a lower income than is the case for Boston 

 and Detroit. Only 3 percent of the 

 Birmingham respondents had one or both of 

 their parents bom outside of the United 

 States compared to U2 percent in Boston 

 and 23 percent in Detroit. A slightly- 

 higher proportion of the Birmingham respond- 

 ents were en^jloyed as conpared to the other 

 two cities. Finally, a higher proportion 

 of the Boston respondents received an edu- 

 cation beyond the eighth grade than was the 

 case in Birmingham and Detroit. 

 Percent*^ of 

 r«*pon4anta 



Boston Ortrolt 



FIGURE IV . —PERCENTAGE OF 

 NEGRO RESPONDENTS 



The personal characteristics of the 

 households and homeraakers were tabulated 

 for each of the canned salmon consumption 

 groups defined for this study. These 

 tabulations revealed frequency of canned 

 salmon usage in these cities to be asso- 

 ciated with the following characteristics : 

 marital status, size of household, occupa- 

 tion of respondent's husband, age of home- 

 maker, family income, education of home- 

 maker, nativity of parents of horaemaker, 

 religion of family, and race. 



In Binningham, the salmon users groups 

 contained a higher proportion of: manual 

 laborers; homemakers 26 to kS years of age; 

 a family income ranging from $5,000 to 



$7,000; homemakers with 8 to 12 years of 

 education; and Negro families. In addition, 

 Detroit users included a higher proportion 

 of homemakers bom in a southern state; 

 manual laborers did not comprise a signifi- 

 cant percentage of the Detroit users. Bos- 

 ton users respondents included a larger 

 number of homemakers with a college educa- 

 tion; more families in the highest income 

 group; and a larger number of families 

 affiliated with a religion other than 

 Catholic or Protestant. 



A statistical summary of findings 

 based on classification of the data by 

 personal characteristics of the households, 

 together with data on product image men- 

 tioned in the next section, will be made 

 available for a limited time upon request 

 to the Fish and Wildlife Service by persons 

 having a need for such data. 



MOTIVATIONAL ANALYSIS 



The motivational analysis in this sur- 

 vey was carried out by two different methods . 

 The first method was to ask the respondents 

 open questions as to why they use or do not 

 use salmon so that they could spontaneously 

 mention any reason or motive. The inter- 

 viewers, who were all familiar with probing 

 techniques, were instructed to probe as 

 deeply as possible for any reasons which the 

 respondents did not bring out immediately. 

 There was a series of such open questions. 

 The users were asked what would induce them 

 to serve more salmon; non-users were asked 

 why they do not use salmon; those who had 

 stopped using salmon were asked why they 

 had stopped. The sporadic users were asked 

 why they used salmon so seldom, those who 

 now use salmon less often than in the past 

 were asked why they had reduced the number 

 of times they served it. Similarly, there 

 were open questions as to why people liked 

 their preferred can size; their preferred 

 color; d.nd why they made impulse purchases. 

 The responses to these open questions (as 

 well as the responses to the more usual 

 direct questions ) have been summarized in 

 the first part of this report. 



Determining the 

 Product Image~ 



The second method which was used to 

 study motivations is statistical and 

 requires some technical explanation. The 

 first step in this analysis was to deter- 

 mine the "image" of the product — that is, 

 what each respondent thought of salmon. 



