Table 1 gives the intended and actual sample size for the different methods used in the study.
Table 1: Sample Size by Methods Used
| Method | Intended Sample Size | Actual Sample Size |
|---|---|---|
| Postal questionnaire | 187 | 40 |
| Consumer schedule | 3000 | 3,821 |
| Retailer schedule | 50 | 67 |
| Price survey | 50 | 73 |
For the postal questionnaire, a list of wholesalers, suppliers, hotels and restaurants from the telephone directory were drawn up. A list of all kapenta producers was added to this. All companies/organisations on this list were then sent a questionnaire together with a postage prepaid envelope for returning the completed questionnaire. Some of the companies identified no longer or never did trade in fish or fish products or had gone out of business.
Consumers were randomly selected for interviews at selected retail outlets. In Harare, Chitungwiza and Bulawayo, retail outlets were selected on the basis of high and low income areas. The study focused on low to medium income areas as these were considered more representative of the population. A total of 3,821 people were interviewed, 65 % were female and 35 % were male. The higher proportion of females in the sample is attributed to the fact that women are more likely to buy food for the household. No one below the age of 16 years old was interviewed.
In urban areas, one full day was spent by one market researcher in a retail outlet. In the smaller towns and rural areas, all main retail outlets selling foodstuffs were sampled. The intended sample size was based on the number of interviews which could be carried out in one day multiplied by the number of market researchers and number of working days feasible, taking into account budgets and logistics. The number of retail outlets and the number of consumers interviewed by location, are given in Table 2.
Table 2: Number of retail outlets included in sample by area and number of consumers interviewed
| Area | Province | No.of Retail Outlets | No.of Consumers |
| Urban: | |||
| Chitungwiza | Harare | 7 | 453 |
| Harare | Harare | 19 | 350 |
| Bulawayo | Bulawayo | 9 | 478 |
| Masvingo | Masvingo | 6 | 253 |
| Mutare | Manicaland | 3 | 308 |
| Gweru | Midlands | 5 | 408 |
| Kwekwe | Midlands | 3 | 207 |
| Kadoma | Midlands | 1 | 196 |
| Zvishavane | Mash.West. | 4 | 302 |
| Esigodini | Mat.South | 3 | 31 |
| Sub-total | 60 | 2,986 | |
| Rural: | |||
| Neshuro | Masvingo | 1 | 51 |
| Mwenezi | Masvingo | 2 | 47 |
| Murewha | Mash.East | 3 | 206 |
| Mutoko | Mash.East | 0 | 120 |
| Gwanda | Mash.East | 1 | 68 |
| Kotwa | Mat.South | 2 | 109 |
| Maphisa | Mat.South | 0 | 18 |
| Tsholotsho | Mat.North | 3 | 118 |
| Mutare Rural | Manicaland | 1 | 98 |
| Sub-total | 13 | 835 | |
| TOTAL | 73 | 3,821 |
For the retail interview schedule, managers of each of the retail outlets were interviewed. Some managers were unable to participate in the survey as they were too busy. Sixty-seven interviews were carried out.
The price survey was also carried out at the same retail outlets. Seventy-three surveys were carried out.
Personal interviews with producers, wholesalers and suppliers were aimed at obtaining information from the main players in the market, as well as a representative sample of kapenta producers on size and type of operation (cooperative or private company). Twenty-four interviews were held with wholesalers, kapenta distributors and producers.
