India - Survey on Level of Living of TRIBALS July-June 1988-89, NSS 44th Round

The sample design is, as usual, stratified two-stage with the census village as the first stage unit in the rural sector and UFS block as the first stage unit in the urban sector. The second stage units were households for all schedules.The sample design in the rural sector was decided with a view to providing good estimates for the tribal enquiry. Except in the north-eastern region, the tribal population was concentrated in some districts within the states having considerable tribal population and even in those districts they were found to be unevenly distributed geographically. Therefore special stratification and selection procedures were adopted not only to net sufficient number of tribal households in the sample but also to improve the design in general for the tribal enquiry.

Sampling frame of villages: The list of 1981 census villages constitute the sampling frame for selection of villages in most districts. However in Assam (where '81 census was not done) and a few districts of some other states (where the available lists of villages were not satisfactory), 1971 census village lists were used as frame.

Stratification : In Haryana, Jammu & Kashmir, Punjab, Chandigarh, Delhi, Goa, Daman & Diu and Pondicherry where there were practically no tribal population, the strata used in NSS 43rd round were retained. In Meghalaya, Mizoram, Nagaland, Arunachal Pradesh, Sikkim, Dadra and Nagar Haveli and Lakshadweep also the strata of 43rd round were retained because of the high percentage of ST. population in these States/U.T.'s. (The strata of 43rd round have been retained in the case of Sikkim as the distribution of tribal population is more or less uniform over all the districts).In the remaining states fresh stratification was carried out as described below.

In these states all districts accounting for the bulk of the states's tribal population were selected for formation of strata with concentration of tribal population. Besides these districts, tribal concentration strata were demarcated also in some other districts with relatively small tribal population in order to ensure coverage of as many different ethnic groups as possible.

Within each district so identified for formation of tribal concentration strata, the tehsils with relatively high concentration of tribal population, together constituted one stratum. These tehsils were selectd in such a way that together they accounted for the bulk (70% or more) of the district tribal population and the proportion of tribal to total population and the proportion of tribal to total population in this stratum was significantly greater than that of the district as a whole. The strata so formed were not always geographically contiguous. These tribal concentration strata are called STRATUM TYPE -1. Further, all the strata of Meghalaya, Mizoram, Nagaland, Arunachal Pradesh, Dadra & Nagar Haveli, Lakshdweep and Sikkim are also considered as stratum type-1. All the remaining strata in the rural sector (in any State/U.T.) were called stratum type -2.

General and special sample villages : There were two types of sample villages in this round. The first type was the general sample in which all enquiries were carried out. The second type was designated as "special sample villages" in which only schedules 3.1, 29.1, 29.2 and 29.3 were canvassed. The special sampleswere intended for augmenting the general sample for the tribal enquiry. These special sample villages were selected only from the tribal concentration strata (stratum type 1 ) of the 16 States and the U.T. OF Andaman & Nicobar Islands 1) of 16 States and the U.T. of Andaman & Nicobar Islands. There were no special sample villages in the remaining States/U.T.'s. The special samples were called sample type-1 and the general samples, sample type-2.

Schedules 29.1 was canvassed both in rural and urban sectors. Six households each from the frame of schedules tribe households in the villages of special as well as general samples falling in the tribal strata were canvassed. On the other hand only two households were selected from the sample villages of stratus type 2 as well as from each sample block in urban areas of the schedules 29.1. In the special sample villages requiring hamlet group formation, the hamlet group having maximum number of tribal population will be designated as area type 1 and the other hamlet group selected at random from the remaining hamlet groups will be called area type 2. From the area type 1, 4 tribal households will be selected, while only 2 tribal households will be sampled from area type 2 for he schedule - 29.1.

Detailed procedures of samp[ling may be seen in INSTRUCTIONS TO FIELD STAFF : VOLUME I attached as external resource.


There was no deviation from the original sample deviation.


Sample weights were NOT included in each of the data files