Nepal - Survey of Own Account Construction (Rural) 1993, Second Round

1. Survey Frame
This survey is designed based upon the area sampling method. All wards under the administration of the Village Development Committees (VDe) constitute~ the survey frame. Besides, wards under three newly declared mun~cipalities viz . Gaur, Byas (Damauli) and Tulsipur which were not covered by the Survey of Own Account Construction (Urban) were also included in the frame. While preparing the survey frame problems related to the recent change in the boundaries and name of the VDC's were encountered. Since the base data for "the frame (numberl oE VDC's, area under VDC's and number of household by VDC's and by wards _etc.) were obtained from the results of Population Census-1991, these changes were ignored . That means if the selected ward was found affected from the changes then the area was surveyed in the boundaries at the date of Populaeion Census-1991 .

2. Stratification
Following the criteria used in the Urban Own account Construction Survey whole rural area was divided into three strata: Kathmandu Valley, Hills and Tarai. Besides, two domains (Domain I and Domain II) were formed in each stratum,. Formation of such domains became necessary in order to separate_ some semiurbanized centers (Domain I), especially the adjoining areas of several municipalities administered yet by the Village Development Committees from the rest of rural area (Domain II). It was assumed that the type, size and quality of construct·ion activities in these area significantly differ from those in the rest of rural area. According to this classificatiron three newly declared municipalities not covered by the Urban Survey automatically fall in Domain I. Separation of area for Domain I was made at the ward level.

3. Sample size and allocation scheme
The Survey of Own account Construction (Rural) designed with a single stage sampling procedure has considered a VDC ward as the sampling unit. In each stratum within the domain, wards were selected directly from the frame. All VDC's were arranged according to the geographical location of their districts in the direction from east to west and the wards were selected based on serpentine systematic sampling procedure starting from the north corner of the easternmost part. Selection of the wards in each domain was made independently. A suitable sample size based on a calculated variance could not be exercised due to lack of appropriate data. However, some auxiliary information were analyzed in finding a way out to determine the sample size . In a previous survey on economic activities conducted by CBS, around 5 % households in rural wards were found engaged in own_account construction. Similarly, on the basis of the urban part of this survey it could be assumed that the observation of around 500 ultimate sampling units (households engaged in own account construction activities) can provide estimates at the statistically acceptable level of precision. So, sample size was determined at 100 wards. Ultimately, 102 wards were selected and 633 households (identified as engaged in own account construction activities during the reference period) from those selected wards were interviewed.

For more details please refer to the Report on The Survey of Own Account Construction (Rural), Section 4 (Sampling Design).


Please refer to the Report on The Survey of Own Account Construction (Rural), Section 5 (Estimation Procedure).