1.5 Research Methodology
1.5.4 Sampling procedures
Multi-stage cluster sampling techniques were utilized to select the respondent for this study. The total geographical area was divided into 7 provinces and 3 provinces were taken for this assessment. However, this is not achievable as a natural cluster may vary considerably in size.
The technique/ stages involve the procedures below;
(a) Zoning Maldivian geography to three zones – North, Central and South.
(Column 1, Table 1.2)
(b) In three zone have 7 provinces as shown in Column 2, Table 1.2 and each province consists of 3-4 atolls (column 3, Table 1.2).
(c) From 7 provinces, three provinces were randomly selected for this study. That is upper north, central and upper south province (Column 4, Table 1.2). The upper north and upper south were most densely populated and situated at the two ends of the country, which makes most difficult to seek tertiary health care. Central province is situated at the capital province, which can excess easily to tertiary health care.
(d) From the selected provinces, atolls are randomly selected based on health care delivery system (Column 5, Table 1.2)
(e) From the selected atolls, health care was selected based on the levels of service (Regional Hospitals, Atoll Hospitals, Health Centers and Health Post/
Family Sections (Column 6, Table 1.2
(f) From the selected island total population was identified on each 3 selected atolls.
(Column 3, Table 1.3)
(g) Sample was determined by using Krejcie and Morgan (1970) table for determining sample size from a given population (Appendix A) (Column 4, Table 1.3)
Table 1.2 Multi-stage cluster sampling
(1) (President’s Office) (3) (4) (5) (6)
Zone Province Atolls Selected province Selected atolls Selected health care facilities
1. North Upper North
Alif Dhaalu Atoll Mahibadhoo (Atoll Hospital)
Maamigilli (Health Center)
Dhidhoo (Health Post)
3. South South (Mathi Dekunu Province)
Uppe South (Dhekunu Province)
Gaafu Atoll Gaafu Dhaalu Atoll
Gnaviyani Atoll Seenu Atoll.
South
(Mathi Dekunu
Province)
Gaafu Dhaalu Atoll Thinadhoo (Regional Hospital)
Madaveli (Health Center)
Rathafandhoo (Health Post)
Table 1.3 Sample size determinations
(1) (President’s Office) (3) (4) (5)
Selected atolls Selected islands Population/ Total Total Sample/ Atoll
(Krejcie and Morgan
Stages involve in determining samples
At the initial stage, the researcher clustered the geography area of the whole country into 3 zones (figure 2). In the second stage, the researcher identified provinces in each zone (7 provinces). At the next stage the researcher identified 3 provinces and selected island from each province to conduct survey (Column 6, Table 1.2).
The researcher selected 3 levels of health care services in the country so that she can compare the levels of participation among the population at different health care availability. As summarized in Column 4 and 5 of Table 1.2, from zone 1 selected Upper North Province and from that Haa Dhaal (HD) Atoll was selected for this study. From the HD atoll, the researcher selected 3 main islands namely Khuludhufushi Island with total 6,998 populations, Hanimadhoo Island with total 923 populations and Finey island total 430 populations.
The researcher determined sample by using Krejcie and Morgan’s (1970).
Based on this, the researcher has calculated proportionate to size for each island.
From Khuludhufushi Island sample size is 307 people, Hanimadhoo Island sample size is 41 people and Finey Island sample size is 19 people. Thus, the total sample size in HD Atoll is 367 people. In the same manner from zone 2 North Central Province was selected. From North Central province Alif Dhaal (AD) Atoll was selected for this study. In AD Atoll 3 Main Island was Mahibadhoo with total 1,780 populations, Maamigilli with total 1,671 populations and Dhidhoo with total 116 populations. The researcher calculated proportionate to size for each island, Mahibadhoo Island sample size is 172.6 people, Maamigilli Island sample size is 162 people and in Dhidhoo Island sample size is 11 people. This makes total of 346
respondents. Finally from zone 3 South provinces Gaafu Dhaalu (GD) Atoll was selected for this study. From GD Atoll, Thinadhoo Island with total 4,442 populations, Madaveli Isaland with total 1.065 populations and Rathafandhoo Island with total 492 populations. Proportionate to size was calculated and from Thinadhoo Island sample size is 267 people, Madaveli Island sample size is 64 and Rathafandhoo Island 30 people. This makes the total of 361 respondents. Therefore the total respondents for this study are 1,074 people.
(h) From selected island respondent were selected by using probability proportionate
to size (PPS) as show in Table 1.3 (column, 5). Sampling with probability proportionate to size allows the larger clusters to have a great chance of being selected.
The selection of the respondents was made by using simple random sampling technique. To select the sample according to this technique the following steps has been followed.
- Meet the province office head to explain the research objectives and work schedule to gain maximum consideration.
- Collected list of people in island along with house name to reduce work load.
Selection of the study sample was made using simple random sampling technique. To select the sample according to this technique, the following four step of sampling has been done. First step is identifying the number of adult population including name of the houses with age 18-65 years. The list was taken from island office or province office. The researcher identified among the total population particular to catchment area in the given population and inclusion criteria as follows:
- 18 - 65 years of age - Male or female participant - Residence of the community - Willing to provide answer to study Exclusion criteria:
- Respondent who cannot speak or listen - Who have mental or psychiatric problem - Not living in island or residence of that island
For example if an island has 6900 population that fit to this study, the researcher uses four-digit number for each person, beginning with 0001 until 6900. Using the Random Number Table, the researcher identified the number of respondents fit for sample from each identified island. The researcher chooses respondents from each island until total number of respondent are complete. For in-depth interview researcher collected the list of committees in the community
Figure: C Map of Maldives