 during Among the 1 major states, in Maharashtra was second only to Punjab in terms o per capita income. At constant ( ) prices, the per capita state income doubled rom to , rom Rs.,45 to Rs. 4,85. The average annual growth rate o per capita state income was.8 percent during the six years preceding (Directorate o Economics and Statistics, Government o Maharashtra, 001; EPW Research Foundation, 1998). During the year , the share o the primary, secondary, and tertiary sectors in the total state income was 17, 0, and 5 percent, respectively. The corresponding shares in the year , when the state was created, were 4, 6, and 40 percent, respectively, which indicates a remarkable decline in dependence o the state economy on the agricultural sector and an increase in the share o the secondary and tertiary sectors. The secondary and tertiary sectors account or a much larger share o state income (8 percent) than the national average (70 percent). The latest available data on actory employment or 1998 indicate that Maharashtra continues to lead the country in average daily actory employment. During the period , the annual growth rate in the total number o actories was.8 percent. Under the state-sponsored Employment Guarantee Scheme, which has been implemented since 197, 95 million person-days o work was provided during the year (Directorate o Economics and Statistics, Government o Maharashtra, 001). According to estimates o the Planning Commission, more than one-third o the state population (7 percent) is below the poverty line and there is little rural-urban dierence in this respect (8 percent in rural areas and 5 percent in urban areas) (Central Statistical Organisation, 1999). As ar as social development is concerned, Maharashtra occupies the second position among the major states, next to Kerala, in terms o two key indicators literacy and inant mortality. During the period , the literacy rate or males has increased rom 49 percent (among the population age ive years and above) to 86 percent (among the population age seven years and above). The emale literacy rate has increased more than threeold rom 0 percent to 68 percent during this period. In 001, Maharashtra s overall literacy rate was 77 percent, whereas or Kerala it was 91 percent (Director o Census Operations, Maharashtra, 001a; Oice o the Registrar General and Census Commissioner, 001). In 1999, the inant mortality rate or Maharashtra was 48 deaths per 1,000 live births, whereas or Kerala it was 14 deaths per 1,000 live births (Oice o the Registrar General, 001). Maharashtra has a rich heritage o contributions rom social reormers and leaders like Mahatma Phule, Maharishi Karve, Agarkar, Karmaveer Bhaurao Patil, and Dr. Bhim Rao Ambedkar. Even today, Maharashtra has an excellent network o dedicated social workers and voluntary organizations working in the ields o education, rural development, and empowerment o women. According to the 1991 Census, 11 percent o Maharashtra s population belonged to scheduled castes and 9 percent belonged to scheduled tribes. From , Maharashtra had the distinction o being the most urbanized state in India, but in 001, Maharashtra (4 percent urban)

3 lost this position to Tamil Nadu (44 percent urban). During the period , the urban population in the state has increased rom 8 percent to 4 percent o the total population. The number o cities with one million or more population in the state increased rom our in 1991 to seven in 001 and the population o Mumbai increased rom 9.9 million in 1991 to 11.9 million in 001 (Director o Census Operations, Maharashtra, 001b). According to 001 census o India, 48.9 percent o the population o Mumbai lives in slum areas (Director o Census Operations, Maharashtra, 001c). The total population o Maharashtra doubled in the three decades ater the creation o the state in 1961, reaching 78.9 million in 1991 and 96.7 million in 001. Population growth in Maharashtra during the last our decades has contributed 10 percent to the decadal population growth o the country. The average exponential population growth rate in Maharashtra declined rom.9 percent in to.04 percent in Maharashtra has been a destination or a large inlux o migrants rom other parts o India. Net migration has contributed 19 percent to its population growth during the last decade. The crude birth rate in Maharashtra declined rom. births per 1,000 population in 1971 to 1.1 in The crude death rate declined rom 1. deaths per 1,000 population in 1971 to 7.5 in 1999 and the inant mortality rate declined rom 105 deaths per 1,000 live births in 1971 to 48 in 1999 (Oice o the Registrar General, 001; Oice o the Registrar General, 1999b). The expectation o lie at birth in Maharashtra in is estimated to be 65. years or males and 68.1 years or emales (Ministry o Health and Family Welare, 1999a). According to the Sample Registration System, the total ertility rate or Maharashtra declined rom 4.6 children per woman in 1971 to.7 in Maharashtra was one o the pioneers in the ield o amily planning. Even beore independence, valuable work in this ield was done by Pro. R.D. Karve, who started a birth-control clinic as early as 191, and Shakuntala Paranjape, who continued his work. The government amily welare programme in Maharashtra was launched in 1957 and, as an important step towards a decentralized approach, it was handed over to Zilla Parishads as early as in Maharashtra has achieved many national awards or the excellent perormance o its amily planning programme. Since its inception in the irst 5 years, Maharashtra s amily welare programme perormed more than 10 million sterilizations and about hal a million IUD insertions (Study Group on Population o Maharashtra, 199). Maharashtra s couple protection rate increased rom 5 percent in 1980 to 51 percent in Yet, at.7 children per woman, the total ertility rate or Maharashtra is well above the replacement level o.1 children per woman, which is the goal to be reached by 004 according to the New Population Policy announced by the Government o Maharashtra in May 000 (Department o Public Health, Government o Maharashtra, 000). 1. Questionnaires NFHS- used three types o questionnaires: the Household Questionnaire, the Woman s Questionnaire, and the Village Questionnaire. The overall content and ormat o the questionnaires were determined through a series o workshops held at IIPS in Mumbai in 1997 and The workshops were attended by representatives o a wide range o organizations in the population and health ields, as well as experts working on gender issues. The questionnaires canvassed in Maharashtra were bilingual, with questions in both Marathi and English.

4 The Household Questionnaire listed all usual residents in each sample household and also any visitors who stayed in the household the night beore the interview. For each listed person, the survey collected basic inormation on age, sex, marital status, relationship to the head o the household, education, and occupation. The Household Questionnaire also collected inormation on the prevalence o asthma, tuberculosis, malaria, and jaundice, as well as three risk behaviours chewing paan masala or tobacco, drinking alcohol, and smoking. Inormation was also collected on the usual place where household members go or treatment when they get sick, the main source o drinking water, type o toilet acility, source o lighting, type o cooking uel, religion o the household head, caste/tribe o the household head, ownership o a house, ownership o agricultural land, ownership o livestock, and ownership o other selected items. In addition, a test was conducted to assess whether the household uses cooking salt that has been ortiied with iodine. Finally, the Household Questionnaire asked about deaths occurring to household members in the two years beore the survey, with particular attention to maternal mortality. The inormation on the age, sex, and marital status o household members was used to identiy eligible respondents or the Woman s Questionnaire. The Woman s Questionnaire collected inormation rom all ever-married women age who were usual residents o the sample household or visitors who stayed in the sample household the night beore the interview. The questionnaire covered the ollowing topics: Background characteristics: Questions on age, marital status, education, employment status, and place o residence provide inormation on characteristics likely to inluence demographic and health behaviour. Questions are also asked about the background characteristics o a woman s husband. Reproductive behaviour and intentions: Questions cover dates and survival status o all births, and current pregnancy status and uture childbearing intentions o each woman. Quality o care: Questions assess the quality o amily planning and health services. Knowledge and use o contraception: Questions cover knowledge and use o speciic amily planning methods. For women not using any contraceptive method, questions are included about reasons or not using contraception and intentions or uture use. Sources o amily planning: Questions determine where a user obtained the amily planning method. Antenatal, delivery, and postpartum care: The questionnaire collects inormation on whether women received antenatal and postpartum care, who attended the delivery, and on the nature o complications during pregnancy or the last two births since January Breasteeding and health: Questions cover eeding practices, the length o breasteeding, immunization coverage, and recent occurrences o diarrhoea, ever, and cough or young children. Reproductive health: Questions assess various aspects o women s reproductive health and the type o care sought or health problems. 4

5 Status o women: The questionnaire asks about gender roles, women s autonomy, and violence against women. Knowledge o AIDS: Questions assess women s knowledge o AIDS and the sources o their knowledge, as well as their knowledge about ways to avoid getting AIDS. In addition, the health investigator in each survey team measured the height and weight o each woman and each o her children born since January This height and weight inormation is useul or assessing levels o nutrition prevailing in the population. The health investigators also took blood samples rom each woman and each o her children born since January 1996 to assess haemoglobin levels. This inormation is useul or assessing prevalence rates o anaemia among women and children. Haemoglobin levels were measured in the ield at the end o each interview using portable equipment (the HemoCue) that provides test results in less than one minute. Severely anaemic women and children were reerred to local medical authorities or treatment. In addition, health investigators tested the blood samples taken rom children born since January 1996 to determine the level o lead in the blood, using LeadCare Analyzers. For each village selected in the NFHS- sample, the Village Questionnaire collected inormation on the availability o various acilities in the village (especially health and education acilities) and amenities such as electricity and telephone connections. Respondents to the Village Questionnaire were also asked about development and welare programmes operating in the village. The village survey included a short, open-ended questionnaire that was administered to the village head, with questions on major problems in the village and actions that could be taken to alleviate the problems. 1.4 Survey Design and Sample Implementation Sample Size and Reporting Domains The NFHS- sample in Maharashtra was designed to provide estimates or the state as a whole, or urban and rural areas, and or Mumbai. The survey was also designed to provide separate estimates or slum and non-slum areas o Mumbai. The sample is not large enough to provide reliable estimates or individual districts. A target sample size o 4,000 completed interviews with eligible women was initially divided between urban and rural areas by allocating the sample proportionally to the population o these two areas. The NFHS-1 nonresponse rates at the household and individual levels were used to estimate the sample size that would be required to achieve the target number o completed interviews in NFHS-. The sampling rates used in urban and rural areas take rates o nonresponse into account based on urban and rural nonresponse rates rom NFHS-1. In order to provide separate estimates or Mumbai and slum and non-slum areas o Mumbai, a higher sampling rate was used in Mumbai than in other urban areas. The target sample size or Mumbai was set to be,000 completed interviews with eligible women, eectively raising the overall target sample size or Maharashtra to 5,500 eligible women. Sample Design There were three sampling domains: rural areas, urban areas excluding Mumbai, and Mumbai. Within each o the sampling domains, a systematic, multi-stage stratiied sampling design was 5

6 used. The rural sample was selected in two stages: the selection o Primary Sampling Units (PSUs), which are villages or groups o villages (in the case o small linked villages), with probability proportional to size (PPS) in the irst stage, ollowed by the selection o households using systematic sampling within each selected PSU in the second stage. In the two urban domains (Mumbai and urban areas excluding Mumbai), a three-stage sampling procedure was ollowed. In the irst stage, wards were selected with PPS. From each selected ward, one census enumeration block (CEB) was selected with PPS in the second stage, ollowed by selection o households using systematic sampling within each selected CEB in the third stage. Sample Selection in Rural Areas In rural areas, the 1991 Census list o villages served as the sampling rame. The list was stratiied by a number o variables. The irst level o stratiication was geographic, with districts classiied into six contiguous regions. The district composition o the six geographic regions (based on the 9 districts in Maharashtra at the time o the 1991 Census) is as ollows: Region I: Region II: Region III: Region IV: Region V: Region VI: Thane, Raigad, Ratnagiri, Sindhudurg Nasik, Dhule, Jalgaon Ahmednagar, Pune, Satara, Sangli, Solapur, Kolhapur Aurangabad, Jalna, Parbhani, Bid, Latur, Osmanabad, Buldhana, Akola, Amaravati Yeotmal, Wardha, Nagpur, Nanded Bhandara, Chandrapur, Gadchiroli In each region, villages were urther stratiied by village size and the percentage o the population belonging to scheduled-castes or scheduled-tribes (SC/ST). Table 1.1 provides details o the sample stratiication in rural areas, along with the population o each stratum. The inal level o stratiication was implicit or all the strata, consisting o an ordering o villages within each stratum by level o emale literacy (obtained rom the 1991 Census Village Directory). From the list arranged in this way, villages were selected systematically with probability proportional to the 1991 Census population o the village. Sample villages larger than 500 households were segmented into to three or more segments, and two segments were selected randomly using the PPS method. Small villages with 5 49 households were linked with one or more adjoining villages to orm PSUs with a minimum o 50 households. Villages with ewer than ive households were excluded rom the sampling rame. The domain sampling raction, i.e., the probability o selecting a woman in rural Maharashtra () was computed as: n N where n number o rural women to be interviewed (ater adjusting upward to account or nonresponse and other loss), N projected rural population o eligible women in the state in May

7 Table 1.1 Sampling stratiication Sampling stratiication procedure in rural areas, Maharashtra Stratiication variables Stratum Region Village population Percent SC/ST population Population > > > 500 > > 1600 > > > > > > 17.0,69,664,8,75,4,570,467,80,885,8,940,0,159,747,95,,065,171,71,59,01,151,095,849,790,660,791,694,8,819 Total NA NA NA 48,88,146 Note: The level o emale literacy is used or implicit stratiication. SC: Scheduled caste; ST: Scheduled tribe NA: Not applicable : Not used or stratiication 1 The population shown is the 1991 Census population, excluding persons living in villages with ewer than ive households. The probability o selecting a PSU rom rural Maharashtra ( 1 ) was computed as: a si 1 si where a number o rural PSUs selected rom the state, s i population size o the i th PSU, Σs i total rural population o the state. A mapping and household listing operation carried out in each sample area provided the necessary rame or selecting households at the second stage. The household listing operation involved preparing up-to-date notional and layout sketch maps o each selected PSU, assigning numbers to structures, recording addresses o these structures, identiying residential structures, and listing the names o heads o all the households in residential structures in the selected PSUs. Household listing in segmented PSUs was carried out only in the selected segments. The work was carried out by our teams, each comprising one lister and one mapper, under the supervision o one ield supervisor and one ield executive. The teams were trained rom 6 January 1999 in Pune by an oicial rom CORT, Vadodara, who was earlier trained in a workshop conducted by IIPS. The mapping and household listing operation was carried out between 8 January 1999 and 10 April The households to be interviewed were selected with equal probability rom the household list in each selected enumeration area using systematic sampling. 7

8 The probability o selecting a household rom a selected rural PSU ( ) was computed as: 1 On average, 0 households were initially targeted or selection in each selected enumeration area. To avoid extreme variations in workload, minimum and maximum limits were put on the number o households that could be selected rom any area, at 15 and 45, respectively. All the selected households were visited during the main survey, and no replacement was allowed i a selected household was absent during data collection. Sample Selection in Urban Areas Urban Areas Excluding Mumbai The 1991 Census list o urban wards was arranged according to districts and within districts by the level o emale literacy, and a sample o wards was selected systematically with probability proportional to population size. Next, one census enumeration block (CEB), consisting o approximately households, was selected rom each selected ward using the PPS method. As in rural areas, a household listing operation was carried out in the selected CEBs and, on average, 0 households per block were targeted or selection. Mumbai For Mumbai, a slightly dierent sampling procedure was used. The 1991 Census list o sections was divided into two components one consisting o sel-selected sections and the other consisting o the remaining sections. The number o PSUs was allocated to each component according to its share o the population. For sel-selected sections, the number o PSUs (i.e., CEBs) in each section was allocated according to the share o section s population in the total population o sel-selected sections. The allocated number o CEBs were then selected rom each sel-selected section using the PPS method. For the component consisting o the remaining sections, the list was irst arranged by the level o emale literacy and a sample o sections was selected systematically with probability proportional to size. Next, two CEBs were selected rom each selected section using the PPS method. As in rural areas and in urban areas excluding Mumbai, a household listing operation was carried out in the selected CEBs in both sel-selected and non-sel-selected sections and, on average, 0 households per block were targeted or selection. The domain sampling raction, i.e., the probability o selecting a woman rom an urban domain (Mumbai or other urban areas) in Maharashtra () was computed as: n N where n number o women to be interviewed rom the urban domain (ater adjusting upward to account or nonresponse and other loss), N projected population o eligible women in the urban domain in May

9 The probability o selecting a ward/section rom an urban domain ( 1 ) was computed as: 1 a si si where a number o wards/sections selected rom the urban domain, s i population size o the i th ward/section, Σs i total population o the urban domain. Note that the probability o selecting a ward/section ( 1 ) was set to one or sel-selected wards/sections. The probability o selecting a CEB rom a selected ward/section ( ) was computed as: k Bi where k the number o CEBs to be selected rom the ward/section, B i population size o the i th block, ΣB i total population o the ward/section. The household listing operation provided the necessary rame or selecting households in the third stage o sample selection. The probability o selecting a household rom a selected block ( ) was computed as: Bi 1 Sample Weights In Maharashtra, the sample is weighted at the level o the sampling domain. Sample weights or households and women are based on design weights, to adjust or the eect o dierential nonresponse in dierent geographical areas. The method o calculating the weights is speciied below. Let R Hi and R Wi be the response rates or households and eligible women, respectively. Then the household weight (w Hi ) is calculated as ollows: w Hi w R Di Hi where w Di the design weight or the i th domain, calculated as the ratio o the overall sampling raction (F n/n) and the sampling raction or the i th domain ( n i /N i ). Note that n Σn i and N ΣN i. 9

10 An eligible woman s weight (w Wi ) is calculated as ollows: w Wi R w Hi Di R Wi Ater adjustment or nonresponse, the weights are normalized so that the total number o weighted cases is equal to the total number o unweighted cases. The inal weights or households and eligible women are: n i WHi w Hi ni i W Wi wwi ni where n i reers to the actual number o cases (households or eligible women) interviewed in the i th domain. For the tabulations on anaemia and height/weight o women and children, two separate sets o weights were calculated using a similar procedure. In this case, however, the response rates or anaemia (or both women and children) are based on the percentage o eligible women whose haemoglobin level was measured, and the response rates or height/weight (or both women and children) are based on the percentage o eligible women whose height or weight was measured. Sample Implementation A total o 18 PSUs were selected, o which 16 were urban and 8 were rural. O the urban PSUs, 101 were selected rom Mumbai. Table 1. shows response rates or households and individuals and reasons or nonresponse. Nonresponse can occur at the stage o the household interview or at the stage o the woman s interview. The last row o the table shows the overall eect o nonresponse at the two stages. The survey achieved an overall response rate o 9 percent. As expected, the overall response rate is slightly lower in urban areas (91 percent) than in rural areas (9 percent). The overall response rate is considerably higher in slum areas o Mumbai than in non-slum areas or in other parts o Maharashtra. O the 6,9 households selected in Maharashtra, interviews were completed in 91 percent o the cases, 5 percent o the selected households were absent or an extended period, in percent o the selected households either no member or no competent respondent was at home when the household was visited, 1 percent o households were ound to be vacant, and 1 percent reused to be interviewed. The household response rate the number o households interviewed per 100 occupied households was 98 percent in both urban and rural areas, as well as in Mumbai. However, in slum areas o Mumbai the household response rate was 100 percent. In the interviewed households, 5,79 women were identiied as eligible or the individual interview. Interviews were successully completed with 94 percent o eligible women. The response rate or women was slightly lower in urban areas (94 percent) than in rural areas (95 n w w Hi Wi 10

11 Table 1. Sample results Sample results or households and ever-married women age by residence, Maharashtra, 1999 Mumbai Urban Rural Total Slum Non-slum Total Result Number Percent Number Percent Number Percent Number Percent Number Percent Number Percent Households selected Households completed (C) Households with no household member at home or no competent respondent at home at the time o interview (HP) Households absent or extended period (HA) Households postponed (P) Households reused (R) Dwelling vacant/address not a dwelling (DV) Dwelling destroyed (DD) Dwelling not ound (DNF) Other (O) Households occupied Households interviewed Households not interviewed Household response rate (HRR) 1 Eligible women Women interviewed (EWC) Women not at home (EWNH) Women postponed (EWP) Women reused (EWR) Women partly interviewed (EWPC) Other (EWO) Eligible women s response rate (EWRR) Overall response rate (ORR) 4, , , , , , , , , , , , , , , , , , , , , , , , NA 97.5 NA 97.8 NA 97.6 NA NA 94.7 NA 97.5, , , , , , , , , , NA 9.5 NA 95.0 NA 94.1 NA 95.5 NA 91. NA 9.7 NA 91.1 NA 9.9 NA 91.8 NA 95.5 NA 86.4 NA 91. Note: Eligible women are deined as ever-married women age who stayed in the household the night beore the interview (including both usual residents and visitors). This table is based on the unweighted sample; all other tables are based on the weighted sample unless otherwise speciied. NA: Not applicable 1 Using the number o households alling into speciic response categories, the household response rate (HRR) is calculated as: C HRR 100 C+ HP + P+ R + DNF Using the number o eligible women alling into speciic response categories, the eligible women response rate (EWRR) is calculated as: EWC EWRR 100 EWC+ EWNH+ EWP+ EWR + EWPC + EWO The overall response rate (ORR) is calculated as: HRR EWRR ORR 100

12 percent), and lower in non-slum areas o Mumbai than in slum areas. Nonresponse at the individual level was primarily due to eligible women not being at home. Very ew eligible women reused to be interviewed (1 percent). The inal sample or Maharashtra consisted o 5,80 successully interviewed households and 5,91 ever-married women age In Mumbai, interviews were completed with a total o,45 households and,010 eligible women, o which 1,19 households and 1,177 eligible women were rom slum areas. 1.5 Recruitment, Training, and Fieldwork Field sta or the main survey were trained in Pune by the oicials o CORT, who were trained earlier in a Training o Trainers Workshop conducted by IIPS. Training in Maharashtra consisted o classroom training, general lectures, and demonstration and practice interviews, as well as ield practice and supplementary training or ield editors and supervisors. Health investigators attached to interviewing teams were given additional specialized training on measuring height and weight and testing or anaemia in a centralized training programme conducted by IIPS in collaboration with the All India Institute o Medical Sciences (AIIMS), New Delhi. This specialized training took place at IIPS. It included classroom training and extensive ield practice in schools, anganwadis, and communities. Additional training on lead testing was conducted at IIPS by the U.S. Centers or Disease Control and Prevention, IIPS, and AIIMS. Seven interviewing teams conducted the main ieldwork, each team consisting o one ield supervisor, one emale ield editor, our emale interviewers, and one health investigator. The ieldwork was carried out between March 1999 and 0 June Coordinators and senior sta o CORT monitored and supervised the data collection operations. IIPS also deputed one research oicer to help with monitoring throughout the training and ieldwork period in order to ensure that correct survey procedures were ollowed and data quality was maintained. From time to time, project coordinators, senior research oicers, and other aculty members rom IIPS, as well as sta members rom ORC Macro and the East-West Center, visited the ield sites to monitor the data collection operation. Medical health coordinators appointed by IIPS monitored the nutritional component o the survey. Field data were quickly entered into microcomputers, and ield-check tables were produced to identiy certain types o errors that might have occurred in eliciting inormation and illing out questionnaires. Inormation rom the ield-check tables was ed back to the interviewing teams and their supervisors so that they could improve their perormance. 1.6 Data Processing Completed questionnaires were sent to the CORT oice in Vadodara or data processing, which consisted o oice editing, coding, data entry, and machine editing, using the Integrated System or Survey Analysis (ISSA) sotware. Data entry was done by ive data entry operators under the supervision o senior sta at CORT who were trained at a data-processing workshop in Vadodara. Data entry and editing operations were completed by August Tabulations or the preliminary report as well as or the present inal report were carried out at IIPS in Mumbai. 1