Page 2 9B04A022 • Current uses of custom work, • Methods used to arrange for custom work, • Level of formality of current arrangements, • Level of satisfaction with current method of buying/selling custom work, • Reasons for satisfaction or dissatisfaction, • Awareness of OMR (remember: Janzen has been advertising and attending trade shows for quite awhile), • Level of knowledge of OMR, • Reaction to the OMR concept, • Benefits seen in the OMR concept, • Potential usage problems with the OMR concept, • Likelihood of trial use, • Attitude toward the proposed pricing structure of OMR, • Demographic characteristics of users/providers of custom work, and • Demographic characteristics of likely triers versus non-triers of OMR. RESEARCH DESIGN Wilson suggested the study should be conducted using a telephone survey of 400 Ontario farmers. Although telephone surveys were more costly than mail, Wilson was very concerned about a low response rate to a mail approach. His experience indicated a response rate of over 70 per cent using the telephone compared to less than 20 per cent using mail. Wilson proposed an overall sample of 400 Ontario farmers stratified by two factors — region and farm size. It was felt stratification was important to achieve proper sample balance. Regional stratification would be done using the “value of custom work” as a control. The Ontario Ministry of Agriculture and Food (OMAF) measured the value of custom work and reported same on a county basis. The first column in the upper part of Exhibit 1 shows this value summed for each of the four regions included in the study. The second column shows the percentage of the total attributable to each region. Wilson felt that sampling efficiency would be maximized by an equal allocation of the sample to each region. His suggested allocation is shown in the third and fourth columns of Exhibit 1. The key reason for this suggestion was to ensure sufficient observations in each region to draw statistically sound conclusions. Weighting would be required in subsequent data analysis to return the sample population to the actual proportions. It was proposed that the allocation of observations based on farm size would be done on a proportional basis. The lower part of Exhibit 1 shows the actual distribution of farm size into four categories as determined by the most recent Census of Agriculture. The farm size control was used to ensure that farms in each
Page 3 9B04A022 category were included in the sample in the same proportion with which they existed in the population of Ontario farms. Exhibit 2 shows the number of sample observations that would be selected randomly from each region and farm size combination. It was proposed that the sample be drawn from the subscription list of Country Guide. This list was selected as the sampling frame for two reasons: it had a high proportion of qualified readers, and it reached the part-time farmer segment better than most other farm publications. MOVING AHEAD After obtaining approval from Janzen on the research objectives and design, Wilson’s next step was to develop the questionnaire to be used in the research. Although Wilson had a lot of experience designing data collection instruments, this was always a challenging part of an marketing research project.
Page 4 9B04A022 Exhibit 1 ALLOCATION OF SAMPLE TO REGIONS Region Value of Custom % of Total Proposed % Number of Work (millions) in Population of Total in Sample Observations Southern $60.3 40% 25% 100 Western $51.3 35% 25% 100 Central $15.0 10% 25% 100 Eastern $16.7 15% 25% 100 Gross Sales % of Total Proposed % Number of Category Gross Sales in Population of Total in Sample Observations Small Less than $50,000 40% 40% 160 Medium $50,000 to $100,000 30% 30% 120 Large $100,000 to $250,000 20% 20% 80 Extra Large Over $250,000 10% 10% 40 Exhibit 2 ALLOCATION OF SAMPLE TO REGIONS AND FARM SIZE Southern Western Central Eastern Total Small 40 40 40 40 160 Medium 30 30 30 30 120 Large 20 20 20 20 80 Extra Large 10 10 10 10 40 Total 100 100 100 100 400