who buys canned sardines,...unitedstatesdepsurtmentoftheinterior,freda.seaton,secretary-...

WHO BUYS CANNED SARDINES,

U. S. Department of the Interior

Fish and Wildlife Service

Bureau of Commercial Fisheries

Washington 25, D. C.

AND WHY?

Circular 90

United States Depsurtment of the Interior, Fred A. Seaton, Secretary-

Fish and Wildlife Service, Amie J. Suomela, Comrhl ssioner

Bureau of Commercial Fisheries, Donald L. McKeman, Director

WHO BUYS CANNED SARDINES, AND WHY?

A Study of Consxnner Motivation in Three Cities

Prepared in the Branch of Economics

United States Fish and Wildlife Service

Circular 90

Washington, D. C. : June I96O

ABSTRACT

This report contains the results of a marketing

research project directed toward the improvement of

promotional and merchandising techniques for the marketing

of canned sardines. The factors which influence consumers

to buy California and Maine sardines are examined in

detail. Special emphasis is placed upon motivational

analysis using psychological techniques in probing for

information as to why consumers are attracted to canned

fishery products with certain styles of pack, taste, color,

or other attributes. The relationships of race and income

to the product image are given psLrticular attention.

PREFACE

While the trend of total canned fish production in theUnited States has been generally upward since the end of World Max II,

the fortunes of the various cemned fish industries have been diverse.Ceinned tuna, the leader, has been breaking production records but hasbeen encountering increasing competition from Japanese imports. Anacute supply problem grips the canned salmon industry and the market-ing problems mainly ensue from a decline in the salmon runs. Cannedsardines, and particularly the California sardine industry, is con-fronted with both supply problems and the loss of traditional markets.

Marketing reseaurch has something to offer the csmned fishindustries. Indeed, the most apparent common denominator in themarketing problems of all these industries is their need for bettermerchandising methods.

The prime objective of this market research study is to aidthe domestic canned fish industries to expand markets for their products,

Results and findings of the study are especially directed toward theimprovement of promotional and merchandising techniques. However, the

study has also a direct bearing upon other important aspects of canned

fish marketing such as the adaptation of the product to meet specific

consumer preferences.

This report describes the results of a svurvey of vrtiat

induces the consumer to buy canned sardines (or the motivationalfactors which influence the buying habits of household consumersof canned sardines). The survey also examines the buying habitsof household consumers of canned salmon and tuna. Separatereports have been issued for those products.

The study was made by the A. J. Wood Research Corporationof Philadelphia, Pa., under contract to the United States Bureau ofCommerciaLL Fisheries. It was financed with funds made availableunder the Saltonstall-Kennedy Act, approved July 1, 195^ (68 Stat.

376).

The survey was conducted under the general supervisionof Walter H. Stolting, Chief, Branch of Economics. Preliminarystatistical and planning work was done by Adolph Scolnick,Analyticed. Statisticism. The report was edited and adapted forpublication by Alton T. Murray and Frans L. Widerstrom, Jr.,Economists.

TABLE OF CONTENTS

Page

Introduction 1Consumer preferences for ceinned sardines : 2

Use of canned sardines 2

Sardine pack preference 3Sardine size preference 3Domestic versus imported sardines kBuying habits kServing habits ^Interest In saxdlne-paste spread 6Recipe sources 6Ordering sardines in public eating places 6Reasons for not using or seldom using cemned sardines 6Canned sardine adveirtislng 6Personal characteristics 7

The sardine can: 7Si ze of can 7Cans vlth and without keys 8Sardine can illustrations 8

Maine versus California sardines : 8Preferences 8Appearance of California sardines 8Appearance of Maine sardines 8Comparison of Maine and California sardine users 8

Race, income^ and sardine use: 9Objective 9Method 9Consumer preferences 9Use of canned f1 sh 9Consumer product attitudes 10

Orangeburg County restilts

:

10Purpose 10Use of CEUined sardines 10Sardine pack preference 10Sardine si ze preference 11Domestic versus imported sardines 11Serving habits 11Maine versus California sardines 11

Motivational analysis : 12Determining the product image 12Measuring the motivational difference 13Index of possible market gain 13Observations based on computed Indexes of possible marketgain 13

Suggestions 13Tables 15

TABLE OF CONTE^rrS - Continued

Page

Statistical tables : 17Use of canned sardines : 17

Table 1.—Which kind of canned fish do you like best? ... 17

Table 2.—^Ilave you served sardines during the last 12

months? 17Table 3.—During the past four weeks, about hovj often did

you serve canned sardines? 18

Table It.--Who in your family eats sardines? 18

Sardine pack preference : 19

Table 5.—Wiich sardine pack do you usually buy? 19Table 6. --Which kind (pack) do you buy occasionally? .... 19Table 7. —Why do you use the other kinds (packs)? 20Table 8.—Which sardine oil pack do you consider to be

the best quality? 20Table 9.—i^y do you use sardines packed in that

preparat ion? 21Sardine size preference : 22

Table 10.—Do you prefer large or small sardines? 22

Table 11.—Why do you prefer small sardines? 22Table 12 .—^About how many inches long are the sardines

you prefer? 23

Domestic versus imported sardines : 2ii

Table 13.—Do you use only domestic or only importedsardines or both? 2i;

Table lli.—Why do you buy only domestic or only importedsardines? 2k

Table 1$.—Do you think domestic sardines are better,about the same, or not as good as importedsardines ? 25

Table 16.—If domestic sardines had their bones and skinremoved, would you . . . ? 25

Buying habits : 26

Table 17.--Was your most recent sardine purchase plannedbefore you went to the store or did youdecide to buy them while in the store? 26

Table 13.—What made you decide to buy them while in thestore? 26

Serving habits : 27Table 19.—Do you sometimes serve sardines for dinner? .. 27Table 20.—Why don't you serve sardines for dinner? 27

Table 21.—How many small sardines do you serve per adultfor a snack? 28

Table 22.

—

Hovj many small sardines do you serve per adultfor a meal? 28

Table 23.—^Were sardines served in your home when youwere a child? 29

Table 2k.—What would induce you to use more sardines? .. 29

TABLE OF CONTENTS - Continued

Page

Statistical tables: - ContinuedInterest in sardine paste spread : 30

Table 25.—li" sardines were processed into a pastespread, do you think you would use such a

product . . . ? 30Recipe sources

:

31Table 26.—Where have you obtained sardine recipes or

information? 31Ordering sardines in public eating places : 32

Table 2?.—Have you ordered sardines in a public eatingplace during the past two months? 32

Table 28.—Other than yourself, has anyone eating withyou ordered sardines in a public eatingplace in the past two months? 32

Reasons for not using or seldom using canned sardines: 33Table 29.—Is there any special reason why you do not

serve sardines ? 33Table 30.—Did you ever use sardines in the past? 33Table 31.—How long ago did you serve them? 3UTable 32.—'.Vhy did you stop serving sardines? 3UTable 33.—Why do you use ssirdines so seldom? 35Table 3U.—Did you use sardines more often in the past?. 35Table 3^.—Does anyone living in the house like

sardines ? 36Table 36.—Why have you reduced the number of times you

serve sardines? 36Canned sardine advertising: 37

Table 37.—Have you seen or heard any advertising forcanned sardines? 37

Table 38.—iVhere did you see or hear advertising forcanned sardines? 37

Sardine can : 38Table 39.—Do you buy sardines packed in the Large or

small can? 38Table kO.—How many small sardines would you like to

find in the U-ounce can? 38Table I4.I.—Do you prefer a sardine can with or without

a key? 39Table k2.—How much more would you pay for a can with

a key? 39Table h3.—Which of these scenes on a sardine can is

more attractive? UO

TABLE OF CONTENTS - Continued

Page

Statistical tables: - ContinuedMaine versus California sardines : Ill

Table hk.—Where do the domestic sardines you buy-

usually come from? ill

Table k^,—Does the appearance of California sardinesplease you when you open the can? Ul

Table hS,—In what way do California sardines not pleaseyou? U2

Table U7»—^Does the appearance of Maine sardines pleaseyou when you open the can? k2

Table I4.8 . --Comparison of Maine and California sardineuse]?s in Detroit k3

Race , income and sardine use : UhTable U9.—Consumer preferences, by race and income khTable 50.—Race, income, and consumer attitude toward

canned sardines hSTable 51.—Race, income, and use of canned fish U5

Characteristics : 1;6

Table 52.—Selected socio-economic characteristics ofhouseholds, homemakers hS

Appendix : U7Survey methods: hi

Questionnaire Ii7

Sample design U?Fie Id work 50Data processing procedures 50Sampling errors 51

Questionnaire S^

viii

WHO BUYS CANNED SARDINES, AND WHY?

INTRODUCTION

In market research it is important toknow how many people do what . It is evenmore important to know why .

The methodology of supplying answersto how-many-people-do-what is well estab-lished. The first part of this report is

concerned with the interpretation of theresults of household consumers' responsesto questions on what buying habits, servinghabits, etc., they have. In effect, thisinterpretation amounts to deciding how manyhousehold consumers prefer particularattributes of canned sardines in relationto other groups with different preferences.The selection of a random sample repre-sentative of all the householders in theareas surveyed was determined by statis-tical methods in common use. Also includedin the first section of the report is theanalysis of consumer responses to the useof two related motivational researchtechniques—the open question and the probe.These techniques represent an initial stepin the process of learning the why ofconsumer buying habits

.

tVhile the study of marketing behaviorover several decades hcis developed a numberof methods of investigating the why of con-sumer habits, motivational research is

relatively new. Practitioners in the fieldof motivational research sometimes disagreeas to the emphasis to be placed upon thespecial techniques drawn from any one ofthe social sciences such as statistics,psychology, economics, and sociology. Theprincipal techniques of motivational re-search in the field of consumer marketingbehavior, however, are derived frompsychology.

The second part of this report is con-cerned with the results of the motivationalanalysis of the marketing behavior of con-

sumers of canned sardines based on otherresearch techniques. Motivational marketsurveys require the services of a stafftrained to interrogate consumers withspecial probing techniques, and a highlyskilled research staff is needed tointerpret the results of the recorded re-sponses. Moreover, motivational researchstudies are much more expensive thanconsumer surveys using conventionalstatistical methods. This situation has

a direct bearing upon the size of themotivational research survey which can be

made for a fixed sum available for con-sumer research. As a compromise betweenthe maximum population coverage to findout how many people do what with cannedfish and the limitation imposed by thecost of motivational research into whythey used it, three urban markets v;ere

selected for study instead of a nationalsurvey.

The populations under study consistedof households within the urba'ized areasof Boston, Massachusetts; Detroit, Michigan;and Birmingham, AlabaFia. In addition

,

Negro households located in the rural areasof Orangeburg County, South Carolina weresurveyed, Negro households in rural areasof the southern states represent an impor-tant market for canned sardines . TheOrangeburg County results will be summarizedin this report and will show the cross-classification of various marketing data byrace for Birmingham and Detroit. A westerncity was not included in the survey becauseof a lack of funds to cover the cost ofinterviews. Area probability samples wereselected to represent the populationscovered and interviews were completed withthe homemaker or person mainly responsiblefor planning the meals as follows

:

Area Number of Interviews

Birmingham

The frequency of serving cannedsardines in the past k weeks varied some-what among the three cities. The sardine-

user households in Birmingham averaged 2.7servings per household compeured to 2.k in

both Boston «md Detroit.

Sardines were eaten by all familymembers in 62 percent of the sau-dine-user

households in Birmingham, 36 percent inBoston, and 55 percent in Detroit.

Sardine Pack Preference

Seurdiae users in Boston usually buysardines packed in olive oil. Sixty- sevenpercent of the respondents bought theolive-oil pack. Saurdines packed in vege-table oil were bought by only 18 percent;tomato sauce, by 5 percent; sild (herring)oil, by 9 percent; and mustard sauce ornatural pack by 1 percent each.

Sardines packed in vegetable oil arethe most popular in both Birmingham andDetroit. Of the ssurdine users in Birming-ham, k6 percent usually bought sardinespacked in vegetable oil; 22 percent, oliveoil; 16 percent, natural oil; 7 percenttomato sauce; k percent mustard sauce and2 percent sild oil. The percentages forDetroit are as follows: vegetable oil, 57percent; olive oil, 29 percent; natural oil,

tomato sauce, and musteird sauce, 5 percenteach; and sild oil, k percent.

Loy6d.ty to the preferred preparationis strongest in Boston tAere 83 percent ofthe sardine users remain consistent buyersof the same oil or sauce. In Detroit, 69percent of the users adhered to theirchoice of preparation. The proportion ofthese users in Birmingham is 52 percent.Most of the users in Birmingham and Detroitwho sometimes buy another prepeurationchange either to tomato sauce (35 percentand 3I+ percent, respectively) or to mustardsauce (33 percent and 31 percent, respec-tively). Those ^o switched said they didso mainly because they 'like variety."

^An important technique used inmotivational research is the "openquestion"—one which seeks the whyof consumer behavior. Such ques-tions permit the respondent toreply freely and do not restricthis choice of answers to the

limited categories imposed bythe direct or closed type. Ifthe respondent's reply is meeoi-

Ingful it gives a reason as towhy he thinks or feels the wayhe does. A response of thetype 'Just because I like it"would not be considered ade-quate and it would be theresponsibility of the inter-viewer to focus the respond-ent on more specific areas in\rtiich to answer. "Hie focusingprocess is known as probing;it is not used in InstancesTrtiere the initial reply isdeemed satisfactory by thespecially-trained interviewer.In cases >rtiere probing is used,it must be handled skillfullyso as not to bias the respond-ent's answer ._7

Tbe first open question asked ofsardine users sought the reasons for theirsardine pack or preparation choices. Ananalysis of the responses of consumersindicates that there was scane tendencyfor those who buy sardines packed in oliveoil to mention as reasons "better taste,

"

'more digestible," or "better for health. "

Other respondents in this category alludedto preferences by husbemds or children forthe olive-oil pack. On the other hand,there was a tendency for those who boughtsardines prepared in vegetable oil to men-tion "lower price" more often than thosewho used other sardine packs.

In all three cities, sardines packedin olive oil were considered to be of thebest quality--by 52 percent in Birmingham,65 percent in Boston, and ^5 percent inDetroit. Soybean oil was considered thehighest quality pack by 28 percent of theDetroit users, but was mentioned by only8 percent in Birmingham and 3 percent inBoston.

Sardine Size Preference

Respondents were asked two questionsabout the size of the sardines they pre-ferred. In the first instance, they wereasked \Aether they preferred the tmn}

}

orthe large sardines; and in the second,they were asked to specify the approximatelength in inches of the sardines they likedto buy.

About 80 percent of the respondents

said they preferred small to large sardines.

"Better taste" and "better flavor" vere the

principal reasons given for selecting the

small size. When specifying the number of

inches, nesirly 90 percent indicated a pref-erence for sardines 5 inches or less in

length.

Domestic versusImported Sardines

More than one-half of the users in

Birmingham, 56 percent, and Detroit, 53percent, served only domestic sardines.Approximately the same proportion of Bostonusers, 3^ percent, served only importedssurdines. In each of the three cities,about one -third of the sardine users servedboth dcmestic eind imported canned sardines.'Lower price" was the leading reason givenby those who use only dcasestic sardines.The vast majority of those who use onlyimported sardines say that imports "tastebetter."

When evaluating the relative qualityof dcanestic and imported sardines, 55 per-cent of the users in Detroit considereddomestic about the same as imported . Only26 percent of the Boston users and 3^ per-cent of those in Birmingham were of thisopinion. Seventeen percent of the usersin Detroit thought domestic sardines werebetter . The percentage of users in thisgroup in Boston was 6 percent; in Birming-ham, 20 percent. Domestic sardines weretermed not aa good by ik percent of the

Detroit users, 36 percent of those in

Boston, and 17 percent in Birmingham.

Sixteen percent of the Boston users,

19 percent of those in Birmingham, and 21percent of the Detroit users felt thatthey would eat more domestic sardines ifthe bones and skin were removed.

Buying Habits

Only 12 percent in Boston, 21 percentin Birmingham, and 31 percent of the

Percentageof sardine

1 1 imported dCBtestic both

60 ^

Boston Detroit Birmingham

FXGITRE II. —PERCENTAGE OF SARDINE USERS WHO BOUGHT ONLY DOMESTIC,

ONIZ IMPORTED, OR BOTH TYPES OF CANNED SARDINES

sardine users in Detroit bought sardines onimpulse at the time of their last pierchase;

the remainder made a planned purchaise. Theexplanation of their behavior by impulsebuyers included "on sale," "Just happenedto notice them," and "Just wanted to havethem on hand."

Serving Habits

Slightly more than one-half of thesardine users in Birmingham sometimesserved sardines for dinner compared withonly about one-eighth of those in Bostonand nearly one-third of the users inDetroit.

Ôpen questions and the probingtechnique also may be used whenattempting to ascertain >rtiy theproduct is or is not used in acertain way. Specific, spontane-ous responses of the type elicitedonly after skilled probing areImportant sources of informationfor those interested in expandingthe market for canned sardines. 7

"Hie analysis of the respondents'answers to an open question eind probesrevealed that the chief reasons for notserving sardines for dinner &re the feel-ing that sardines are primarily a luncheon,snack, or picnic food; or that there arenot enough of them to make a full dinnermeal.

Sardine users in Birmingham served anaverage of 4.9 small sardines per adult fora snack compared to an average of 5.9 inDetroit and 6.3 in Boston.

Birmingham seurdine users served anaverage of 5^3 sardines per adult for ameal. This average is based on the 53 per-cent of those in the city ô knew thenumber of sardines they serve. The averageof 5.'* in Detroit and 7^6 in Boston werebased on only 39 percent and l7 percent ofthe users in the respective cities.

In 78 percent of the seurdine-userhouseholds in Birmingham, sardines wereserved in the heme 'vên the respondent wasa child. For Boston and Detroit, the com-parable percentages were 71 percent and 70percent, respectively.

^The replies of consumers toopen questions and their re-sponses to the use of the prob-ing technique revealed therelative Importance of pricereduction and other considera-tions as motives for the morefrequent use of canned sardines._/

Seventeen percent of the sardine usersin Birmingham and 15 percent in Detroitindicated that a lower price would leadto an increase in their consumption of theproduct. Only k percent mentioned a pricereduction as an incentive in Boston. Theremoval of the bones and better cleaningat the cannery were mentioned as induce-ments to greater consumption by 16 per-cent of the users in Detroit, 7 percentin Boston, and 6 percent in Binningham.However, k6 percent in Boston, 27 percentin Birmingham, and 21 percent in Detroitsaid that nothing would induce them to eatmore sardines. In addition, about one-fifth of the users in all three citiessaid that they eat sardines often enoughat present.

Percentage ofsardine users

50

40 \-

30

20

10

Boston Detroit Birmingjiam

FIGURE III. —PERCENTAGE OF SARDINE USERSWHO SAID THAT NOTHING WOULD INDUCE

THEM TO USE MORE SARDINES

Interest in Sardine -

Paste Spread

Sixty-three percent of the sardine

users in both Boston and Detroit said that

they would not use sardines that had beenprocessed into a paste spread. The com-

parable percentage for Birmingham was 37percent. In all three cities, the mostfrequently mentioned use for such a spreadwas on crackers.

Recipe Sources

Approximately 8 out of 10 of the

sardine users in all three cities could notname their sources of sewdine recipes. The

most frequently-named soxirce of infonnationwas "friends," but even this source wasmentioned by no more than 10 percent in

any of the three cities.

Ordering SardinesIn Public Eating Places

The great majority of sardine users

—

more than 95 percent in each city- -had not

ordered sardines in a public eating placeduring the two months prior to the inter-view. Similarly high percentages of the

respondents in the user group indicatedthat no one eating with them had orderedsardines in the same period.

Reasons for Not Using orSeldom Using Canned Sardines

/"Motivational research tech-niques- -the open question and the

probe- -were used to great advan-tage when the siirvey sought thereasons for not using, seldomusing, or discontinuing the useof canned sardines./

The "never users" (those who had notused sardines in the past 12 months) wereasked if there was any special reason whythey had not served sardines. By far the

most frequently mentioned reasons werethose associated with the lack of appealto the senses--taste, smell, and appearance.Also "too oily" was mentioned as a reasonby 35 percent in Boston, 20 percent inDetroit, but only 11 percent in Birmingham.

sardines in the past. Approximately60 percent of these "past users" hadserved sardines within the last h years.

The respondents in this group said theyhad stopped using canned sardines becausethey did not like the product's taste orsmell; the respondent's husband or chil-dren did not eat sardines; the family wasreduced in size; or for health or dietreasons

.

The leading reasons given by the"sporadic users" (those who had seirved

saurdines in the past 12 months, but not inthe past k weeks) for using canned sardinesso seldom slIso referred primarily to a lackof sense appeal either to the homemaker orto other members of these families. Approx-imately one-fifth of the "sporadic users" inBirmingham and one-fourth in both Bostonand Detroit said they only used sau^dines

for snacks or on picnics. Twenty-one per-cent of the "sporadic users" in Boston, 28percent in Itetroit, and kl percent inBirmingham said they had used sardinesmore often in the past. A reduction inthe size of the family and heeLLth or dietreasons were most frequently mentioned asmotives for the less frequent use of theproduct

.

Nearly 60 percent of the sardine non-users ("never users" and "sporadic users"combined) in Birmingham and Detroit re-ported that at least one member of thehousehold liked sardines.

Canned Sardine Advertising

About 80 percent of all respondents

in Birmingham said they had not seen or

heard any advertising for canned sardines.

The comparable proportions for Boston and

Detroit were 87 percent and 88 percent,

respectively.

Among those who had been exposed tocanned sardine advertising, newspapers werereported most frequently as the source byrespondents in Birmingham and Detroit.Newspapers and magazines were mentioned assources by approximately the same propor-tions of the respondents in Boston.

Less them k out of 10 of the "neverusers" in each of the three cities had used

Personal Characteristics

The socio-economic characteristics 2/of the households and homemakers in all

three cities differed considerably with re-spect to race, religion, income, birthplaceof parents , employment status , and edu-cation. In Birmingham, 36 percent of thehouseholds were Negro ccanpared with 19 per-cent in Detroit and only 2 percent in

Boston

.

Percent*^ ofrespondents

?w

respondents indicated that the preferredsardine can size was not ri^t for theirhouseholds.

The respondents were shown a full-sizesketch of a li-ounce sardine can and ^ereasked to state the number of sardines theywould like to find in this size can. TheBinr.inghain average of 7.5 small sardiiies tothe can was somewhat less than 10.6 averagein Boston and 10.7 in Detroit.

Cans With and

Without Keys

subtly more than 80 percent of thesardine users in Birmingham, 86 percent inDetroit, and more than 90 percent in Bostonpreferred a sardine can with a key. How-ever, hi percent of those who expressedsuch a preference among the Birminghamsardine users would not be willing to paymore for a can with a key. In Detroit, 53percent and in Boston, 68 percent wouldnot pay more. The majority of those whowould be willing to pay more for a canwith a key indicated a willingness to payonly an additional one cent a can.

Sardine Can Illustrations

Respondents were shown sketches of afish scene and a marine scene. They vjere

then asked to indicate which would be themore attractive on a sardine can. Seventypercent of the respondents in Birminghamand 55 percent in Detroit preferred thefish scene. In Boston, the marine scenewas preferred by 53 percent; the fishscene, by hh percent.

MAINE VERSUS CALTFCRNIA SARDINES

Preferences

Three -fourths of the sardine users in

all three cities who serve domesticsardines indicated that they usuallybought Maine sardines. In Birminghain , lU

percent of the users usually bought Cali-fornia sardines; in Boston 10 percent andin Detroit 19 percent.

Appearance ofCalifornia Sardines

Fifty-two percent of the sardine usersin Birmingham thought that the appearanceof California sardines was satisfactory

when the can was opened; in Boston, 35 per-cent were of this opinion; and in Detroit,

U5 percent. The proportion who did notfind the product's appearance satisfactoryin Birmingham was 15 percent; in Boston,

6 percent; and in Detroit, 19 percent.

/The skilled interviewerssuccessfully used the probetechnique to discover the waysin which the appearance of theCalifornia sardines did notplease the latter group ofrespondents ._/

In response to an open question andthe subsequent probe, those users who werenot pleased with the appearance of theCalifornia sardines mentioned "look toolarge" most frequently as the reason.

Appearance ofMaine Sardines

In Birmin^am, 82 percent of thesardine users were pleased with the appear-ance of Maine sardines when the can wasopened; only 10 percent were not. Fifty-one percent in Boston thought that theappearance was satisfactory; only 6 per-cent were of the opposite opinion. Theappearance favorably impressed 80 percentof the Detroit users; only 12 percentindicated that it was not pleasing to them.

Comparison of Maine andCalifornia Sardine Users

Detroit was the only city thatprovided a sufficient number of users ofCalifornia sardines to make possible a

comparison of Maine and California sardineusers. The items used in the comparisonof the Detroit user groups are summarizedas follows:

Sardine users who usually boughtMaine sardines served them an average of2,8 times in the four weeks prior to theinterview. California sardines wereseived 2,9 times during the same periodby those who usually bought that product.

Sixty-five percent of the Mainesardine users bought the product packed invegetable oil most frequently. One-fourthof th-r California users indicated thatthey usually purchased sardines packed intomato sauce; only one -twentieth of theMaine sardine users usually chose thisstyle of pack.

Ibe appearance of California sardines

in the open cem was pleasing to more than

90 percent of the California seurdine users,

but to only U5 percent of the Maine productusers. Eighty percent of the users in bothgroups indicated that the appearance ofMaine sardines in the open can was satis-

factory.

More than 80 percent of the Maineseurdine users indicated a preference for a

can with a key, canpared with 65 percent ofthe California users.

Impulse buying is somewhat morefrequent among California users (38 percent)than among Maine users (27 percent).

Preference for the large-size sardinecan is more pronounced among Californiausers, 33 percent, than among Maine users,k percent.

California users tend to have higherincomes than Maine users. Slightly morethan ho percent of the California userswere in the highest income bracket, $7>000a year or more. Only 17 percent of theDetroit families were in that group.

Eighteen percent of the Maine usersindicated that they would use more domesticsardines if bones and skin were removed,compared with 10 percent of the Californiasardine users.

RACE, INCOME, AND SARDINE USE

Objective

A major domestic market for Maine andCalifornia sardines has been traditionallyamong Negroes and low income groups in theSouthern States. There has been a tendencyto attribute the decline in these marketsin recent years to the rising standards ofliving generally among Negro consumers.

One of the objectives of this surveywas to investigate jrtiether or not there isa psychological attitude of "poor man'sfood" associated with Negro families' useof sardines. Conversely, as Negro families'Incomes rise, do they switch to tuna andother canned fish because they associate itas a food for "higher class" families?

Method

In order to investigate this problem,white £Lnd Negro households in Birminghamand Detroit were classified into threeincome groups, rsuiging from low inccme torelatively high Inccme .3/ The attitudesof these groups were examined on a numberof guestloiiB.

Consumer Preferences

Among Negproes in Blzmingham in thelowest income group, 12 percent mentionedsardines as their choice compared with 10percent in the next higher Income categoryand 9 percent in the group having thehighest annual income.

Sixteen percent of the Detroit Negrorespondents in the lowest income classnamed sardines as their preference. Thelike percentage in the middle income groupwas 7 percent and in the highest Incomecategory, 11 percent. However, the modestdecline in the mention of sardines asincome Increased was not restricted toNegro families. There was a parallelmovement among >riiite respondents as Incomeincreased. For example, among Detroitwhite households in the lowest incomegroup, 10 percent Indicated that sardinesare "liked best." The ccraparable percent-age in the next hl^er Inccme class was 5percent and in the bluest Income groirp,

k percent.

Use of Canned Fish

Tbere was some increase in the use ofcanned tvuia, salmon, and sardines amongNegro families as Inccme Increased. Ofthe Negro families in Birmingham, 6k per-cent of those in the lowest Income groupused tuna in the 12 months prior to theinterview; and 82 percent in the middle in-come category also used the product duringthat period. The comparable percentages

3/ The income groups were designated asfollows: (1) annual incomes of lessthan $2,999; (2) $3,000 to $^^,999; and(3) $5,000 and over.

for Negro respondents in Detroit were 8hpercent, 79 percent, and 8h percent.

Eighty-seven percent of the Birming-ham Negro respondents in the lowest incomeclass and 97 percent in the higher Incomegroups had used canned seQjnon during the

12 months preceding the interview. In

Detroit, 76 percent in the lowest incomeclass, 84 percent in the middle group, and

97 percent in the highest income categoryalso reported using canned salmon.

There is a tendency toward the

increased use of sardines by Negro respond-ents as their income increases. In Birming-

ham, nearly 80 percent in the lowest incomeclassification and 90 percent in the higherincome categories reported serving c£uined

sardines in the 12 months prior to theinterview. The percentages among Detroithomemakers were 72 percent and 89 percent,respectively. The proportion of \rtilte

families using sardines in the 12 monthsprior to the interview in those cities didnot increase with income . However, the

proportions using tuna and salmon didincrease with increases in income.

Consumer Product Attitudes

The association between income and the

attitude that sardines &re a food for Ne-

groes or poorer people is the opposite ofthat ^ich was anticipated. As inccane

increased among Negroes, the feeling thatsardines are a food eaten by Negroes orpoor people decreased. The proportion ofBirmingham Negroes in the lowest income

class who disagreed with the statement thatsardines are a food for poorer people was

23 percent; in the next higher group, 25percent; and in the highest income classi-fication, U7 percent.

ORANGEBURG COUNTT RESULTS

Purpose

Two hundred Negro households in rural

Orangeburg County, S. C, were included in

the survey because of the Importeuice of

such hoviseholds as a market for cannedsardines. The analysis of the results ofthis phase of the survey indicates differ-ences and similarities between Negro house-holds located in Birmingham and Detroit and.

rural households situated in OrangeburgCounty.

Use of Canned Sardines

Two-thirds of the Orangeburg Countyrespondents named salmon as their preferredcanned fish; 8 percent, tirna; and ik per-cent, sardines. The proportion choosingsardines was nearly three times that ofthe over-all average of 5 percent for anrespondents in the three cities.

Only 20 percent of the rural home-makers had not served sardines in the 12months prior to the interview. This per-centage was ccanparable to that for Negrohouseholds in Birmingham, 17 percent, andsomewhat lower than that of Detroit Negrohouseholds, 27 percent.

The frequency of serving sardines inthe ii-«eek period averaged 2.3 times in

Orangeburg County and 2.0 times in Negrohouseholds in Birmingham and Detroit. Allsardine-user respondents in Detroit aver-aged 2.4 servings during the same period;in Birmingham, 2.7 servings.

Sardines are eaten by all familymembers in 90 percent of the sardine-userhouseholds in Orangeburg County, comparedwith 76 percent of the Negro households inBirmingham and 64 percent in Detroit.


Sardines peicked in vegetable oil arethe most populeir in Orangeburg County with

39 percent of the homemakers reportingthis preference. Negro respondents inBirmingham, 56 i>ercent, and Detroit, 78percent, also named the vegetable-oil pack.Natui"al oil and tomato sauce—21 percentand 19 percent, respectively--are next inpopularity in the rural area, followed byolive oil £md mustard sauce at 10 percenteach.

Loyalty to the preferred product isstrongest in the rural area and NegroDetroit households where 69 percent of the

sardine users remain consistent buyers of

the same oil or sauce. The comparable per-centage for Negro user-households inBirmingham is 48 percent. Most of the users

in Orangeburg County and those in Birming-

ham change either to mustard sauce (62 per-cent and 35 percent, respectively) or to

tomato sauce (42 percent and 33 percent,respectively). Among Negro sardine-users

10

In Detroit, the changes are to tomato saucehd percent; olive oil, 19 percent; andnatural oil, 15 percent. Those who switchsaid they did so mainly because they "likevariety .

"

Sardines packed in olive oil areconsidered to be of the best quality by all

three groups of Negro ssu-dine-user respond-ents. Soybean oil and cottonseed oil packswere named also by significant percentagesof these respondents.


Nearly three-quarters of theOrangeburg County respondents prefer gmaii

rather than large sardines. The leadingreasons given were "better taste" and"sweeter." More than one-third of thesehomemakers did not know the length of thesardines they liked best. Another thirdstated a preference for 2- to 5-inch fishand the remainder, 2 inches or less.

Danestic versusImported Sardines

Only danestic sardines are used by 89percent of the Orangeburg County respond-ent households. In Birmingham, 57 percentof the Negro respondents used domesticsardines and in Detroit, 67 percent. About30 percent of the homemakers in the urbanareas served both dcanestic and Importedsardines, cooipared with 6 percent in therural group. Lack of awareness of otherkinds (imported) was by far the most impor-tant reason given by Orangeburg Countyrespondents for their preponderant use ofthe domestic product.

Serving Habits

More than three-quarters of theOrangebvirg County hanemakers sometimesserved sardines for dinner, ccanpared with63 percent of the Negro group in Birminghamand only 3'* percent in Detroit.

Two-thirds of the rural respondentsserved only 1 or 2 small sardines per suiult

for a snack. The over-all average was only2.2 sardines for a snack portion. Servingsmade by the comparable group in Birminghamaverage 3.9 sardines; In Detroit, 5.3.Only 2.5 8«u:dlnes were served per adultfor a meal in the Orangeburg County group;

5.6 and 5-5 sardines were the average peradult meal in Birmingham and Detroit,respectively.

Nine out of ten rural homemakersreport that sardines were served in theirhomes when they were children. About 8out of 10 Negro urban respondents made thesame repoirt.

Tliirty-slx percent of the OrangeburgCounty homemakers indicated that theyseiTred sardines often enough at the pre-sent time. An additional 7 percent indi-cated that nothing would induce them toserve more sardines. A lessening orelimination of unpleasant odors was namedas a reason for increased use by 15 per-cent. Better cleaning and removal ofbones was listed by 11 percent.

"Lower price" was mentioned as aninducement for increasing their sardineconsvmiptlon by 22 percent of the Negrorespondents in Birmingham and 2k percentin Detroit. Only in Detroit was the odorproblem mentioned by as many as 10 percentof the homemakers. Of the Birminghamgroup, 23 percent indicated that presentservings of sardines were about the amoxintthey would want; in Detroit, the propor-tion was 17 jjercent. Twenty-one percentin Blrmlnt^am and ik percent in Detroitwere of the opinion that nothing wouldInduce them to serve more saxdlnes.

Maine versusCalifornia Sardines

Eighty-seven percent of the OrangeburgCounty domestic-sardine users usuallybought Maine sardines; 13 percent, theCalifornia product. The proportions forNegro households using Maine sardines inBirmingham were 7!* percent, emd 16 percentusing California sardines; those in Detroll^

72 percent and ik percent, respectively.

Ifesu-ly 60 percent of the ruralrespondents had no opinion whether appear-ance of California sardines was pleasingto them in the open can. In the BirminghamNegro group, 59 percent indicated that theproduct's appearance was satisfactory; inDetroit, k9 percent expressed this opinion.

More than 90 percent of the Negrorespondents in the rural area cmd in

11

Birmingham were pleased by the appearanceof Maine sardines in the open can; the

comparable percentage in Detroit is Qkpercent

.

MOTIVATIONAL ANADfSIS

The motivational ansd^sis in this

svtrvey was carried out by two differentmethods. The first method was to askrespondents open questions as to why theyuse or do not use sardines so that theycould mention spontaneously euiy reason ormotive. The interviewers, who were allfamiliar with probing techniques, wereinstructed to probe as deeply as possiblefor any reasons which the respondents coxild

not bring out immediately. There was aseries of such open questions. The userswere asked what would induce them to servemore ssurdines; nonusers were asked \rhy they

do not serve saxdlnes; those who hadstopped using sardines were asked vhy theyhad stopped. The sporadic users were askedwhy they used sardines so seldom; those whonow use sardines less often than in thepast were asked why they had reduced thenumber of times they served them. Similar-ly, there were open questions as to ^ypeople use sardines packed in a particularoil or sauce; why they used only domesticor imported sardines; why they preferredlarge or rnnall SEurdines. Iê responses tothese open questions ( as well as the re-

sponses to the more usuttl direct questions )

have been summarized in the first part ofthis report.

Determining theProduct Image

The second method >rtiich was used tostudy motivations Is statistical and re-quires some technical explanation. Thefirst step in this analysis was to deter-mine the "image" of the product--that is,

what each respondent thought of sardines,what characteristics she attributes to them,

what associations the product evokes.Therefore, each respondent was asked vrfaether

she agreed or disagreed with a series ofstatements, each representing a chsuracter-

istic of sardines; for example: "Sardineshave a good flavor."

fTbe motivational technique usedin this phase of the anedysis is

known as the "guided association

question." Although the respond-ent is asked only whether or notshe agrees or disagrees with thestatement, the intenriewer actual-ly records the intensity of theanswer. Thus, Strong agreementor disagreement (as well as lessintensely expressed feelings oropinions) is noted by the inter-viewer. In addition, the state-ments on the questionnaire weresometimes phrased positively andsometimes negatively—as forexample: "Sardines have an un-pleasant smell." This was donein order to minimize ât iscalled a "halo" or clusteringeffect \rtiereby a favorable atti-tude toward a product tends tomake respondents attribute allfavorable characteristics tothe product. "Hie skill andtraining of the interviewer areas important to the proper hand-ling of guided association ques-tions as these requirements arewhen dealing with open questionsand the probing technique^

The product-image phase of the studyconducted in Blrmlngîam revealed that morethan 90 percent of the respondents in thatcity regard sardines as a convenient foodand that sardines usually sure not eatenby children only. Between 73 and 85 per-cent of the respondents were of the opinionthat saurdines are not too troublesome toprepare; are used a great deal by Negroes;do not have many uses; are not expensive;and have em impleasant smell.

In Boston, the image is scne^rtiat

different. Eighty-seven percent of theserespondents agreed that sardines usuallyare not eaten only by children. However,only 20 percent eigreed that they are useda great deal by Negroes

—

68 percent neitheragreed or disagreed. The only other lead-ing characteristic on \rtiich there wasagreement between Birmingham and Bostonhcmemakers was on the statement "Sardinesdo not have many uses." In each of thosecities, three-quarters of the respondentsagreed with the statement. Only one othercharacteristic was agreed to by more than70 percent of the respondents in Boston

—

that sardines are good only if they are awell-known brand.

12

Nine out of ten Detroit respondentsagreed that sardines usually are not eatenonly by children. This statement was theonly one agreed to by as many as 90 percentof the respondents. As in Birmingham, 75percent to 8k percent of the respondentsagreed that sardines do not have many uses,that they are a convenient food and not tootroublesome to prepare. All of the otherstatements were eigreed to by less than 70percent of the respondents.

Rather wide differences of opinionamong the respondents in the three citieswere noted for the items "sardines have anunpleasant smell," "are food for poorerpeople, ' and "are used a great deal byNegroes."

Measiu'ing theMotivational Difference

The aim of motivational analysis is todetermine the characteristics which havethe greatest influence on the behavior ofthe respondent. The approach used in thisstudy to measure the strength of a motivewas to determine first the ratio of heavyusers among those who agree with the state-ment and compare it with the ratio of heavyusers among those who do not agree with it.The difference between these ratios, whichwill be called the motivational difference,indicates whether agreeing with the state-ment has influence and measures the extentof the motivational strength of the state-ment. The greater the difference betweenthe ratios, the stronger the influence ofthe specific statement. The selection ofheavy user groups Is justified becausethere is an interest in converting not onlythe never users into reguleur users but alsoto transform the light users into heavyusers of canned seirdlnes.

Considerable agreement exists amongthe three cities with respect to the itemswith the highest motivational differences.Good flavor is first in all three citiesand is clearly the most important motive.The motivational differences sire high for"are not undesirably oily, " "have a pleasantaftertaste," "are a food of high quality,"and "have a nice appesurance in the can."Ease of prepatration is eilso an Importantmotive. In Boston, "not troublesome to pre-pare" ranked high as a motive. "Not hard tomake look good ' was listed as an important

item in Detroit; it also was a significantconsideration in Birmingham.

Index of PossibleMarket Gain

As a third step, the extent to >rtiich

a motivating characteristic is alreadyattributed to the product by hcmemakersmust be measured.iy Conversely, the pro-portion of homemakers irtio are to be con-vinced that canned sardines have a specificdesirable quality must be established. Itis in this group that the potential marketgain is greatest. The result of multiply-ing the motivational difference by thepotential to be convinced yields &n indexof the possible market gain.

Observations Based onComputed Indexes ofPossible Market G€d.n

Only the itans "sardines have apleasant aftertaste," "are not hard tomake look good, " and "are not undesirablyoily" are among the leaders in all threecities. This is explained for the mostpart by the fact that the survey citiesare located in different regions of thecountry. The variations in the list amongthe cities axe a reflection of differencesin socio-economic characteristics andmarketing behavior.

Ccraraon to both Boston and Detroit arethe items "sardines have a good flavor" and"have many uses." linportant in Birminghamand Boston is the ch€u:acteristic "niceappearance in the can." "Sardines are onlygood if a well-known brand" is among theleading Indexes in Birmingham and Detroit.

SUGGESTIONS

The following suggestions have beenderived from the analysis and summary ofthe data:

hj There is no need to try to convincethat sector of the public that a prod-uct has a certain characteristic ^eneveryone within the sector recognizesthat this is the case.

13

Only a small proportion of the consum-ing public has been exposed to promotionalor advertising material for sardines. Inview of the limited budgets of most sardinepackers, it would be advisable for them tochannel advertising and promotional expend-itures through a central agency of an in-dustry association as is done in Maine. Bymeans of such a program, a consistent, co-herent, and more widespread promotionalcaii5)aign could be undertaken.

Household consumer demand for sardinesis not as strong as the demand for othercanned fish; only 5 percent of the respond-ents in any of the survey cities expressedtheir preference for sardines. If any sub-stantial progress is to be made in increas-ing sardine consumption, it is particularlyimportant that a strong and sustained effortbe made to raise the prestige of the product.Of nesirly equal importance vould be theintroduction of product innovations whichwould give the product wider appeal.

Promotional and advertising effortsmust be directed toward increasing thefrequency with which sardines are used andalso toward increasing the proportion ofuser households. There are serious weak-nesses in both these areas at the presenttime.

and serving sardines. Results of theresearch would provide a constant supplyof recipes and product uses which thenwould be presented to the public throughadvertising and promotional media. At thepresent time, the vast majority of house-wives are unable to indicate any sourcefrom which sstrdine recipes are obtained.

Special prices on the larger c£m ofsardines should be associated with newrecipe ideas in an effort to increase thesales of the large can. Cons\mier satis-faction with the smaller can is very hi^at present; however, it is related to thelimited use made of sardines. Indicationsare that this satisfaction could be over-come by means of special price inducementsfor the piirchase of the larger cem and bymaking housewives aware of uses >rtiich

Involve greater quantities of sardines.

A promotional program directed at therestaurant trade is required. The propor-tion of consumers ordering sardines inpublic eating places Is very small. Spe-cial advertising materisLl designed toinform restaurant owners and cooks shouldbe prepared and distributed. The purposeof this program would be to Increase thenumber of sardine offerings on restaurantmenus.

Another area in which there is a

possibility of increasing sardine consump-tion is indicated by the different servinghabits noted in the three cities. In bothBoston and Detroit, seirdines are servedmuch less frequently as a dinner item thanin Birmingham. One of the functions of anyeducational campaign designed to increasesardine use should be to widen the practiceof serving sardines for dinner.

A method of increasing the use of theproduct as a dinner item is suggested bythe small number of sardines served per per-son at any meal (about 5 to 7 sardines).If sardines are to be used as a main course,it is important that recipes be developedwhich would require sardines to be used inlarger quantities.

It would be advisable for seu-dine

packers to contribute to the support of afull-time experimental kitchen. The pri-meiry function of such an establishmentwould be to investigate ways of prepeurlng

As new methods of serving or packingsardines are developed, particular emphasisshould be placed on sauces or additivesthat would reduce the amount of oil. Italso would be advisable to disguise thetaste of sardines so as to make the quali-ties presently associated with aftertastemore acceptable to consumers.

Because of the differences whichappear among the leading indexes for thethree cities, at least some newspaper andmagazine promotional themes should havemore limited appeal to consumer buyingmotives. Limitations of this sort woxild

enable the advertiser to vsiry or controlhis activities on a local or regionalbasis

.

Unless the shape of the can is changed,consumers perfer that a key be supplied.

14

TABIES

The tables showing the percentagedistribution of the responses to each ofthe questions pertinent to this report areincluded in the next section. A weightedbase vas employed for the computation ofeach percentage distribution. This baseis shovn at the bottom of each tablecolumn . Weighting the actued. number ofinterviews completed in each city wasnecessary since a smsLLl number of thesampling units were subsampled to avoid&n excessive nvmiber of interviews in anyone interviewing assignment. IMs proce-dure was necessary in those sample areasegments which had grown considerablyin nunber of households since 1950. Inaddition, the total Detroit area wasdivided into zones which were eitherpredominately white or Negro with theformer sampled at one-half the rate of the

latter zone; weighting was employed torestore the proportionableness of the racedistribution in this city.

No weighting was attempted for house-holds selected for the sample but notinterviewed (refusals, vinable to contact,etc.). The actual number of completedinterviews and the weighted base for thetoted, respondent population in each cityare shown below.

City

BirminghamBostonDetroit

Actual numberof interviews

585

553609

Weightedbase

669572916

Percentage distributions were computedirtienever the weighted base was 25 or more.

15

STATISTICAL TABLES


TABLE 1.—WHICH KIND OF CANNED FISH DO YOU LIKE BEST?

Kind of canned fish All respondentsBirmingham Boston Detroit

Percent Percent Percent

TunaSalmonSardinesShrimpDon't know

Weighted base

100

k2k25

38

(669)

100

56

17k

Ik

9

(572)

100

hi276

11

9

(916)

TABI£ 2. --HAVE YOU SERVED SARDINES DURING THE LAST 12 MONTHS?

Response All respondentsBirmingham Boston Detroit


Yes


TABLE 3.— DURING THE PAST FOUR WEEKS, ABOUTHOW OFTEN DID YOU SERVE CANNED SARDINES?

Number of timesThose vho have served sEurdlnes

In the last yeeu:

R1 rm1 ngham


TABI£ 5.—WHICH SARDINE PACK DO YOU USUALLY BUY?

Kind of pack Sardine usersBirmingham


TABLE ?.—WEf DO YOU USE

ON

O

HOS


Size

TABLE 10.— DO YOU PREFER LARGE OR 9«ALL SARDINES?

Sardine usersBirmingham Boston Detroit


LargeSmallDon't knov

Weighted base

100

116k

5

:i92)

100

Ik81

5

(iito)

100

10

7911

(291)

TABLE 11.—WHY DO YOU PREFER SMALL SARDINES?

ReasonsSardine users vho preferonly small sardines

Birmingham


TABIE 12. —ABOUT HOW MANY INCHES LONG ARE THE SARDINES YOU PREFER?

Length Sardine usersBirtaingham Boston Detroit


100 100 100

2 inches or less2 to 5 inchesMore thsm 5 inchesDon't knov

Wei^ted base

21711

7

(192)

21

Ik

5

1772k

7

(lltO) (291)

23


TABLE 13.—DO YOU USE ONLT DOMESTIC OR ONLT IMPORTED SARDINES OR BOTH?

Responses Saurdine usersBirmi ngham Boston Detroit


ImportedDanesticBothDon't know

Weighted base

100

856306

(192)

100

3h15

292

(1^)

100

11533^^

2

(291)

TABLE Ik. --mi DO YOU BUY ONIY DOMESTIC OR ONIY IMPORTED SARDINES?


TABLE 15. --DO YOU THIM DOMESTIC SARDINES ARE BETTER,

ABOUT THE SAME, OR NOT AS GOOD AS IMPORTED SARDINES?

Responses Sardine usersBirmingham Boston Detroit


BetterAbout the sameNot as goodDon't know

Weighted base

100

20341729

(192)

100

6263632

(1^)

100

17

551414

(291)

TABIE 16.—IF DOMESTIC SARDINES HAD THEIR BOHESAND SKIN REMOVED, WOULD YOU . . . ?

Responses

Buying Habits

TABLE 17."WAS YOUR MOST RECENT SARDINE PURCHASEPLANNED BEFORE YOU WENT TO THE STORE OR DIDYOU DECIDE TO BUY THEM WHILE IN THE STORE?

Sardine usersResponses Birmingham Boston Detroit


100 100 100

PlannedImpulseDon't know

77212

126931

Weighted base (192) (11^) (291)

TABLE 18.—WHAT MADE YOU DECIEE TO BUY THEM WHILE IN THE STORE?

Responses Impulse btiyers of sardines

Serving Habits

TABLE 19.—DO YOU SOMETIMES SERVE SARDINES FOR DINNER?



100 100 100

Yes

Serving Habits

TABLE 21. --HOW MARY SMALL SARDINES DOYOU SERVE PER ADULT FOR A SNACK?

Number of sardines SeLTdine usersBirmingham

Serving Habits

TABI£ 23.—WERE SARDINES SERVED IN YOUR HCME WHEN YOU WERE A CHILD?



Yes

Interest inSardine Paste Spread

TABLE 25."IF SARDINES WERE PROCESSED INTO A PASTE SPREAD,DO YOU THINK YOU WOULD USE SUCH A PRODUCT . . . ?

Uses Sardine usersBirmingham

RecipeSources

TABLE 26. --WHERE HAVE YOU OBTAINED SARDINE RECIPES OR INFORMATION?

Sources

Ordering Sardines inPublic Eating Places

TABLE 27. —HAVE YOU ORDERED SARDINES IN A PUBLICEATING PLACE DURING THE PAST TWO MONTHS?

Responses Sardine users

Reasons For Not Using Or SeldomUsing Canned Sardines'

TABLE 29.—IS THERE AHY SPECIAL REASON WHY YOU DO NOT SERVE SARDINES?

Reasons For Not Using Or SeldomUsing Canned Sardines

TABIE 31.—HOW LONG AGO DID YOU SERVE THEM?

Number of yearsSardine non-users who had not

served seirdines in past 12 monthsBirmingham Boston Detroit


100 100 100

1 to 2 years2 to 3 years3 to 4 yestrs

5 years or moreDon't know

Weighted base

22

1921

2513

(88)

15

39li+

311

(84)

192813328

(138)

TABLE 32.—WHY DID YOU STOP SERVING SARDINES?

Reasons For Not Using Or SeldomUsing Canned Sardines'

TABLE 33."WHY DO YOO USE SARDINES SO SELDOM?

Reasons For Not Using Or Seldom

Using Canned Sardines

TABLE 35. --DOES ANYONE LIVING IN THE HOUSE LIKE SARDINES?

Responses Sardine non-usersBirmingham Boston Detroit


YesNoDon't know

Weighted base

100

63

298

100

271360

(477) (431)

100

56

2

(625)

TABI£ 36.—WHY HAVE YOU REDUCED THE NUMBEROF TIMES YOU SERVE SARDINES?

Reasons

Sardine non-users who servedsardines In past 12 months andused sardines more In the pastBirmingham

Canned Sardine Advertising

TABLE 37. --HAVE YOU SEEN OR HEARD ANYADVERTISING FOR CANNED SARDINES?

Responses All respondentsBirmingham Boston Detroit


100 100 100

YesNoDon't know

Weighted base

1980

(1)

(669)

13861

1287

(1)

(572) (916)

l/ Less than one percent.

TABLE 38. --WHERE DID YOU SEE OR HEAR ADVERTISING FOR CANNED SARDINES?

Soiirces

Those exposed to advertisingfor canned sardines

Birmingham

Sardine Can

TABLE 39. --DO YOU BUY SARDINES PACKED IN THE LARGE OR SMALL CAN?

Can size Sardine usersBirmingham Boston Detroit


LargeSmallDon't know

Weighted base

(1)

1790

(192)

(1)

12911

(11^)

(1)

20

91

(291)

1/ Totals more thsm 100 percent as some respondents gave more thanone answer.

TABLE I4O.—HOW MANY SMALL SARDINES WOULD YOULIKE TO FIND IN THE i+-OUNCE CAN?

Number Sardine usersBirmingham Boston Detroit

6 or less

7 or 8

9 or 1011 or 12More than 12Don't know

Weighted base

Average


100100

Sardine Can

TABI£ In."DO YOU PREFER A SARDINE CAN WITH OR WITHOUT A KEY?



With keyWithout keyDon't know

Weighted base

100

8116

3

(192)

100

91

9

(1^)

100

86Ik

(291)

TABLE 42.—HOW MUCH MORE WOULD YOU PAY FOR A CAN WITH A KEY?

AmountSardine users \dio prefer a

sardine can vith a keyBirmingham Boston Detroit


No more1 cent2 cents

3 centsMore than 3Don't know

Weighted base

cents

100

hi27138

32

(156)

100

68166

7

3

(127)

100

5326Ik

3

31

(251)

39

Sardine Can

TABLE if3.- -WHICH OF THESE SCENES ONA SARDINE CAN IS MORE ATTOACTIVE?

Scene Sardine usersBirmingham Boston Detroit


100 100 100

Fish scene 70 hk 55Marine scene 28 53 42Don't knov 2 3 3

Weighted base (192) (ll(0) (291)

ItO


TABLE l*-4. --WHERE DO THE DC»ffiSTIC SARDINES YOU BUY USUALLY COME FROM?

State

Sardine users who usedomestic sardines

Birmingham Boston Detroit


MedneCaliforniaDon't know

Weighted base

(1)

77IkIk

(164)

(1)

791013

(61)

(1)

761913

(255)

l/ Totals more than 100 percent as some respondents gave more than~ one answer.

TABLE 45. --DOES THE APPEARANCE OF CALIFORNIASARDINES PLEASE YOU WHEN YOU OPEN THE CAN?



YesNoDon't know

Weighted base

100

521533

(192)

100

356

59

(140)

100

451936

(291)

41


TABLE 1^6. --IN WHAT WAY DO CALIFORNIA SARDINES NOT PLEASE YOU?

ReasonsSardine users vho do not like

appearance of California sardinesBirmingham


TABLE 48. --COMPARISON OF MAINE AND CALIFORNIA SARDINE USERS IN DETROIT 1/

Preferred pack 2/Users vho usually bviy

Maine Californiasardines S8u:dines

Percent Percent

Vegetable oilTomato sauceOlive oilOther

Appearance - Open can California sardines

PleasingNot pleasingDo not know

Open cein Maine sardines

PleasingNot pleasingDo not know

Preference for can with key

With keyWithout key

Last pvirchase, by type

PlannedImpulse

Effect of skin, bone removal vhen packed

If done, would eat more ssurdines

If done, would eat about same quantityDo not know

Weighted base

65

5

2317

'^

•K3

Characteristics

TABI£ 52.—SELECTED SOCIO-ECONOMICCHARACTERISTICS OF HOUSEHOLDS, HOMEMAKERS

APPENDIX

SDRVEY METHODS

Questionnaire

Ihe development phase of the studyconsisted of 57 depth Interviews conductedby specialists in this type of Interviewing.Respondents were chosen for these Inter-views in a non- systematic, but eLLso non-random method. These interviews consistedof informal and casual discussions coveringaspects of household consumers' preferencesfor canned fish. The respondent was al-lowed to take whatever direction she wantedto in these discussions, following her ownnatural inclinations. No attempt was madeto limit or restrict the discussion topredetermined areas of interest. Only whenthe respondent had exhausted some topic didthe interviewer attempt to give some fur-ther direction to the interviews by askinga very general and open "why" or "how"question.

In addition, a number of specifictechniques were used in these initial inter-views as further aids in eliciting consumerattitudes and motivations. Such techniquesas word association, sentence completion,response projection, role taking, andcartoon tests were used.

After the first few of these 57 inter-views were completed, discussions were heldby the staff and the Interviewers. Sug-gestions were made regeurding proceduralchanges in order to Increase the prospectsfor more complete and detailed information.

Using these 57 preliminary interviewsas a basis, a list of associations wasdeveloped showing all of the relevant areasto be explored in the full scale study.This "item list" formed the basis for de-velopment of the "guided association"portion of the questionnaire.

Seveml drafts of the pre-test ques-tionnaire were then developed, in consul-tation with staff members of the U. S. Fishand Wildlife Sei-vice, with each draft re-ceiving limited field tests by a specied-istin interviewing techniques. An improveddraft of the questionnaire was also for-warded to members of the fish canningindustry for their ccmments and suggestions.

A full scale pre-test of the ques-tionnaire was carried out in the threeurbanized areas selected for the fullscale svirvey, Boston, Massachusetts; Bir-mingham, Alabama and Detroit, Michigan. Atotal of 6l pre-test interviews were com-

pleted. Ihese were distributed approxi-mately eqiially among the three urbemizedareas. A complete review of all questionsincluded in the pre-test was carried outwith differences in local interpretationnoted especially. Based on this reviewfinal revisions in the questionnaire weremade and specific Instructions to the In-

teinriewers prepared. A copy of the ques-tionnaire is included in this Appendix.

Sample Design

The sample design for this study wasconstructed with two basic requirements inmind. First, the sampling techniques em-ployed must be consistent with the demandsof sound research methodology; they must betechniques by which veLLld inferences may bedrawn from the sample for the populationgroup under investigation. Tbe only knownway to meet this requirement is throughprobability sampling. With probabilitysamples, the chance of observing a givenindivldusil or element of the population ofInterest is known. It permits the re-searcher to not only control the samplingareas, but also to measure them. It is

this property, the measureLbility of area,which lends validity to the conclusionsdrawn from probability samples.

Second, the sample design must beeconcffliically and statistically efficient;that is, it should, for the budget allottedand resources available, provide the mostaccurate estimates of the characteristicsstudied. The estimates derived from thesample must be of sufficient accuracy tobe used with confidence. Selection of themost efficient design implies knowledge ofthe sources of variation affecting a setof ssmple observations or measurements.The problem of sample design is to makethat judicious selection among the manytechniques available for controlling thesesources of variation, and hence theeventual sampling area, which will achievean appropriate balance between adminis-trative efficiency and statistical

hi

efficiency. The specific techniquesemployed in the sample designs constructedfor this study Include:

1. Grouping the eligible populationinto small clusters or samplingunits comprising an efficientinteJTviewer daily work load.

2. Grouping the sampling \inits

into city and subxirban zones,in each of the three urbjmlzedareas surveyed, to provideapproximately proportionaterepresentation

.

3. A further grouping of thesampling units within eachzone into geogi^phic or areastrata, with an equsLL numberof sampling units in eachstratum, to ensure adequatedistribution of the sample to

all segments of the populationof interest.

k. Using equal probabilities forthe selecting of sampling unitswithin strata and thereby con-

siderably simplifying the formu-las necessary for valid compu-tation of the estimates and oftheir standard errors.

A strict probability sample impliesthe application of completely objectivemethods for the selection of respondents.In the absence of a list of households orpersons eligible for interview, the re-

quired objectivity is met through the useof area probability sampling techniques.To be satisfied with simple area samplingtechniques is not enough, however. Inge-nuity in the use of available resourcesand facilities can considerably increasethe efficiency of one tirea probabilitysaniple over another.

United States Census Population andHousing data, both published and unpub-lished, are our major resource in thedesign of efficient probability samples.Unpublished data for small areas, such as

enumeration districts used in collectingcensus data, may be purchased on specialorder from the Bureau of the Census. In

open country areas maps indicating thelocation of dwelling units are available

from State Highway Commissions. Thissupplementary information may be used fora variety of purposes in the design of asample including stratification, assigningselection probabilities, or for the con-struction of approximately equal-sizedsampling units. The sample design outlinedbelow makes use of 1950 census data toestablish the area strata and for theassignment of the sampling units vithlnthese strata. Although these data were notused for the direct assignment of selectionprobabilities, the sampling plan adopted issuch that the chance for any segment of theareas surveyed, to be represented in thesample, was approximately proportionate tothe number of occupied dwelling units con-tained within the sequent whether it wasan enumeration district, census tract,township, urban place, city block, orportion of an enumeration district, etc.

A sample representative of all house-holds in the urbanized areas of Birmingham,Alabama; Boston, Massachusetts; Detroit,Michigan was selected for this study. Inaddition, a representative sample of cLll

non-'vrtiite households located in the ruralportion of Orangeburg County, SouthCarolina was chosen. Bureau of the Censusdefinitions of households, dwelling xinits,

urbanized eirea, rural territory etc. wereemployed. The sample designs for thethree urbanized areas will be describedfirst. These designs were stratified one-steige sample designs, constructed inaccordance with the principles outlinedabove. Careful control in all steps ofthe sample selection made it possible toknow exactly the chance every householdcluster or sampling unit had of fallinginto the sample.

The first step in the sample designconsisted of listing and ordering geo-graphically the census tracts in thecentred, city portions of each of the threeurbanized areas. In Detroit , those censustracts with 10 percent or more of thedwelling units occupied by non-white house-holds in 1950 were listed and orderedseptLrately. Similarly, ordered lists ofthe 1950 Census Enumeration Districts wereprepared for those portions of the threeurbajiized areas ^ich fall outside thecentral cities. Geographic or area stratawere then constructed within the centred,

city zones and the suburban zones for each

k8

of the urbanized areas using the orderedlists and 1950 census data on the number ofoccupied dwelling units or households foundIn each census tract, block or enumerationdistrict. These strata, seventy in numberfor each urbanized area, were constructedto contain approximately the same nvmiber

of households in each.

Each of the seventy strata was thendivided into a number of «™»n area seg-ments having boundaries which could easilybe identified In the field by the inter-viewers. Each such area segment conteiinedone or more clusters of households or sam-pling units. The number of sampling xinlts

or Interviewer work loads assigned to eacharea segment was based on data available onthe nxmiber of occupied dwelling xmits lo-cated within these segment boundaries.These data were obt«d.ned from a variety ofsources including 1950 block statistics,1950 enumeration district statistics, statehighway maps, etc.

The geographic strata in each citywere all constructed to contain the samenumber of sampling units with the exceptionof Detroit. In the cenla-al city portion ofDetroit, the area strata for the tracts inthe white zone (that is, the tracts with atleast 90 percent of their 195O dwellingvinits occupied by white families) were con-structed to contain twice as many samplingunits as the remaining area strata estab-lished for the Detroit urbemized areasample. Initially, two sampling units wereselected with equal probability and withoutreplacement from each of the geographicstrata, yielding a sample total of lUOsampling units for each urbemlzed area.

The sample selection was accomplishedby choosing two random numbers for eachstratum between one and the total number ofsampling units in the stratum. Thus, thesampling rate was the same for aill geo-graphic strata within a city with the ex-ception of those comprising the white zonein Detroit referred to above. Since thesestrata contained twice as msmy samplingunits as the remaining geographic stratain Detroit they were sampled at one-halfthe rate of the remaining strata in thaturbanized area. The disproportionate sam-pling in Detroit was deemed necessary toyield sufficient interviews with non-^rtilte

families for separate tabulation.

The number of strata and samplingunits for the central cities and the re-maining portions of the three urbemlzed£ureas axe shown in the following table:

Appendix Table 1

WJMBER OF STRATA AND SAMPLING UNITS OFURBANIZED AREAS INCLUDED IN THE

MOTIVATION SURVEY

to take through the segment for listingpiirposes. These lists were then returnedto the Philadelphia office of the A. J.

Wood Reseetrch Corporation where they werechecked. Next, the dwelling units on eachlist which were associated with the se-

lected sampling units were marked for inter-viewing. For example, if a given area seg-ment was assigned three sampling units andthe random selection had designated thesecond sampling unit, the list was firstdivided into three equal parts and then thedwelling units listed in the second of thethree parts were marked for interview.The few sampling vinits in each urbanizedarea ^rtiich contained more than 12 house-holds selected for interview were sub-saii5>led. The lists were then returned tothe interviewers for interviewing.

The intei-viewers were instructed tointerview the sample (marked) householdson the list and any other household (not

shown on the list) found between a samplehousehold and the next one listed. Thushouseholds which might have been omittedin the pre-listing were included; andchanges occurring after the pre-llstingwere accounted for. Interviews in the

sample households were conducted with the

person mainly responsible for planning the

meals. Where the person designated forinterview was not at home on the firstcall, succeeding calls up to a total ofthree were made on different days orevenings (in seme instances more th£in

three calls were made )

.

The sample design for the OrangeburgCovmty, South Ceirolina sample was similarin many respects. After preparing anordered list of the enumeration districtsfalling in the rural portion of the county,

sampling units were assigned to the enu-meration districts according to the numberof dwelling units occupied in 1950 by non-

irtiite households contained in each. Thesesampling units were then grouped into geo-graphic strata, 21 in total, with eachstratum containing 55 sampling units. Twosampling units were selected at randomwithout replacement from each stratum,

yielding a total of 42 sampling \inits forthe sample. Next, maps of each of the area

segments containing a selected samplingunit were prepared and the interviewerslisted all dwelling imits falling withinthe area segnent, classifying these

dwelling units according to whether theywere occupied by white households or non-white households or were vacant. Thelocation of each dwelling unit was markedon the segment map and numbered; this samenumber was used on the listing sheet.

Field Wor^

Training sessions with the super-visors and interviewers were conducted ineach of the survey aureas by members of thePhiladelphia office staff of the A. J.

Wood Research Corporation. Initial fieldwork was checked for quality and under-standing of the instructions. A copy ofthe intei^riewer instructions is also in-cluded in this Appendix.

In addition to the check of theinitial interviews, the area supervisorswere required to conduct a preliminaryedit of eill work turned in and to check10 percent of each interviewer's work bytelephone. A further verification checkon the field staff was carried out by thehome office by means of a check card mail-ing to 33 percent of the respondents ineach city.

A total of 2,385 households weredesignated for interview in this survey;

706 In Birmingham, 7k3 in Boston, 716 inDetroit and 220 in Orangeburg. Interviewswere completed in 1,9^7 of the samplehouseholds. The reasons for the non-inter-views are tabulated in Appendix Table 2.

Data Processing Procedures

All questionnaires were edited uponreceipt in the Philadelphia office andthose which were incomplete or containedquestionable responses were returned tothe field supervisors for re-interview.The coding department then prepared tabu-lations of the open-end questions from asample of the completed Interviews fromeach survey area. Codes for these ques-tions were established and coding Instruc-tions prepared and reproduced.

The questionnaire and coding proce-dures were explained and reviewed withthe coders. The open-end questions wereresearved for coding by the most experi-enced coders only. Answers to open-endquestions which were not readily

50

classified into specific code categorieswere held aside for review by the codingsupervisor and project director. Specificcategories for the latter cases were es-tablished when necessary.

The work of all coders was checked bythe coding supervisor until an acceptablelevel of coding consistency was achievedboth between and within coders. There-

after a 10 percent check for the purposeof maintaining this consistency level wascarried out.

The punch cards were then preparedand weighted as follows: In Detroit, theinterviews completed in sampling unitsselected from the irtiite zone were dupli-cated once since these interviewer assign-ments had one-half the probability of beingincluded in the sample as did the remainingsampling units chosen for this survey inthat city. In addition the punch cards forinterviews completed in assignments whichhad been subsampled were weighted accordingto the subsampling rates. No attempt wasmade to substitute or weight for householdsdesignated for the sample but not inter-viewed.

The punch cards then received athorough eiTor and consistency check on the

lEM Electronic Statistical machine. Wherenecessary the punch cards were correctedby reference to the specific questionnairescorresponding to the cards in question.

Sampling Errors

The sampling error for a parti culeu:estimate serves as a guide to the confi-dence with which this estimate ceui be used.

It is a measure of the closeness of thesample estimate to the result which wouldbe obtained from a complete census of thepopulation sampled, using the same ques-tionnaire, interviewp and interviewingprocedures.

Practically all of the estimatesdeveloped from the data collected in thisstudy are simple percentages of the re-spondents having a particular opinion orcharacteristic. In technical terminology,these percentsiges are actually combinedstrata ratio estimates, since the sampledesign employed extensive geographicstratification and cluster sampling, inwhich the number of respondents in eachcluster was subject to randcan samplingvariation. Thus, sampling errors werecomputed using the formula for thevariance of a ratio estimate.

This formula contains variance meas-ures for the cluster average of both thenumerator and denominator of the computedproportion or percentage estimate, as wellas a covariance measxire for these twoaverages. These measures were computedfrom the average variance between clusterswithin strata.

The chances are approximately 2 to 1that the error, due to sampling, in apsurticular estimate, will not exceed onestandard error; the chances are 19 to 1against a deviation as large as two stand-ard errors from the result which would beobtained with a ccmplete census using thesame procedures.

Estimates of the standard errors forseveral items included among the guidedassociation questions are shown inAppendix Table 3>

51

Appendix Table 3

SAMPLING ERRORS

Question

stratum No.

City

_3,'t Segment No.

7

QDESTIONNAIRE

5 Unit No. 6Bureau of the BudgetNo. 1*2-5901

Auth. Explrea June 30, 1959

1ihterview that person mainly responsible for planninc the meals served in the household

1

(BEAD EACH STATHfflNT

IWSERTING EACH VAHITrYOF CANNED FISH)

A-3

(READ EACH STATE^ENTINSERTING EACH VARIETYOF CANNED FISH)

D-1

PAHT D

SARDINE SECTION

A-1. Have you served canned sardines during the last 12 months?

\JF "NO," SKIP TO NOH-SARDINE USER SECTIOiTI

Yes-Bo--

-35-1- -2

During the past k weeks, about how often did you earve canned sardines?

fPld not serve -X

Don't know -Y

1 time 3

2 times -U

3 times -5It times -6

5 times or more( SPECIFY )

cISARDINE USERS ONLY I

Who In your family eats sardines.

IF NO SARDINES SERVED IN LAST It

W£EKS. SKIP TO NON-USER SECTION

Entire family-?-

or onlyHusband,Respondent,Children under 5>

Children 5 - 10,

Children 11 - 15,

Male children over 15,

—

Female children over 15,-Other adults?

-36-1

- -2

- -3

- -k

- -5

- -6

- -7

- -8

- z2.

Which preparation in which sardines are packeddo you usually buy . .

.

COTTONSEED AND SOYBEANOIL ARE VEGETABLE OILS

2a. Why do you use that kind?

Natural, 37-1Olive oU, 2

Vegetable oil, 3

Tomato sauce, -U

Mustard sauce, -5

Slid oil? 6

Other (SPECIFY)

3 . Do you sometimes use another kind?

I

IF "YES"I

3a. Which kind?

3b. Why do you use the other kind(8)?

Yes 38-1Bo— 2

Natural 3

Olive oil U

Vegetable oil 5

Toaato sauce -6

Mustard sauce -7

SUd oil ^Other (aPBCm)

U. Do you buy sardines which come In the largeor In the small can?

ISHOW CARD f1 TO RESPONdEotI

5 . Do you use only domestic or only Importedsardines or both?

nF~ONLY DOMESTIC OR IMPORTED |

lArge 1»0-1

S»ll 2

Imported -3

Domestic -U

"Both 5

58

5a. Why7_

[ASK QUESTION 5b IF DOMESTIC OR BOTH~1

5b. Where do the domestic eardlnes youbuy usually come from...

6. Do you think domestic sardines are better, aboutthe same, or not as good as imported sardines?

7. If domestic sardines had their bones andskin removed, would you...

8

.

Does the appearance of California sardinesplease you when you open the can?

I IF "HO^

8a. In what way do they not please you7_

Maine, 1»1-1

or California? -2

Better -3

About the same -kNot as good -5

eat more of them, -6

or eat about the sajne? -7

les -8

CM 9

9. Does the appearemce of Maine sardines pleas

»

you when you open the can?

I IF "NO" 1

9a. In what way do they not please you?

10. What would induce you to use more sardines?

11. Do you prefer a sardine can with or without a key?

11a. I IF WITH. KEY |How much more would you payfor a can with a key?

12 . Do you sometimes serve sardines for dinner?

[ IF "NO"I

12a. Why don't you serve them for dinner?

With U5-IWithout - -2

Ho more -3

1^ k

^ 5

3^ 6More (SPECIFY) _

Yes- U6-1Ho 2

13. The last tiae you bought sardines, did you plan to buythem before you went to the store; or did you decide onthem at the store?

Flannod- -i»7-l

- -2

59

D-3

IIF "IMPULSl: *]

13a. What made you decide to buy them?

PROBE

Ik. How many small sardines do you serve peradult for a snack.7

pg SERVED FOR MEALS, ASK 6. l5~l

1^5. How man;, small sardines do you serve per

adult for a meal?

16

.

How many small sardines would you liketo find In this size can?

ISHOW CARD #2 , k oz . CAJTI

17. Do you prefer large or small sardines?

17a. Why do you prefer them?

1 or 2 lib-l

3 or U -2

5 or 6 -j

7 or 8 -li

9 or 10--- -.,

I or 2-— -6

3 or U -7

5 or 6 -8

7 or 8 -9

9 or 10 -0

Less than 5 or 6 1*9-1

7 or 8 -2

9 or 10 -3

II or 12 -I4

More than 12 -5

Large -6

Small J_

50-

PROBE

18 . About how many inches long are the seirdines you like to buy?

19- If sardines were processed Into a pastespread, do you think you would use sucha product. .

.

ISHOW RESPONDEMT CARD #3~|

20. Which of these scenes on a sardine can is

more attractive to you?

21

.

Were sardines served in your home whenyou were a child?

22. Have you ever gotten sardine recipes or information from a.

on crackers? 5I-Ion sandwiches? -2

as an Hors d'oeuvre? -3

Other (SPECIFY)

Would not use -6

Fish sceneMarine scene-

Yes -

No--

Newspaper advertisement?--Magazlne advertisement?

—

Television advertisement?

-

Radio advertisement?Friend?

Other source (SPECIFY)

-52-1- -3- -5

- -7- -9

Newspaper food column?Magazine food column?

TV home service program?Radio home service program? -8

label on the can? -0

-6

23 . Have you ordered sardines In a publiceating place during the last two months?

Yes-No—

-53-1- -2

IIF "YES"

I

23a. How many times have you ordered sardinesin a public eating place in the last two

months ?

1 to 3 times 3

U to 6 times >*

7 to 9 times -5

10 to 12 times -6

Over 12 times --- -7

60

DJ*

23b. What kinds of dishes did you order?[SPBCIFY"|

23c.

23d.

Generally speaking, what day of the veek didyou order sardines in a public eating place?

IF MORE THAN ONE DAT MENTIOITO"CIBCLB ALL DAYS MENTIONED

Generally at what time did you order sardinesIn a public eating place?

IF MORE THAN ONE TIME MENTIONED,CIRCLE ALL TIMES MENTIONED

24. Besides yourself, has anyone eating with you orderedsardines In a public eating place In the past two months?

25.

21*a. Who was It?

Sardines packed In what oil do you considerthe highest quality? (READ LIST)

Monday ' 54 -1

Tuesday -2

Wednesday -3

Thursday-" -k

Friday 5

Saturday — -6

Sunday -7

Morning -8

Lunch -9

Afternoon -0

Dinner -X

Evening— ^Yes 55-1Ho -2

Spouse -3

Child -k

Friend or other relative -5

Olive oil 56-1Cottonseed oil -2

Soy-bean oil- -3

Slid oil -k

61

SAHDIHES - NON-USER SECTION

I

ASK OF PEOPLE WHO HAVE NOT SERVED SARDINES IN LAST 12 MONTHS]

1. Is there any special reason why you don't serve sardines?

2. Did you ever use sardines In the past?

IIF "YES"_|

2a. How long ago did you serve them?

2b. Why did you stop using sardines?

35-

-36-1

37-

rASkED OF PEOPLE WHO SERVED SARDINES IN PAST 12 MONTHS gg'.NOT IN PAST h WEEKS]

3 . Why do you use sardines so seldom?

-38^

It. Did you use sardines more often in the past?

he.. Why have you reduced the number of times you serve them?

]jes 39-1No -2

^UO^

[Yes- 1*1-1

No 2

Spouse -3

Male child -1*

Female child 5

Other (SPECIFY)

62

E-1

FART E

IASK ALL RESPONDEin'S |

1 . Have you seen or heard any advertising for canned tuna7

) IF "YES" I

la. Have you seen or heard It...

2. Have you seen or heard any advertising forcanned salmont

pjr~"YES"J

2a . Have you seen or heard It . .

.

3 . Have you seen or heard any advertisingfor canned sardines?

ROTATE i^JESTIOHS

1. 2 and 3

3a. Have you seen or heai^l it..

k . Have you seen or heard advertising forcanned shrimp?

I

IF "YES^

Ua . Have you seen or heard It . .

.

5

.

Do you use canned shrlmpT

I

IF "YES" I

5a. Do you use the veined or de-velnedvariety, or both?

IIF "BOTH"n

5aa. Which do you like better...

6. Which kind of canned fish do you Ilka best.

Yes 77-1Ho - -2

In magazines, -3

On radio, -it

On television, -5

In newslapers? -6

Yes 7Ho 8

In magazines, -9

On radio, --- -0

On television, -X

In nevspapers? ^Yes 78-1Ho -2

In magazines, -3

On radio, -k

On television, -5

Ta newsj>apers7 -6

Yes 7No— - -8

In magazines, -9

On radio, -0

On television, -X

In newspapers?-- -Y

[Yes 79-1No 2

Veined - 3

De-velned -h

fBoth— - 5

Veined,- -6

or de-velned? -7

Tuna, -8

Salmon, -9

Sardines, -0

Shrimp? ^

63

M.WITAL STATUS:Marr'.ed

SingleWidowed, divorced, etc.-

CUSSIFICATION DATA

5-70-1- -2

- li

la. How many people eat. dinner at homewith you'; (CIHCLE ONE)

123U56789 10or more

71-1 -2 -3 -it -5 -6 -7 -8 -9 ^I

IF NOT SINGLE, ASK QUESTIONS lb ASP Ic [

lb. How many children eat dinner at

home? (CIHCLE ONE)

1 23U56789 10or more

Ic . What are their approximate ages?

IIF MAHBIED, ASK QUESTION Id

|

Id . What Ij^ your husband ' s J ob ?

Executive, professional, mer-chant or own business— ' 72-1

Clerical or sales personnel -2

Manual skilled, semi-skilledor unskilled worker -3

Retired, unemployed, or student -k

Other (SPECIFY)

Yes-No--

Do you work?

IIF "YES

,

" ASK QUESTION 2a|

2a. What Is your Job?

Executive, professional, mer-chant or own business

Clerical or sales personnelManual skilled, semi-skilled

or unskilled workerRetired, unemployed, or student

Other (SPECIFY)

-73-1- -2

riiAJg) RESPONDENT CARD fT]

3. Would you tell me which letter indicates

the age you are? (CIRCLK ORE)

A B C D K

71* -1 -2 -3 -^ zl

fUXD RESPONDENT CARD #T]

k. Would you tell me into which group your

total family income falls? (CIRCLE ONE)

75-1

E

What was the last grade you completed in

school?Less than 8th 76-I8th through 12th-College-

I

ASKED ONLY OF COLORED HOUSEHOLDS|

How long have you lived in this city?

1 year or less

More than 1 to 5 years• More than 5 to 10 years

More than 10 to 15 yearsMore than I5 years

I

IF LESS THAN 1? YEARS , ASK QUESTION 6a|

6a. Where did you come from?

Northern state--Southern state --

Foreign country-

ASK ALL RESPONDENTS

Where were you born?

Northern state

Southern stateForeign country (SPECIFY)

-77-1- -2

IF NATIVE BORN

Were both of your parents Yes

born in this country? No

IIF "NO"

I

8a. In what country (s) were they born?

(SPECIFY)Father

-78-1- -2

Mother

9. Ten years ago did you live in.

open country,suburbs,city?

-79-1- -2

- -3

The religious backgroupd of a family at

times influences eating habits. With what

religion is your family most closely asso-

ciated?

Protestant -'*

Catholic -5

Jewish 6

Other (SPECIFY)

AUTOMATIC CLASSIFICATION

RACE: White 80-I

Negro -2

Other non-white -3

12. LOCATION OF DWELLING:City—

-

Suburb

-

-k

Clty_

Interviewer's Slgnature_Time Interview Coapleted:

A.M. ^P-M.

6k

65INT.OUP..D.C.60- 7*917

-69- -e'^- -tf3-

-tf^ <e- -ee- -y^-

O -(^ ^eft- -W3-

-ee- -ee- t^ -ee-

66

5 WHSE 00202

who buys canned sardines,...unitedstatesdepsurtmentoftheinterior,freda.seaton,secretary-...

Documents