Sampling Procedure
Sampling Frame:
The frame of the survey was developed basically from the list of the National Economic Census (NEC) 2018. The list of establishments from NEC 2018 contains the complete name and address of establishments along with other variables such as annual income, annual expenditure, number of persons engaged etc.
Before deciding about sampling methodology, the nature and characteristics of NEC 2018 frame was reviewed and considered it as a main frame. Beside main frame, the lists of currently running micro hydro-powers and waste collection/disposal industries were also updated to the main-frame. Those lists were obtained from AEPC and Federation of Waste Collection and Disposal Industries.
Stratification:
Separate survey scheme was designed for each section of NSIC covered in the survey. Industries in section B, D and E of NSIC were completely enumerated, in case of section C of NSIC both complete enumeration and sampling method were adopted based on the following criteria.
· For large manufacturing establishments (if number of persons engaged was more than or equal to 10), all establishments were enumerated.
· For small manufacturing establishments (if number of persons engaged was less or equal to 9), sampling method was adopted. However, if number of establishments in each province at 4 digit NSIC level was less than 5, then such establishments were completely enumerated to ensure their representation in the survey.
For small size establishments of section C, three categories were made based on number of persons engaged i.e. 1-2 persons engaged, 3-5 persons engaged and 6-9 persons engaged.
Using these categories, strata were made based on province, NSIC and persons engaged categories. The total number of establishments in the sampling frame for small size establishments of manufacturing section by province is presented in Table A.
Table A: Total number of establishments in the sampling frame by province and persons engaged category (for small size establishments of section C)
Persons engaged category
Province 1-2 3-5 6-9 Total
persons persons persons
Province no. 1 5,970 3,780 1,070 10,820
Madhesh 3,225 2,642 565 6,432
Bagamati 7,349 6,473 1,980 15,802
Gandaki 3,041 2,305 629 5,975
Lumbini 4,480 3,666 982 9,128
Karnali 810 536 107 1,453
Sudurpaschim 1,548 1,134 254 2,936
Total 26,423 20,536 5,587 52,546
Source: National Economic Census, 2018
Analytical Domain:
The survey was designed to provide the national estimates of the NSIC 4 digits classification and provincial estimates of the NSIC 3 digits classification.
Sample Size and Sampling Method:
The total enumeration size (census and sampling) of the survey was planned to enumerate 14,000 establishments. Out of them, about 9,200 establishments were expected to be enumerated in census category for the NSIC sections B, D, E and large size manufacturing establishments of section C.
For small size manufacturing establishments, a total of 4,738 establishments were fixed as a sample size, which were distributed to each stratum (province, NSIC, persons engaged category) using power allocation with power 0.7. The samples from each stratum were then selected by using systematic sampling method. The summary of sample allocation is presented in the table B.
Table B: Summary of the allocation of samples in provinces by persons engaged category
(for small size establishments of section C)
Province Persons engaged category
1-2 3-5 6-9 Total
persons persons persons
Province no. 1 418 340 149 907
Madhesh 251 253 97 601
Bagamati 524 527 257 1,308
Gandaki 253 241 98 592
Lumbini 333 322 140 795
Karnali 93 77 27 197
Sudurpaschim 152 135 51 338
Total 2,024 1,895 819 4,738