Philippines - Annual Survey of Philippine Business and Industry (ASPBI) 2015, 44th round

The primary consideration in the determination of sample size for the survey was its manageability at the optimum level of estimated budget without compromising the reliability and accuracy of survey results and the timeliness of retrieval of accomplished questionnaires from respondents.

Sample sizes for each of the strata were determined separately by employment stratum to achieve a minimum precision of 4% margin of error for a 90% confidence interval on employment estimates. Then, the sample size for industry stratum was determined by applying specified sampling rates and allocation rule for each employment stratum. Total sample size was obtained by aggregating separately each estimate at the industry stratum.

For sampling purposes, the survey frame was divided into the five primary strata. Stratum 1 comprises of Mining and Quarrying (B) and Electricity, Gas, Steam and Air Conditioning Supply (D); Stratum 2 consists of certainty industries including the Information and Communication Technology (ICT) core industries and Business Process Management (BPM) industries; and tourism characteristic industries; Stratum 3 for GOCCs; Stratum 4 is composed of establishments with TE 100 and over; and all other establishments not classified in strata 1-4 were grouped as Stratum 5 and were treated as Non-Certainty Stratum. Strata 1-4 were considered as certainty stratum.

The Stratum Code, Sampling Units, Sampling Indicator are as follows:

1 - Establishments under section B and D - Certainty
2 - Establishments under the certainty industries ( SICT Core Industries, BPM industries and Tourism characteristic industries) - Certainty
3 - GOCCs - Certainty
4 - Establishments with TE of 100 and over - Certainty
5 - All other establishments not classified in strata 1-4 - Non-Certainty

A certainty stratum is defined as the employment stratum whose sampling ratio is 100 percent. In this stratum, all establishments are taken as certainty samples, i.e. the selection probability is 1 and the sampling weight is 1. On the other hand, a non-certainty stratum is the employment stratum where only sample establishments are taken.

The corresponding sampling rates by employment stratum for establishments in the non-certainty employment size strata are as follows:

1 - 9 (10 %)
10 - 19 (15 %)
20 - 49 (25 %)
50 -99 (50 %)


For TE of less than 20

The sample size for establishments with TE of less than 20 was determined by applying the specific sampling rates for each employment stratum (that is, separately for strata TE 1 - 9 and TE 10 - 19) of the same industry stratum (5-digit PSIC) at the national level.

No allocation was done in these strata as the sample sizes were determined at the 5-diigit PSIC (industry domain) and national level.

Minimum sample size is set to 3 establishments and maximum of 5 establishments per cell (industry domain and employment stratum). However, when the total number of establishments (N) in the cell is equal to or less than the set minimum sample size, all establishments in that cell are taken as samples.

The estimated sample size for establishments for these employment strata is about 10,923.


For TE of 20 and over

The sample size for establishments with TE of 20 and over was obtained by applying the specific sampling rates for the ith employment stratum in the hth industry stratum at the national level and then allocating to geographic domain (region).

Proportional allocation was used in allocating the number of samples in the sampling strata of TE 20 - 49 and TE 50 - 99 of the industry domain at the regional level. Basis of allocation is the total number of establishments (N) in the region by industry domain and by employment stratum. The formula used was,

nhij = (Nhij/Nhi) x nhi
where:
Nhij = total number of establishments in the jth region of the ith employment stratum in the hth industry domain
Nhi = total number of establishments in the ith employment stratum of the hth industry domain
nhij = number of sample establishments in the jth region of the ith employment stratum in the hth industry domain
nhi = Number of sample establishments in the ith employment stratum of the hth industry domain

The sample sizes were likewise controlled to a minimum of 3 establishments and maximum of 5 establishments per cell (industry domain, employment stratum in the region). However, when the total number of establishments (N) in the cell is equal to or less than the set minimum sample size, all establishments in that cell are taken as samples.

The aggregate sample size for TE of 20 and over in all sectors is about 19,691.

Thus, the total number of samples for the 2015 ASPBI is about 30,614.

For all sectors, sample establishments in the sampling strata of TE of less than 20 are selected using systematic sampling by industry domain and employment stratum at the national level. For each industry domain and employment stratum, the establishments are sorted by region, province from largest actual employment to smallest actual employment, business name and ECN.

For each of the sampling strata of TE of 20 and over (i.e. TE 20-49 and TE 50-99) sample establishments are selected using systematic sampling within the region. For each region in the employment stratum and industry domain, the establishments are sorted by province from largest actual employment to smallest actual employment, business name and ECN.

Systematic sampling is chosen so that the sample employment values are spread out, resulting from having representative samples for each TE size in the employment stratum. Likewise, this mode of sampling will somehow provide implicit stratification of TE by employment size group, thus avoiding all sample establishments with low TE values or high TE values.


Field operations of the 2015 ASPBI was scheduled from April to July 2016.

Distribution was undertaken for a period of one (1) month starting April 13, 2016 to May 20, 2016.
Collection of questionnaires was from May 2, 2016 to July 15, 2016.

The total number of sample establishments for the survey operations was 30,552 of which 28,715 or 94.0 percent responded.

Total response rate include receipts of "good" questionnaires, partially accomplished questionnaires, reports of closed, moved out or out of scope establishments.


For certainty stratum, the weight is equal to 1.

For non-certainty,

Weight Adjustment Factor for Non-Response

To account for non-response in the non-certainty strata, the adjustment factor (n/n') was multiplied with the sampling weight (W) of each of the sampling unit. The sampling weight, defined as N/n, was recomputed as

W' = N/n*n/n' = N/n'

Thus, the adjusted weight (W'sj) for employment stratum in TE of 1-9 or TE of 10-19 was

W'sj = Ns/ns * ns/n's = Ns/n's

where:
Ns = total number of establishments in the employment stratum in TE 1-9 or TE 10-19 in the sth industry domain
n's = number of responding establishments in the employment stratum in TE 1-9 or TE 10-19 in the sth industry domain

For the non-certainty employment stratum for the industry domain with TE 20-99, the adjusted weight (W'spj) was

W'spj = Nsp/nsp * nsp/n'sp = Ns/n'sp

where:
Nsp = total number of establishments in the non-certainty employment stratum with TE 20-99 for the industry domain within each geographic domain (region)
n'sp = number of responding establishments in the non-certainty employment stratum with TE 20-99 for the industry domain within each geographic domain (region)