World Values Survey 2007, Wave 5

Probability sampling, ensuring national representivity Households and respondents selected using the Kish Grid Sampling is the process of selecting certain members of a group in such a way that they will represent the universe. Selection of respondents for the project followed a stratified Multi- Stage Random Selection Procedure as follows:

Selection of Sectors/EAs: Sectors are defined as sampling blocks of equal geographical dimensions with identifiable boundaries, encompassing a substantial number of people. Sectors were divided into high, medium and low density areas. Each of the sectors was thereafter randomly selected from each area using the available street maps already sectorised into different density areas. Where maps are not available, especially for rural areas, an exhaustive list of all sectors was considered. The sample allocated to each density area i.e. high/medium and low was proportionate to the number of sectors in each group. The overall sample for the urban and rural locations determined the number of sectors selected. However, a maximum of five (05) interviews were conducted in each randomly selected sector. All sectors were selected by simple random method via a random numbered table. Group interviewing techniques were adopted for the study across all the study locations. By this design, a team of interviewers under the leadership of a supervisor moved as a group to each selected sector, and then completed the assigned quota for that sector before moving to another sector. This afforded the supervisors the opportunity to closely monitor the interviewers under them. The questionnaire was precoded using the alphabet letters A to K excluding letter I.

Selection of Sampling/Entering Points within each sector: Immediately after the selection of the sectors, the supervisors surveyed each of the selected sectors to determine the sampling/entering points of the sector. These are points where the team started their days interviewing. These can be prominent structures such as churches, mosques, schools, hospitals, etc.

Selection of Dwelling Structure within each sector: In each of the randomly selected sectors, the Days Code was used to determine each interviewers starting point, i.e. [The first house/dwelling structure to enter/approach]. A dwelling structure is defined as a floor of a distinct residential building within a sector of a town/village; where only one household occupied a multistorey building, the entire building [and not the floor] constituted a dwelling structure. Where it is a multi-storey building with multiple occupants, counting of floors was carried out consistently from the upper floor to the ground floor in an unbroken chain from floor to floor. A fixed sampling gap of one in three (1:3) and one in five (1:5) respectively was observed after each successful call in low, medium and high density areas.

Selection of Household: On entering a selected dwelling structure, each interviewer determined the number of households within the structure. Having done that, the interviewer then used the household selection grid to determine the household where the interview would take place. A household is defined as the collective individuals living under the same roof and having a common feeding arrangement and also with a recognised person in the household as the head of household. Only residents who have stayed in the selected household for at least six [6] months were interviewed. Visiting relations who have stayed for less than six months were not regarded as household members. Substitution of Households: In the case where the selected room was unoccupied, interviewers were instructed to replace with the next household. Only one substitution was allowed per dwelling structure.

Selection of Respondents: Respondents were randomly selected among the male and female household members. In order to select the final person to interview within the selected household, all the male and female residents of Mali, aged 16 years and above in the selected household were listed by names and ages on the respondents selection grid on the questionnaires. The listing was done from the eldest to the youngest (males and females combined) and then one respondent was selected using the Kish grid a table of randomly generated numbers.

Call Backs/Substitution Criteria: In the case where the selected adult in the household was not available at the time of the call, interviewers were instructed to make up to two additional recalls on different times of the day including evenings when the selected respondent was said to be at home. However, where the selected adult was not available for interviewing within the days of selection, interviewers were asked to regard such a case as a non-response situation or ineffective call. No substitution of respondents within the same household/dwelling structure was allowed.

Substitution of Households: In the case where the selected room was unoccupied, interviewers were instructed to replace with the next household. Only one substitution was allowed per dwelling structure.

Call Backs/Substitution Criteria: In the case where the selected adult in the household was not available at the time of the call, interviewers were instructed to make up to two additional recalls on different times of the day including evenings when the selected respondent was said to be at home. However, where the selected adult was not available for interviewing within the days of selection, interviewers were asked to regard such a case as a non-response situation or ineffective call. No substitution of respondents within the same household/dwelling structure was allowed.

The sample size for Ghana is N=1534 and includes national population aged 16 years and over for both sexes.