representative samples

Determination of a 'Representative Sample'

%% {[NOT YET INTEGRATED]} %%

The total number of villages in MTB in the year 2010 is being used as a proxy measure to establish a representative sample within the project's 'social area of impact'. The client acknowledges that it will be a project of regional significance, and definition of 'area of influence' will continue to be reformulated and refined during the process of social impact assessment. The significance of this measurement is to establish an appropriate survey size.

FIGURE: 1 %%{[ADJUST TABLE]}%%

Population Size	Confidence Level	Degree of Accuracy	Target Sample Size

MTB total # villages* 72 45% 0.05 28

To help ensure that the survey team interacts directly with a 'representative sample' of the regional area of influence, Consultancy grouped villages according to arithmetic quartiles by population and sought to balance site selection accordingly.

FIGURE: 2 %%{[ADJUST TABLE]}%%

Calculation of MTB quartiles

MTB kabupaten population, mean: 105,341/71 = 1483.676 [1484]

No. MTB villages above mean = 28

// re-factored mean for top half = 71,232=/28 [2,544]

• top [\geq 2,544] = 5 villages
• mid-top [2,544 - 1,484] = 23 villages

No. MTB villages below mean = 43

// re-factored mean for top half (mean re-factored) = 34,809/43 [793]

• mid-bottom [1,483 - 793] = 18 villages
• bottom = [\leq 792] = 26 villages

This methodology extends the number of study villages to achieve a more balanced and representative sample, as well as to better demonstrate how the project is attending to the IFC prescript to ensure that the impact assessment process accounts especially for 'vulnerable communities' (i.e., by giving suitable weighting to the representation of small and isolated communities that may be exposed to project-induced indirect impacts, such as price inflation).

{[ADJUST STATEMENT]} As calculated in the table above and reflected in the graphics below, the distribution of ESMS social baseline study locations achieves a rough approximation of regional composition with respect to population size.

FIGURE: 2 {[INSERT EXCEL IMAGE]}

Primary baseline data collection in each village sets numerical targets for the number of survey participants according to confidence values of 80% and margin of error calculations varying between 0.25 and 0.5. Consultancy will use a combination of household questionnaire surveys, key informant interviews, and focus group discussions to effect a rapid rural appraisal that reaches representative sample population targets //%%{[SECTION]}%%//. The target number of participants for FGD sessions is calculated according to MTB 2010 village population quartile values. ESMS supplementary social baseline implementation schedules were created by calculating village days according to the estimated number of FGDs that can be completed in one day (6 total; refer to //%%{[SECTION]}%%//).

ANNEX 1: %%{INSERT: [Sample Definition]}%%

Selection of Study Sites

Consultancy and client agreed to the following parameters for supplementary baseline site selection:

• sample size should seek to achieve a defensible statistical representation of the social field; no less than 35% confidence level for regional "area of influence" (minimum of 17 villages)
• full representation of directly affected communities [equivalent to 'priority directly-affected communities' (PDAC)in the terminology of the client Public Consultation and Disclosure Plan (PCDP)]
• partial representation (ca. 40% - 80%) of areas likely to be exposed to project-induced direct and indirect socio-economic impacts ('directly-affected communities' DAC)
• partial representation (ca. 10% - 40%) of areas that may indirectly be affected by project-induced socio-economic change, or those area where there is a vocal stakeholder perception or belief that such changes may occur ('indirectly-affected communities' IAC)

Additional parameters influencing Consultancy' proposed field sites selection include:

• proximity and distance from key project activity zones, especially the administrative and economic center of Saumlaki, the capital of MTB, as well as the planned on-shore LSB site and port facilities in neighboring Olilit Village.
• distances from port facilities / accessibility by regular forms transportation [proximity to transportation hubs affects the likelihood a village will experience in-migration]
• distance from healthcare and treatment facilities [availability of, and access to, health and educational services affect the extent to which communities' vulnerability may be impacted by the project]
• available production land in village [derived from BPN statistics, this measurement indicates the potential of a village community to expand agricultural output in response to added population pressures]
• due consideration of existing social, demographic, linguistic, geographical and economic data [MTB has several distinctive language groups, the number of which is a matter of scholarly debate; at least five major groups can be distinguished]
• due consideration of timing, capacity constraints and the logistics of access

Consultancy' proposed selection criteria seeks to extend the geographic coverage of study sites to better account for potential project-induced changes in northern parts of MTB and other more distant villages--such as those that may be affected by speculative flows of immigrants or particularly remote and impoverished villages that are relatively more vulnerable to shocks of regional economic change. We recognize, however, that much of the west coast of MTB's Yamdena Island is not accessible by road, and a number of populated islands in the northwest and northeast of Yamdena cannot be readily reached by means of sea transport meeting client' quality, health and safety standards. Consultancy will work with client to revise and finalize final field site selection.

Following the above parameters and considerations, the Consultancy team recommends selection of the below villages for village-level collection of supplementary baseline data:

ID# Village Kecamatan

81-01-43-006 Adaut Selaru 81-01-54-005 Alusi Kelaan Kormomolin 81-01-42-003 Batu Putih Wer Maktian 81-01-40-017 Bomaki Tanimbar Selatan 81-01-43-004 Kandar Selaru 81-01-50-029 Kelaan Tanimbar Utara 81-01-50-026 Keliober Tanimbar Utara 81-01-40-002 Latdalam Tanimbar Selatan 81-01-50-022 Lelingluan Tanimbar Utara 81-01-40-001 Lermatang Tanimbar Selatan 81-01-43-003 Lingat Selaru 81-01-41-001 Tumbur Wer Tamrian 81-01-54-001 Lorwembun Kormomolin 81-01-40-016 Matakus Tanimbar Selatan 81-01-54-007 Mayano Raya Kormomolin 81-01-40-010 Olilit Tanimbar Selatan 81-01-50-024 Ridool Tanimbar Utara 81-01-50-023 Ritabel Tanimbar Utara 81-01-50-002 Rumngeur Yaru 81-01-41-006 Sangliat Dol Wer Tamrian 81-01-40-009 Saumlaki Tanimbar Selatan 81-01-40-009 Sifnana Tanimbar Selatan 81-01-52-009 Tutukembong Wuar Labobar 81-01-50-025 Watidal Tanimbar Utara 81-01-53-001 Waturu Nirunmas 81-01-43-001 Werain Selaru 81-01-42-004 Wermatang Wer Maktian 81-01-40-015 Wowonda Tanimbar Selatan

For a detailed explication of rationale for the selection of each village, as well as a list of target stakeholders for FGD questions per village, please refer to ANNEX 2.

ANNEX 2: %%{INSERT: [Site Selection and Lines of Focused Inquiry]}%%