University of Hawaii, Ph.D., -1969Economics, agricultural
WONG, Henry LiNan, 1940-AN EXAMINATION OF IMPULSE BUYINGOR IN-WE-STORE PURCHASE DECISIONSAS A CONSEQUENCE OF ~ STOREMERCHANDISING PRACTICES.
r- "------ ------- ------------------------------.--- - - -~ - ~~----- ---j
) This dissertation has been fmicrofilmed exactly as received 70-4320 I
II
\I;\
\t(
IIIL University ~~crof~S: Inc.• Ann Arbor. MIchigan
...._- - - _.--- .-...- _.'-' .,. - .-. _... _-- ..-_.- _.- ~--_.-
AN EXAMINATION OF IMPULSE BUYING OR
IN-THE-STORE PURCHASE DECISIONS AS A
CONSEQUENCE OF IN STORE MERCHANDISING PRACTICES
A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE
UNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN AGRICULTURAL ECONOMICS
JUNE 1969
By
Henry LiNan Wong
Dissertation Committee:
Arnold B. Larson, ChairmanJere R. BoyerBurnham O. CampbellJack R. DavidsonHeinz Spielmann
ii
ACKNOWLEDGEMENTS
The author wishes to acknowledge the cooperation of the management
of Star Supermarkets for providing the test store and the grateful thanks
to the 700 shoppers who allowed themselves to be interviewed.
The author acknowledges the U. S. Office of Education which provided
the NDEA Title IV Fellowship for the course of the study.
Sincere appreciation is due my wife Laurie, whose patience,
encouragement, understanding and typing of the drafts certainly
contributed to the completion of this research.
iii
ABSTRACT
Food retailing is an important segment of Hawaii's economy. The
national trend of consumer expenditures is toward an increasing number
of purchases without advance planning. The present study was based on
empirical information obtained from a group of 700 shoppers in a
representative supermarket in Honolulu, Hawaii, d·~:;L'.;lg a five-week test
period. It was hypothesized that consumers are aware of marketing
techniques; namely, multiple pricing and increased display space on
selected items. The primary objectives of this study were to find the
behavior patterns of supermarket shoppers in Hawaii and their awareness
of merchandising techniques and to ascertain how multiple pricing and
increased display space influence the shopper in Hawaii to purchase
unplanned selected items among the many that are avai1able~ and to
determine the impact these merchandisi~g practices have on sales. The
selections of the test items were based on a study conducted by Jere
Boyer, Food Distribution Specialist, entitled Producing Profits (A
Training Clinic on Retail Produce Department Operation). The selections
were also based on the sales of the supermarket chain and the fact that
these products were representative of Hawaii. The high demand item was
navel oranges; the semi-demand item was turnips; and the low demand item
was bell peppers.
Answers given to the following questions showed similarities between
shoppers in Hawaii and on the Mainland.
1. Who does the family shopping?
2. How many marketing trips are taken per week?
3. How many stores are frequented per week?
iv
4. How many food ads are read per week?
Some findings that showed a marked deviation from Mainland findings
were:
1. A greater number of Island shoppers used a written shopping
list.
2. More Island women shoppers were fully employed.
3. Household incomes of Island shoppers are higher than on the
Mainland.
4. Island shoppers come from more varied ethnic groups.
Aside from these few deviations, Island shoppers are very similar to
their counterparts on the Mainland.
The statistical test, analysis of variance, was used to see if
weekly volume movements were significantly different. The findings w~re:
1. 'The sale of navel oranges could be increased by increasing the
display space even at the rather high cost of 30¢ per pound.
2. The sale of turnips was influenced by multiple pricing.
3. Each of the variables, multiple pricing and increased display
space, increased the sale of bell peppers slightly.
The effectiveness of the merchandising techniques, multiple pricing
and increased display space was identified with the number of unplanned
purchases the marketing technique induced. The results were:
1. The greatest number of unplanned purchases of navel oranges
occurred in the week where the test variable was increased
display space (35.9 pe".cent of all purchases unplanned).
2. The greatest number of unplanned purchases of turnips occurred
in the week where the test variable was multiple pricing
v
(21.1 percent of all purchases unplanned).
3. The greatest number of unplanned purchases of bell peppers
occurred in the week where the test variable was multiple
pricing (44.7 percent of all purchases unplanned).
There was no indication that increased display space or multiple
pricing added significantly to the number of unplanned purchases.
It was concluded that the shoppers' awareness of price and display
changes is minimal at the ranges of prices and display spaces for navel
oranges, turnips, and bell peppers.
vi
TABLE OF CONTENTS
STATE OF THE PROBLEM
LIST OF TABLES
LIST OF ILLUSTRATIONS •
. . . . . . . . .
1
ii
ix
vii
iii
· . .· . .· . .
· . . . .. . . .
. .
. .. . . .
. . . . .
. . .
. . . .. . . . . . .
. . .. .
CHAPTER I.
PREFACE •••
ABSTRACT
. . . . . . . . .CHAPTER II. METHOD OF STUDY
CHAPTER III. RESEARCH ENVIRONMENT
. . · . . . . . . . 7
13
General Background of Hawaii's PopulationTest Store Environment • • • • • • • • • •
· .. . . . . . 1319
CHAPTER IV. SHOPPER CHARACTERISTICS AND BEHAVIOR.
Characteristics of ShoppersShopping Patterns •••••
. . . . . . . .· . . .
. . . . . .
· .
27
2734
CHAPTER V. THE IMPACT OF MULTIPLE PRICING AND INCREASEDDISPLAY SPACE ON SALES OF SELECTED PRODUCE ITEMS
Weekly Purchases Compared • • • • ••Cost Per-Unit Versus Cost Per-Pound
· . . .. .· .
· .· . .
47
6667
CHAPTER VI. UNPLANNED PURCHASES AND CONSUMERAWARENESS OF MERCHANDISING TECHNIQUES ••• · . 71
Unplanned Purchases ••••••••• • • • • • • • •Consumer Awareness of Applied Merchandising Techniques •Reasons for Purchases ••••••••••• • •Reasons for not Purchasing • • • • • • • • • • • • · . .
71768186
CHAPTER VII. SUMMARY AND IMPLICATIONS OF STUDY. · . 92
Suggestions for Further Research •
. . .· . . . . . . .
APPENDIX
BIBLIOGRAPHY
. . . . . . . .. . . . . . . . . . .
. . · . .. .. .
· . . . .· . .
94
96
157
vii
LIST OF TABLES
.Tab1e
I HOURS OF THE TEST STORE • · . . . . . 21
HOUSEHOLD INCOMES OF ALL SHOPPERS •
II
III
IV
ETHNIC CLASSIFICATIONS
AGE OF SHOPPERS • • • •
. . . .· . .
. . . . . .
. . . .
28
29
29
V
VI
PERCENTAGE OF SHOPPERS AND MEMBERS IN A HOUSEHOLD •
EDUCATION LEVEL OF SHOPPERS •
31
32
VII
VIII
OCCUPATION OF SHOPPERS •• •
WHO DOES THE FAMILY SHOPPING?
· . . . . . .. . . . . . .
33
35
IX WHO DOES THE FAMILY SHOPPING? MAINLAND AND HAWAIICOMPARISON •• • • • • • • • 35
X TYPES OF SHOPPING LIST •• • . . . . . . · . . • •• 36
XI
XII
XIII
HOW OFTEN DOES THE SHOPPER VISIT THE SUPERMARKET?
PERCENTAGE OF SUPERMARKET SHOPPERS SHOPPINGSPECIFIED NUMBERS OF TIMES PER WEEK • •
MAJOR SHOPPING DAYS • • • • • • • • •
. . . 37
38
39
XIV NUMBER OF STORES FREQUENTED PER WEEK . . . . . . . . . •• 40
xv
XVI
PERCENTAGE OF SUPERMARKET SHOPPERS PATRONIZINGMORE THAN ONE STORE PER WEEK • • • • • • • • • •
DISTANCE TRAVELED TO THE SUPERMARKET •••••
41
•• 42
XVII
XVIII
REASONS FOR PATRONIZING THE TEST STORE
NUMBER OF GROCERY ADS READ PER WEEK • •
. .. .
• •• 43
. • 44
XIX
xx
XXI
NUMBER OF SUPERMARKET ADVERTISEMENTS READPER AD-READING SHOPPER • • • • • • • • •
"DO TRADING STAMPS CAUSE YOU TO BUY MORE?" ••
TEST SCHEDULE • • • • • • • • • • • • • • • •
· . . .· . . .
44
45
•• 48
XXII DAILY VOLUME MOVEMENTS (IN LBS.) OF NAVEL ORANGES, TURNISP,AND BELL PEPPERS FOR THE DURATION OF THE EXPERIMENT • • •• 50
Table
XXIII
XXIV
xxv
XXVI
ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSU'ICATION, DAYS AND WEEKS, FOR NAVEL ORANGES
ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSIFICATION, DAYS AND WEEKS, FOR TURNIPS •••
ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSIFICATION, DAYS AND WEEKS, FOR BELL PEPPERS
SIGNIFICANCE TESTS FOR NAVEL ORANGES,TURNIPS, AND BELL PEPPERS • • • • • •
· . . .
· . . .
viii
58
61
63
65
XXVII
XXVIII
REASONS GIVEN FOR BUYING BELL PEPPERSPRICED AT 12¢ APIECE • • • •• • • •
REASONS GIVEN FOR BUYING BELL PEPPERSPRICED AT 49¢/LB. • • • • • • • • • •
· . . . . . . . . .
· . . . . . . . . .
68
69
XXIX UNPLANNED PURCHASES AS A PERCENTAGE OF TOTAL PURCHASES • 72
xxx ,
XXXI
UNPLANNED PURCHASES AS A PERCENT OFALL PURCHASES, BY AGE GROUP • • • • • • • • • • •
UNPLANNED PURCHASES AS A PERCENT OF ALLPURCHASES AND THE MAJOR OR FILL-IN SHOPPING TRIP
· . . .
· . . .
74
75
· . . .XXXIII TEST SCHEDULE • • • • • • • • • •
XXXII UNPLANNED PURCHASES OR A PERCENT OF TOTALPURCHASES BY WHO DOES THE FAMILY SHOPPING • . . .
. .77
79
XXXIV
xxxv
SHOPPERS' AWARENESS OF DISPLAYS •••
REASONS FOR PURCHASING NAVEL ORANGES
· . . .· . . . . .
80
82
XXXVI REASONS FOR PURCHASING TURNIPS . . . . . . . . . . . . . 84
XXXVII REASONS FOR PURCHASING BELL PEPPERS •
XXXVIII REASONS FOR NOT PURCHASING NAVEL ORANGES
. . . . .. . .
85
87
XXXIX REASONS FOR NOT PURCHASING TURNIPS • • • • • 89
XL REASONS FOR NOT PURCHASING BELL PEPPERS • · . . . · . . . 90
. . . . . . . . . .
Figure
1
2
3
LIST OF ILLUSTRATIONS
PERCENTAGE DISTRIBUTION OF POPULATION OFHAWAII BY ETHNIC ORIGIN, 1946-1965 •••
AGE COMPOSITION OF POPULATION OF HAWAIIBETWEEN 1900 AND 1960 • • • • • • • •
PER CAPITA INCOME FOR HAWAII AND U.S.BETWEEN 1948 AND 1965 • • • • • • • •
. . . . . . . . .
. . . . .
ix
14
16
17
4 INCOME DISTRIBUTION IN THE STATE OF HAWAII: 1946-1965 18
5
6
MAP OF CENTRAL HONOLULU
LAYOUT OF THE TEST STORE •
· .. . . . . . . . . . . . .
20
22
7
8
9
10
INDEX OF DAILY SALES RECEIPTS OF TEST STORE FORTHE DURATION OF THE EXPERIMENT • • • • • • • • •
INDEX OF DAILY CUSTOMER COUNT OF TEST STORE FORTHE DURATION OF THE EXPERIMENT • • • • • • • • •
INDEX OF DAILY TOTAL PRODUCE SALES OF TEST STOREFOR THE DURATION OF THE EXPERIMENT •
MEMBERS IN A HOUSEHOLD • • • • • • • • • • •
· . . . .· . . . .
· . . . .
24
25
26
30
11 POUNDS OF NAVEL ORANGES SOLD DAILY FORDURATION OF EXPERIMENT • • • • • • • • • . . . . . . ., . . 51
12
13
14
POUNDS OF TURNIPS SOLD FOR DURATION OF EXPERIMENT
POUNDS OF BELL PEPPERS SOLD FOR DURATION OF EXPERIMENT
COMPARISON OF TWO METHODS OF SELLING BELL PEPPERS ••
. . 52
53
54
CHAPTER I
STATEMENT OF THE PROBLEM
The trend of consumer expenditures is toward an increasing number
of purchases without advance planning. Between 1945 and 1959, the
unplanned purchases as a percent of total purchases in supermarkets rose
from 38.2 percent to 50 percent. l Impulse products anq,·their respective,
placements within a supermarket have been shown to greatly affect profits.
The interior of the store should be conducive to aiding the shoppers in
deciding their purchases. The use of such merchandising strategy tends
to increase food sales with very little additional costs.
Impulse buying is affected by factors such as economy, personality,
time, location and culture. Interestingly, these factors may vary from
shopper to shopper with respect to the purchase of the same item. Also
these factors may vary for the same shopper buying the same item under
different buying situations. This results in the delineation of four
broad classifications of impulse buying.
1. Pure Impulse Buying. This is truly impulse buying; the novelty
or escape purchase which breaks a normal buying pattern.
2. Reminder Impulse Buying. Reminder impulse buying occurs when
a shopper sees an item and remembers that the stock at home is
exhausted or low, or recalls advertisement or other information
about the item and a previous desire to buy. The key factor is
remembered prior experience with the product, or knowledge of
lHawkins Stern, "The Significance of Impulse Buying Today," Journalof Marketing, XXVI (April, 1962), 59-62.
2
it, which "sparks" the impulse purchase.
3. Suggestion Impulse Buying. Suggestion buying occurs when a
shopper sees a product for the first time and visualizes a need
for it, even though she has no previous knowledge of the item.
Suggestion buying is distinguished from reminder buying in that
the shopper has no prior knowledge of the product to assist her
in the purchase. Product quality, function, and the like must
be evaluated at the point of sale. The distinction between
suggestion buying and pure impulse buying is that items
purchased on suggestion impulse can also be entirely rational
or functional purchases, as opposed to the emotional appeal
which sparks pure impulse purchases.
4. Planned Impulse Buying. Although "planned impulse buying" may
seem anomalous, it is accurate. Planned impulse buying occurs
when the shopper enters the store with some specific purchase
in mind, but with the expectation and intention to make other
purchases that depend on price specials, coupon offers, and the
like. It is a recently developed consumer buying trait and
likely to be a most significant one. 2
It seems that shoppers enter the store with a general intention to
buy, but they make the actual buying decision at the point of purchase.
The growth of self-service food stores apparently has developed this
kind of planned impulse buying because the shopper can explore, compare
and arrive at a decision unhindered by a sales clerk. The use of
2Stern , ~. £!l., pp. 59-60.
3
advertising and in-store displays provide the shopper with product
information so the shopper can make a decision. Probably the most
important way in which the individual supermarket can adjust to the
increased practice of impulse buying is to induce the shopper to use the
store itself as a substitute for a personal food shopping list.
Especially where there is weak store loyalty, the shopper should be
encouraged to make her impulse purchases in that particular store,
thereby increasing store sales.
Previous studies of consumer behavior have dealt mostly with
planned purchases. Although unplanned purchasing, or impulse buying, is
an extremely important factor in relation to modern day merchandising
and promotional efforts, only limited research has been advanced in this
area. A series of studies of impulse buying was done by E. I. DuPont
DeN~mours and Company, "Consumer Buying Habits Studies," for 1945, 1949,
1954, and 1964. Some of the more interesting results were:
1. Nearly 7 of 10 purchases resulted from a decision made within
the store.
2. Sixty-three percent of the shoppers had no written shopping
lists.
3. The shopper spent an average of 26.1 minutes in a supermarket
per shopping trip. (This means that in a supermarket with
6,500 separate items the individual items have only one-fourth
of a second to catch the shopper's eye.)
4. Chewing gum (family pack) was the item that was most
"impulsive." (Eighty-nine percent were unplanned
4
purchases.) 3
Behavior patterns of supermarket shoppers have been changing with
the times. Mainland studies like Schapker's "Behavior Patterns of
Supermarket Shoppers" have tried to explain changing behavior patterns,
probing into questions such as: 1) What are the reasons for the decline
in supermarket shopping loyalty during the past decade? 2) How many
times per week does the supermarket customer shop? 3) Why is readership
of foods tore advertising in newspapers increasing? 4) Is interest in
saving trading stamps on the wane?4
The general attractiveness of the store environment, including the
aisles, air-conditioning, good lighting, and cleanliness, while it
undoubtedly contributes to patronage for planned purchases, probably has
a particularly great effect on the nature and level of unplanned
purchases. Store layout, through its effect on traffic flows, and
placement of displays also have a profound effect on impulse buying. It
is generally supposed that impulse buying of a particular item can be
increased by devoting a larger area of display space to the item, by
multiple pricing as opposed to pricing per pound, or by pricing on a
unit cost rather than a per pound basis. The effect of each of these
presumed factors is probably influenced by personal characteristics of
the shoppers, especially age, household income, sex, employment status,
store loyalty, knowledgeability with respect to food prices, and use of
3"Today , s Purchases in Supermarkets," E. 1. DuPont DeNemours andCo., The 7th DuPont Consumer Buying Habits Study, 1964.
4Ben L. Schapker, "Behavior Patterns of Supermarket Shoppers,"Journal of Marketing, XXX (Oct., 1966), 46-49.
5
a shopping list.
All previous studies of impulse buying were cOndtlcted on the
Mainland. The relative importance of impulse buying, and the shopper's
response to merchandising practices designed to increase such buying,
might well be different in Hawaii, especially if shoppers differ from
their Mainland counterparts in relevant personal characteristics.
Previously, the effects of multiple pricing and increased display
space on sales and consumer response were studied separately. This study
is a different approach because there was an attempt to link the two
types of studies together. The experiment was designed in such a way
that a consumer questionnaire was used as a vehicle for feedback of the
consumer's activities in relation to the experiment. Thus, the method
combined shopper interviews of awareness of marketing practices with
emphasis on unplanned purchases and the analysis of variance to compare
differences occurring between the control week and the test weeks.
The problem of the researcher was to find the behavior patterns of
supermarket shoppers in Hawaii and their awareness of merchandising
techniques. An additional problem was to ascertain how multiple pricing
and increased display space influence the shopper in Hawaii to purchase
unplanned selected items among the many that are available, and the
impact these merchandising practices have on sales. Related to this
was a study of the effect of unit cost versus per pound cost of a
selected item.
It has frequently been assumed that when the sales of an item had
been increased by multiple pricing and increased display space, the
consumers were aware of this and adjusted their shopping activities
6
accordingly. This assumption was tested. The hypothesis is that
consluners are aware of merchandising techniques; namely, multiple pricing
and increased display space of selected items.
CHAPTER II
METHOD OF STUDY
Typically, food retailers expect that merchandising practices like
multiple pricing and increased display space applied to an item can
increase sales by encouraging an increase in the number of impulse
purchases. This assumption is based on past experiences and previous
studies substantiating this. Food retailers are also cognizant of the
effects of competing items to the test item. Changing any aspect of
the display of the competing items could also be a cause for a change in
the level of sales of the test item. Likewise, seasonal demand for
another product could affect sales of the test item, thus making it
increasingly difficult to obtain an accurate measurement of the effects
of the variable being tested. Basically, any change in the
merchandising of an item can affect many other items in the store. In
light of this, one can see the near impossibility of designing an
experiment for testing one or two variab1es,and at the same time,
eliminating all other inFerferLng'factors.
The purpose of the study was not only to determine how multiple
pricing and increased display space would affect sales but to see
whether consumers would be aware that these merchandising techniques
were directly causing them to purchase these items. Thus a controlled
experimental design coupled with the use of a consumer questionnaire
was selected as the most effective method to cope with the problem.
In this way the difficult task of isolating all other factors was not
necessary for the success of the study. Having normal proceedings of
the store was deemed adequate since the consumer questionnaire was used
8
as a reinforcement for testing the effects of the test variable.
The Moiliili store of the local supermarket chain was selected as
the test store. The fact that only one store was selected was due not
only to the commercial restraints placed by the chain involved but also
to the financial and physical limitations of the researcher. However,
it was felt that the selected store was representative of Hawaii in
terms of location, education, income, and cultural background.
The study called for testing the effect of multiple pricing and
increased display space on a demand, semi~demand and low demand produce
item. A demand item is a fast moving sales item. A low demand item is
a slow moving sales item and a semi~demand item is a sales item that
falls between the two limits. Ideally, different food products should
be tested. A number of limiting factors prevented many products from
"being chosen for this test. Products where price specials were used had
to be discarded. Another group of products where space was limited was
rejected. The management rejected any ideas where a change in the
store's image would have occurred. The allocation of pricing and
increased display space of the test items had to conform to the
management's guidelines.
Since only a limited number of produce items could be tested, a
high demand, semi-demand and low demand item were chosen in an effort to
obtain different kinds of commodities and to see if multiple pricing and
increased display space would affect these items any differently.
Although the experiment was limited to only three items, for which valid
results were found, it i~ felt that these results suggest the kind of
reactions the consumer has formed in regards to merchandising practices
9
for other similar items. Thus the fact that a change in sales can be
made on these items suggests that future study on other items using the
same method could be beneficial. So although one cannot generalize, for
instance, that the same results obtained for the high demand product
tested will be obtained when testing another high-demand item, it can
suggest a possible trend in that direction.
From a study made in Hawaii by Mr. Jere Boyer, Food Distribution
Specialist, navel oranges were found to be demand items because they
accounted for over 8 percent of all produce sales. Bell peppers were
low demand items for they accounted for 1.5 percent of all produce sales.
A semi-demand item was turnips, sales of which accounted for a percentage
between navel oranges and bell peppers.* The selection of these test
items for the study was based on both Boyer's study and the sales of
the selected supermarket chain in Hawaii. The three selected produce
items thus were navel oranges, turnips, and bell peppers.
The management of the supermarket agreed that the three designated
test products would not be advertised during the duration of the
experiment.
The data was collected through personal interviews. The
questionnaire consisted of forty-two questions and took an average of
six minutes to comp1ete~ (The questionnaire can be found in the
Appendix.) The survey took five weeks at the designated store. During
the test period all other factors were held constant through a
combination of experimental and statistical controls. Sales and volume
*Study entitled Producing Profits (A Training Clinic on RetailProduce Department Operation).
10
responses to changes in the display space and multiple pricing for the
three designated products were measured. Inventories of navel oranges,
turnips and bell peppers were recorded daily. With the cooperation of
the store manager the following schedule was carried out:
1st Week (Control): Regular Pricing and Given Display Space
2nd Week: Regular Pricing and Increased Display Space
3rd Week: Multiple Pricing and Regular Display Space
4th Week: Multiple Pricing and Increased Display Space
5th Week: Cost Per Pound Display Versus Cost Per Unit Display
During the five weeks of testing, shoppers were asked a series of
questions pertaining to the purchase of the three designated products.
A series of questions dealing with personal data of the customer was
also asked. Non-buyers of the three designated products were also
interviewed. They were selected at random at a designated spot in the
store. The only requirement was that they must have shopped at the
produce department. Shoppers who bought anyone of the three products
were interviewed. The shoppers were classified into the following
groups: 1) Buyers of navel oranges; 2) Buyers of turnips; 3) Buyers of
bell peppers; and 4) Non-buyers of these three products. (The study
revealed that there were only a few buyers who bought a combination of
these items.) Hence, there were four classifications and each contained
five interviews daily. This meant that a week's survey yielded 140
interviews. This scheme was followed for four weeks. Thus 560
interviews were obtained. The fifth week 140 interviews were obtained
in an experiment dealing with bell peppers only. Consequently, 700
interviews were obtained.
TEST SCHEDULE
Test Period Price Display Space
1st week (Control)Navel Oranges 30¢/lb. 54"Turnips 19¢/lb. 19"Bell Peppers 59¢/lb. 19"
2nd weekNavel Oranges 3 1bs. for 89¢ 54"Turnips 2 1bs. for 39¢ 19"Bell Peppers 4 pes. for 59¢ 19"
3rd weekNavel Oranges 30¢/lb. 86"Turnips 19¢/lb. 24"Bell Peppers 59¢/lb. 24"
4th weekNavel Oranges 3 1bs. for 89¢ f?5"Turnips 2 1bs. for 39¢ 24"Bell Peppers 4 pes. for 59¢ 24"
5th weekBell Peppers 12¢ apiece 15"Bell Peppers 49¢/lb. 15"
11
12
During the first week the norms of sales and volume movements of
navel oranges, turnips, and bell peppers were determined. During the
following weeks variables were changed. It was therefore assumed that
additional sales and volume increases were a result of these variables.
The data gathered during the test period shows the various effects that
multiple pricing and increased display space had on sales and volume
movements of a demand, semi-demand, and low-demand item.
The statistical test, analysis of variance, was used to determine
if weekly volume movements were significantly different. The chi-square
test was used to see if the number of unplanned and planned purchases
of the three selected produce items in the test week was significantly
different from the control week.
The results of the questionnaire were analyzed to give additional
information to help explain the effects of the test variables on sales
of a demand, semi-demand and low demand item.
CHAPTER III
RESEARCH ENVIRONMENT
It was important to understand the other factors involved in the
study before undertaking the actual experiment. Thus, information on the
general background of Hawaii's population, the test store environment and
the test store itself was obtained. A clear understanding of the
existing situation contributes to a better understanding of this study's
findings.
GENERAL BACKGROUND OF HAWAII'S POPULATION
The unique characteristics and location of Hawaii probably affect
consumer behavior and thus might influence the results of this study.
It was necessary to look at some of the differences which set Hawaii
apart from other states. By observing some statewide characteristics of
income and age of the population some guidelines were developed which
were used to make judgment as to whether or not the characteristics of
shoppers interviewed were representative of Hawaii's general population.
One difference from most other states is that in Hawaii there are
many ethnic groups. Figure 1 shows the percentage distribution of
Hawaii's population by ethnic origin, from 1946 to 1965. One can
readily observe that the greatest increase of population in the last
eight years has been among the Caucasians. The percentage of Japanese
has declined slightly over the same number of years. In 1965 Caucasians
accounted for 37 percent of the total population while Japanese
accounted for 29 percent of the total. The remaining population consists
---- - - ---.-.
----
........ - ---.......
Hawaiian &Part Hawaiian
-
.............
-----Japanese
Filipino
-------
--
Caucasian
/---.........
/
20
30
40
~CJJ.l(\l~ ,---- ......
.--..--------. -. -- -.---._._.-._- ·-·-e----- ~ .._........----.......10
/0-0 - 0 0 _0-0-0 _ 0 / Chinese 0_ _o ----0""---0-0
Others
196519601955Year
1950i I I I I i I I I I01) iii Iii iii.
1946
FIGURE 1. PERCENTAGE DISTRIBUTION OF POPULATION OF HAWAII BY ETHNIC ORIGIN, 1946-1965
Source: U.S. Department of Commerce, U.S. Census of Population, 1960, Hawaii, General PopulationCharacteristics, Washington, D. C.
I-'~
15
of 6 percent Chinese, 10 percent Filipino, 15 percent Hawaiian and
part-Hawaiian and 3 percent others.
The population in Hawaii is unique in both numerical size and growth
rate of different etllnic groups.
The age distribution of the population and the income distribution
provided information pertinent to this study. Figure 2 shows the age
composition of the population of Hawaii between 1900 and 1960; it also
shows that the population of Hawaii tends to be on the young side.
Figure 3 shows the rising per capita income in the United States and
Hawaii from 1948 to 1965. Between 1959 and 1965 the per capita income
in Hawaii exceeded that of the U. S. Mainland.
Figure 4 shows the income distribution in the State of Hawaii from
1946 to 1965. It is interesting to note the increasing percentage of
the population at the higher levels of income.
The greatest part of the state population is centered on Oahu where
there are approximately 386 retail grocery stores, according to the
University of Hawaii Extension Service. The major supermarkets which
engage in food retailing are:
1. Foodland Super Market, Ltd. (16)
2. Times Super Market, Ltd. (7)
3. Star Super Market (4)
4. Safeway Stores, Inc. (4) - National chain
5. P & P Super Foods (2)
6. Chun Hoon Super Market (2)
7. Holiday Mart (2)
8. GEM (2)
60
50
40
30
20
10
16
24 years and under
25 to 34 years
-- ..........- ...........~............ --
~ _.-.-., ~ ------.--
35 to 44 years
1900 1910 1920 1930Years
1940 1950 1960
FIGURE 2. AGE COMPOSITION OF POPULATION OF HAWAII BETWEEN 1900 AND 1960
Source: Hawaii Department of Planning and Economic Development.
$3000
./,/
",Hawaii ".",
"-"-",/$2500 -___ _ u.s.
I U$2000 ~ _-
.scd+Jo.-fg-
O
~ ~1500..........------
$1000
1948 9 50 1 2 3 4 5 6 7Years
8 9 60 1 2 3 4 1965
FIGURE 3. PER CAPITA INCOME FOR HAWAII AND U. S"o BETWEEN 1948 AND 1965
Source: U.S. Department of Commerce, Survey of Current Business.~....,
60
---.... 0-/--o~ -.......
.."....-- -.._......._ - - -$i5':000-$99,999
~-- -o_ _0 _ .-.-' """"---- 0 0 _ 0 --~--- -' ~
$3,000-$4,999
$10,000-$14,999
_0
Under $1,000
·v·\0---·'0$1,000-$2,000 ~~
jt\ ~~o
, \/~--- --------...."'-'-~ "------..
-- ...---.. - - ...... -- -- - - - - - -----oJ): -; - iii : i :: iii iii iii iii
1946 1950 1955 1960 1965Year
10
20
50
40
~
I::~ 30$.IQ)~
Source:
FIGURE 4. INCOME DISTRIBUTION IN THE STATE OF HAWAII: 1946-1965....a) U.S. Internal Revenue Service, Statistics of Income, Washington, D.C. co
b) Hawaii Department of Planning and Economic Development, The State of Hawaii Data Book, Honolulu,1967.
19
9. Big Way (2)
10. IGA Stores (10) This is made up of ten independent stores.
The number of stores in each chain is in parentheses.
TEST STORE ENVIRONMENT
The following map shows the location of the test store -- the shaded
triangle -- with three concentric circles drawn around it. The smallest
circle encompasses the area and population that is within a one-mile
radius of the test store; the population within this area is estimated at
47,220. 1 The next largest circle encompasses the area and population
within a two-mile radius of the test store; the population within this
area is estimated to be around 142,200. 2 The largest circle encompasses
all the area and population within a three-mile radius of the test
store. This is the fringe area with respect to the drawing power of
the test store.
Within the one-mile radius of the test store there are five food
stores of similar size. Along with these are a number of smaller,
family-run food stores. The competition becomes greater as we consider
all the stores within a two-mile radius of the test store. Within the
three-mile radius, competition would come from competing food stores and
also from a member of the test store chain.
1Estimated Civilian Population and Dwelling Units in Oahu CensusTracts, Hawaii State Department of Health (July 1,1967),5-7.
FIGURE 5. MAP OF CENTRAL HONOLULU
No
21
The test store has a total selling area of approximately 13,800
square feet. (This includes the liquor department.) All departments
within the store are operated on the principle of self-service with the
exception of the fresh fish section. The meat department affords the
customer the option of ordering special cuts. The store also gives
trading stamps as part of its services.
TABLE I. HOURS OF THE TEST STORE
Sunday 9:00 AM to 5:00 PM
Monday 9:00 AM to 8:00 PM
Tuesday 9:00 AM to 9:00 PM
Wednesday 9:00 AM to 8:00 PM
Thursday 9:00 AM to 9:00 PM
Friday 9:00 AM to 9:00 PM
Saturday 9:00 AM to 7:30 PM
A layout of the test store follows. The location of the navel orange,
turnip, and bell pepper displays are also shown. The displays were
placed in their usual locations. The shoppers in their normal shopping
patterns had the opportunity to see each display once. The displays
were made to look as normal as possible; or in other words, similar to
the store's usual displays. Problems would arise if the displays were
"featured" or if they were end displays, or multiple exposures. The
factors influencing the shopper would no longer be multiple pricing and
Soda i Beverage iPackaged: Meats:
~ I Produce f ~ IBell Peppers Turnips t::::l
I rroduc~ I I profuce U __@ ~Navel
Oranges
MilkButterE
t:j
Iiipo
I II II 1 9 en
(1) t"4r't
I0 .... 0
I; ..0Ii
c::(1)
0 t:j
r'tIi I'i
en g
CJr't
I I
CJ .t::::l
(&)0
I0
IIi
rozenPies
Tea-Coffee
Home ! Cleaners
krozen Jnicl Produce : Ice Cream ~
U iT:;~::::r;~o~s 0
U ; Canned GOOf Soap D
U ?reakfast fofdS DBread Candy
11=;:*1: JJi ce canneJ ~~~~~e~UiCJ]
IMacaro:~g: Food : Rice a
Frozen : Poul lory
Frozen.
Me ~tsI
Icatsup! tSoy sau~Salad Dressing
I Soup h i Sugar aaby Food
ICanned {Meat & FIShJams ~
s:.lPlUr'tlD
FIGURE 6. LAYOUT OF THE TEST STORENN
23
increased display space and it would be difficult to isolate the effects
of these variables from others that arise in these situations.
An account of the store's promotional and merchandising activities
was recorded daily for the purpose of identifying any deviation from the
normal functioning of the store. These included advertised specials and
any interesting or unusual events of the store that could affect the
outcome of the research.
Figure 7 shows the index of daily total receipts for the duration
of the five week study. There does not appear to be any dras tic change
from week to week.
Figure 8 shows the index of daily customer count for the length of
the experiment. The week-to-week changes were rather uniform.
Figure 9 shows the index of daily total produce sales. Again the
week-to-week changes were small.
These three figures show that the chosen time period of the study
at the test store was fairly typical of the store's normal activities.
Within the normal range of drawing power of this store there are
many different types of residential areas, including apartment houses,
and low and high income single-family dwellings, and the store is quite
accessible to commuter traffic. Hence it is judged to cater to a broad
cross-section of food shoppers in Honolulu. From the number of retail
grocery stores in the near vicinity, it would appear to be subjected to
strong competition. By and large, the store seemed well suited for use
as the tes t store.
M
~ 100 J.I----I--+----F---\----,r--~:__-_t~r_-,'---.s
SMTWTFSSMTW TFSSMTWTFS SMTW TFSSMT WTFS
FIGURE 7. INDEX OF DAILY SALES RECEIPTS OF TEST STORE FORTHE DURATION OF THE EXPERIMENT
Source: Daily Records of Test Store.
24
M~ 100 H-----J~_+---~~--__J.-_\_--I__\_--.J_-.s
SM TWTFS SMTWTFS SM TWTFSSMT WTFSSMTWTFS
FIGURE 8. INDEX OF DAILY CUSTOMER COUNT OF TEST STORE FORTHE DURATION OF THE EXPERIMENT
Source: Daily Records of Test Store.
25
CHAPTER IV
SHOPPER cr~CTERISTICS AND BEHAVIOR
This chapter is divided into two parts. The first part deals with
the characteristics of the interviewed shoppers to see if they were
representative of Hawaii. Also, the results were compared to mainland
studies whenever possible to determine the similarities if any existed.
The second part deals with the shopping patterns exhibited by the
interviewed shoppers. Again, the results were compared to mainland
studies whenever possible. The reason for comparison was to see how
similar shopping patterns on the Mainland are to Hawaii and thus, how
valid past and future Mainland studies are for the State of Hawaii.
CHARACTERISTICS OF SHOPPERS
Ethnic Classifications:
The ethnic classifications of the seven hundred shoppers
interviewed appeared to be fairly consistent with the general make-up
of the statewide population. In this study it was found that most
shoppers were either Japanese or Caucasian. They numbered 341 and 265
respectively. The combination of these two classifications resulted in
86.6 percent of the 700 shoppers interviewed. Other ethnic groups,
however, are somewhat under-represented. The breakdown is shown in
Table II.
TABLE II. ETHNIC CLASSIFICATIONS
Number Percentage
Japanese 341 48.7
Caucasian 265 37.9
Chinese 36 5.1
Filipino 18 2.6
Hawaiian 8 1.1
Korean 6 .9
Portugese 5 .7
Other 21 3.0
The age classification of shoppers interviewed revealed that
nearly five out of ten shoppers were in the age group of 20-34 years
old. The average age of the 700 shoppers was 37.2 years. Mainland
studies show that over half the shoppers are between 30 and 49 years
of age. 1 The percentage of shoppers interviewed in each age
classification is as follows:
l"The Family Shopper," E. 1. DuPont DeNemours and Co., The 7thDuPont Consumer Buying Study (1964), p. 1.
28
29
TABLE III. AGE OF SHOPPERS
Percent
Under 20 years 1.9
20-34 years old 48.0
35-49 years old 34.4
50-64 years old 11.6
Over 65 years 4.1
The average household income of the shoppers interviewed was
$11,338 with a standard deviation of $2,599.50. The average household
income would have been higher except for the low student incomes. Table
IV shows the percentage of shoppers in eaCh household income level.
TABLE IV. HOUSEHOLD INCOMES OF ALL SHOPPERS
PercentageHousehold Incomes of Shoppers
$4,000 or under 12.3
$4,000 to $6,999 6.1
$7,000 to $9,999 19.6
$10,000 to $12,999 28.7
$13,000 to $15,999 15.0
$16,000 to $19,999 6.4
$20,000 and over 11.9
225
200
175
(/)'t:l 150...-I0
-m(/) 125::l0=
4-1 1000
I-lQl 75
~50
25
I • .1 2 3 4 5 6 7 8 9 or more
FIGURE 10. MEMBERS IN A HOUSEHOLD
30
31
One reason for higher household incomes in Hawaii than on the
Mainland is that there are more full-time employed members in the
household. The Mainland findings show that over half of the families
have yearly incomes of more than $5,000 and approximately one out of
four families have incomes in excess of $7,500. 2 Family incomes have
risen nationally since this study was published in 1964.
In the classification of members in a household, the interviews of
shoppers disclosed that there was an average of 3.17 persons per house-
hold. Figure 10 shows the number of households and the members in a
household.
The percentages of shoppers in each classification of numbers of
members in a household are shown in Table V.
TABLE V. PERCENTAGE OF SHOPPERS AND MEMBERS IN A HOUSEHOLD
Members in a PercentageHousehold of Shoppers
One 7.7
Two 25.7
Three 31.3
Four 19.5
Five 12.0
Six :.9
Seven 1.0
Eight .3
Nine or more .1
ZIIThe Family Shopper, II .2£.• .ill., p. 1.
32
Mainland studies reveal that more than half of the shoppers buy for
three persons or less. On the other hand t only one-fourth shop for
households of five or more. 3 Table V shows that more than half of the
shoppers interviewed shop for three or less. Only 16.3 percent shop for
households of five or more. Thus t sizes of households in Hawaii and on
the Mainland are quite similar.
The educational level of the shoppers interviewed showed that on
the average they have had at least one year of college. This was higher
than it would normally be 'because of a significant number of college
students interviewed. The percentages were as follows:
TABLE VI. EDUCATION LEVEL OF SHOPPERS
Percent
1-7 years
Finished grade school
1-3 years high scheol
Finished high school
1-3 years college
Finished college
Beyond ccl1ege
3 11The Family Shopper t II .2E.. cit. t p. 1.
3.5
4.3
3.6
39.8
20.4
16.4
12.0
33
The percentages of all shoppers interviewed by occupation were as
follows:
TABLE VII. OCCUPATION OF SHOPPERS
Percent
Housewives 36.4
Professional 16.9
Self-employed 1.7
Skilled 4.9
Semi-skilled 3.0
Clerical 11.1
Service 11.3
Student 9.6
Other 5.1
This tabulation of the occupations of the shoppers interviewed
discloses that more than one-third were housewives. A total of 606
women were interviewed, 281 of whom were not gainfully employed. This
meant that 53.6 percent or slightly better than five out of ten women
interviewed worked at a full time job.
Mainland studies show that three in ten female family shoppers are
employed. 4 Thus, with a larger percentage of women working in Hawaii,
4"The Family Shopper," .QP.. cit., p. 1.
34
ease of shopping, time, and convenience are of great importance to them.
These characteristics of Island shoppers showed the differences
between the Island and Mainland shopper. The results were not surprising
because of the more varied ethnic groups and the fact that there are
more women shoppers gainfully employed in Hawaii.
SHOPPING PATTERNS
Some interesting results dealing with shopping behavior were
obtained from the responses of the 700 interviewed shoppers. Some
implications were drawn when the findings were compared to Mainland
findings.
Table VIII indicates the family member who does the shopping among
those interviewed. About 58 percent of the customers interviewed
shopped alone. About fourteen percent of the shoppers were accompanied
by spouses who assumed a passive role in shopping, and 16.3 percent of
the shoppers were accompanied by children. Eighty-six percent of the
shoppers interviewed were women.
Table IX compares Mainland and Hawaii findings on the question,
"Who Does the Family Shopping?" The shopping pattems in Hawaii tend
to be consistent with Mainland findings.
TABLE VIII. WHO DOES THE FAMILY SHOPPING?
Men Women Total
Alone 76 323 403
With Spouse 9 92 101
With Parent 2 16 18
With Relatives 1 16 17
With Friends 1 46 47
With Children 5 109 114
TABLE IX. WHO DOES THE FAMILY SHOPPING?MAINLAND AND HAWAII COMPARISON
Main1anda Hawaii
PercentWoman Alone 54.7 46.1
Woman with Chi1d(ren) 15.2 15.6
Man Alone 10.5 10.9
Woman and Man 8.6 14.4
Other Combinations 11.0 13.0
a"'rhe Family Shopper," .22.' cit., p. 2.
35
36
TABLE X. TYPES OF SHOPPING LIST
With shopping list
With mental list
No list
Men
10.3
1.4
1.7
Women
Percent
60.9
22.0
3.7
Total
71.2
23.4
5.4
From Table X it can be seen that 498 or 71.2 percent of all shoppers
interviewed had a written shopping list. On the other hand, 164 or 23.4
percent of the shoppers had no written shopping list but did have a
mental list. Lastly, 30 or 5.4 percent of the interviewed shoppers had
neither a written nor a mental shopping list; thus, all their purChases
were influenced by in-the-store decisions.
The shopping pattern in Hawaii with regard to a shopping list
showed a remarkable deviation from that on the Mainland. In fact, it
was almost a complete turn-about. The pattern exhibited on the Mainland
showed that most shoppers there carried no shopping list. The shoppers
tended to respond to in-the-store stimuli for suggestions for their
purchases. 5 The fact that Hawaii has a high percentage of members who
are employed full time suggests the following rationale. Free time is
scarce; therefore, in order to make best use of the limited amount of
5"The Family Shopper," .Q£.. .ill., p. 2.
37
time for family shopping, lists are more frequently used. Higher food
costs in Hawaii may also influence the shopper to prepare a written
shopping list based on a limited budget.
TABLE XI. HOW OFTEN DOES THE SHOPPERVISIT THE SUPERMARKET?
Mainlanda Hawaii
Percent
Once a week 26.9 30.1
Twice a week 25.3 33.3
Three times per week 22.1 15.5
Four times per week 9.8 7.1
Five times per week 15.9 4.0
a"The Family Shopper," .2E.. cit., p. 2.
Mainland studies revealed that the consumers average 2.7 shopping
trips per week. The shopper in Hawaii averages about 2.2 trips. A
comparison between Hawaii and the Mainland is shown in Table XI. The
comparison shows that the Mainland shopper visited the supermarket more
frequently than her counterpart in Hawaii. As noted earlier, about 3
out of 10 female family shoppers are employed on the Mainland compared
with 5 out of 10 female family shoppers in Hawaii. Time is important to
the working female family shopper and since Hawaii has a greater
percentage of working women it would seem logical that fewer shopping
38
trips per customer would result.
Table XII shows that the trend is toward an increasing number of
shopping trips per week, but this is only true up to four or five times
a week. It is very likely that Hawaii has the same type of trend.
TABLE XII. PERCENTAGE OF SUPERMARKET SHOPPERSSHOPPING SPECIFIED NUMBERS OF TIMES PER WEEKa
1954 1961 1963 1965
Percent
Once a week or less 62 59 55 48
Twice a week 18 23 24 26
Three times a week 10 12 13 16
Four times a week 4 4 6 7
Five or more times a week 6 2 2 3
aSchapker, .2E.' cit., p. 47.
Table XIII presents the percentages of shoppers in the sample who
conducted their major shopping trips in the various days of the week.
It shows that major shopping days for most customers are at the end
of the week.
39
TABLE XIII. MAJOR SHOPPING DAYS
Percent
Sunday 13.3
Monday 11.4
Tuesday 10.9
Wednesday 11.2
Thursday 15.3
Friday 17.5
Saturday 20.3
The findings here are consistent with the following items
previously mentioned:
1. The index of daily total receipts for the store.
2. The index of daily total customer count for the store.
3. The index of daily total produce sales.
These three indices were illustrated in Figures 7, 8, and 9,
respectively.
The average number of stores shopped per week is 2.16 with a
standard deviation of .92. The percentage of shoppers and the number
of stores they frequent per week is shown in Table XIV. Mainland studies
show that 61 percent of the shoppers frequent more than one store per
week. 6 An even larger percentage does so in Hawaii.
6 "The Family Shopper," .2E.. cit., p. 3.
TABLE XIV. NUMBER OF STORES FREQUENTED PER WEEK
Percent
40
One store/week
Two stores/week
Three stores/week
Four stores/week
Five or more stores/week
25.7
42.8
23.6
5.6
2.3
41
TABLE xv. PERCENTAGE OF SUPERMARKET SHOPPERSPATRONIZING MORE THAN ONE STORE PER WEEKa
1954 1961 1963 1965
One Supermarket Exclusively
More than one Supermarket
41
59
Percent
29
71
25
75
17
83
aschapker , E,E.. cit., p. 47.
Table XV shows an increasing trend toward patronizing more than one
store. The change is quite noticeable ~n the eleven-year period from
1954 to 1965. In 1954, 41 percent shopped at one supermarket
exclusively while 59 percent shopped at more than one supermarket. In
1965 the percentage of single store shoppers declined from 41 percent to
17 percent and that of multiple store shoppers increased from 59 percent
to 83 percent.
A reason for this result is increasing competition in the food
retail industry. Store loyalty has weakened as the competitive struggle
for sales has become stronger. Thus, it is becoming increasingly
difficult to achieve a distinctive store personality.
The survey of shoppers in Hawaii revealed that the average distance
traveled from home to the supermarket was 1.26 miles with a standard
deviation of 1.18 miles. From Table XVI it is noted that more than half
of the shoppers traveled less than one mile for their shopping needs.
42
TABLE XVI. DISTANCE TRAVELED TO THE SUPERMARKET
Percent
1 mile or less 61.4
Between 1 and 2 miles 16.7
Between 2 and 3 miles 12.4
Between 3 and 4 miles 4.9
Between 4 and 5 miles 2.9
5 miles or more 1. 75
Reasons shoppers gave for patronizing the test market are shown in
the following table. Convenience brought responses from more than
one-half of the respondents. "More specials" accounted for 29.1 percent
of all responses. An interesting reason was "know where everything is,"
and this was the response of 4.7 percent of the shoppers.
Shoppers who had budgets amounted to 344 out of the 700 interviewed,
but only 48.8 percent of this group responded that they usually stayed
within the budget. Shoppers who responded to the positive effect of
promotion or specials amount to 68.4 percent. In light of this, it
appears that it is possible to influence the shopper to buy more through
different selling and merchandising techniques.
43
TABLE XVII. REASONS FOR PATRONIZING THE TEST STORE
Percent
Convenient
Near
On the way
Quality
Service
More Specials
Low prices
Friends
Know where everything is
Other
57.4
29.1
7.4
7.9
3.4
29.5
11.6
6.0
4.7
3.0
(Note: Total percentages exceed 100 percent.)
The average number of grocery ads read per week by Island shoppers was
2.5. Table XVIII shows that 71.3 percent of all shoppers interviewed
read at least one food ad per week. It is interesting to note that the
two largest groups were 28.7 percent who read no food ads per week and
30.4 percent who read five or more food ads per week. Mainland studies
show that 32.2 percent of the shoppers check newspaper ads before
shopping. 7 The much higher use of ads by shoppers in Hawaii may be
attributed to the higher food prices and the higher percentage of
full-time employed family shoppers.
7"The Family Shopper," .2£.. cit., p. 3.
44
TABLE XVIII. NUMBER OF GROCERY ADS READ PER WEEK
Percent
0 ads read per week 28.7
1 ad read per week 12.0
2 ads read per week 16.6
3 ads read per week 8.7
4 ads read per week 3.6
5 or more ads read per week 30.4
TABLE XIX. NUMBER OF SUPERMARKET ADVERTISEMENTSREAD PER AD-READING SHOPPERa
1954
1.7
1961
2.4
1963
2.6
1965
2.8
aSchapker, .2£.. cit., p. 47.
Mainland shoppers read 2.8 ads per week, while Island shoppers
read 2.5 ads or .3 less than their counterparts on the Mainland. It is
interesting to note that the trend is toward an increasing number of ads
read by the ad-reading shopper. Note that in this study the 700 Island
shoppers included shoppers who read no ads. The Island average would
have been higher if only the figures from the ad-reading Island shoppers
were averaged.
45
The shoppers in Hawaii responded negatively to the question, "no
trading stamps cause you to buy more?" The results are indicated in
Table XX whiCh shows the percentage of shoppers for each classification.
TABLE XX. "DO TRADING STAMPS CAUSE YOU TO BUY MORE?"
Yes
10.2
No
Percent
88.6
Indifferent
1.1
In summary the findings in this chapter revealed that the Island
shopper was basically the same as her counterpart on the Mainland.
Answers given to the following questions showed similarities between
shopper in Hawaii and on the Mainland: 1) Who does the family
shopping? 2) How many marketing trips are taken per week? 3) How
many stores are frequented per week? 4) How many food ads are read per
week?
Some findings that.showed a marked deviation from Mainland results
were: 1) A greater number of Island shoppers used a written shopping
list; 2) More women shoppers were fully employed; 3) Household incomes
of Island shoppers were higher than on the Mainland; and 4) Island
shoppers come from more varied ethnic groups. Aside from these few
deviations Island shoppers are very similar to their counterparts on the
Mainland.
46
Questions asked of the consumers were directly related to factors
leading to impulse buying. For example:
1. Determining the number of stores frequented per week is an
indication of how much competition is involved, thus knowing
how effective in-store techniques are in inducing impulse
buying in that particular store.
2. Determining whether a shopper has a shopping list indicates how
strongly in-store techniques can contribute to or alter a
shopping list.
3. Determining numbers of food ads read shows how effective food
ads are in drawing a customer in and making up his list.
4. Determining how many shopping trips are made per week shows the
need for better in-store stimuli to persuade the shopper to buy
more.
5. Determining who does the family shopping helps to determine if,
for example, two persons shopping together do more impulsive
buying than one person shopping alone.
In the next chapter, some of these characteristics of shoppers that
are relevant to unplanned purchases will be discussed in more detail.
CHAPTER V
THE IMPACT OF MULTIPLE PRICING AND INCREASED DISPLAY SPACEON SALES OF SELECTED PRODUCE ITEMS
This chapter is divided into three parts. The first part evaluates
the effectiveness of the marketing techniques studied by using the
analysis of variance test on the sales in the control week and in the
week in question for each of the three products. The second part of the
chapter examines the differences in the percentages of purchases that
are unplanned between the test weeks and the control week for each of
the three items. The third part discusses the fifth week experiment
where two different sales methods for bell peppers were employed.
Table XXI shows the test schedule used for the three selected
products. The variables, multiple pricing and increased display space,
were varied for three selected items from week to week in the following
manner:
1st week (control: Normal price and normal display space
2nd week: Multiple pricing and normal display space
3rd week: Normal pricing and increased display space
4th week: Multiple pricing and increased display space
This schedule resulted in a balanced experiment with two variables.
During the fifth week bell peppers were sold on a per-pound basis
versus a unit-cost basis. This meant that there were two displays of
bell peppers. One display sold bell peppers at 49¢ per pound. The
other display sold pre-sized bell peppers at 12¢ each. (The l2¢ each
bell pepper was identified with a small white sticker.) The two prices
were essentially equivalent.
TABLE XXI. TEST SCHEDULE
48
Test Period Price Display Space
1st week (Control)
Navel Oranges 30¢/lb. 54"Turnips 19¢/lb. 19"Bell Peppers 59¢/lb. 19"
2nd week
Navel Oranges 3 lbs. for 89¢ 54"Turnips 2 lbs. for 39¢ 19"Bell Peppers 4 pes. for 59¢ 19"
3rd week
Navel Oranges 30¢/lb. 86"Turnips 19¢/lb. 24"Bell Peppers 59¢/lb. 24"
4th week
Navel Oranges 3 lbs. for 89¢ 86"Turnips 2 lbs. for 39¢ 24"Bell Peppers 4 pes. for 59¢ 24"
5th week
Bell Peppers -.12¢ apiece 15"Bell Peppers 49¢/lb. 15"
49
The volume movement of the three selected produce items was recorded
daily. The produce manager was responsible for compiling the daily
totals of the volume movements of navel oranges, turnips and bell peppers
for the duration of the five-week experiment.
The volume movements of navel oranges, turnips, and bell peppers are
shown graphically in Figures 11, 12, and 13 respectively. Figure 14
shows the volume movement of the fifth week experiment of bell peppers
selling at a per-pound basis versus a unit-cost basis. During the
period of the study there was no discernible change in quality or
availability of any of the three produce items.
The technique, analysis of variance, is a useful tool because the
total variation of the variable being studied can be separated into
components that are of interest to the experimenter. In this study the
researcher was interested in the effects of two marketing practices,
namely multiple pricing and increased display space, on sales of three
selected produce items. Four analysis of variance tables were drawn up
for each of the selected produce items. One analysis of variance table
includes all of the weeks. The remaining three analysis of variance
tables consist of comparing weeks two, three and four with week one, the
control week.
The term "analysis of variance" refers to a general method of
statistical inference. It consists of a body of tests of hypotheses and
methods of estimation, using statistics which are linear combinations of
sums of squares of linear functions of the observed values. The total
variability in a set of observations can be partitioned to determine the
relative amount of variations attributable to the various sources. The
TABLE XXII. DAILY VOLUME MOVEMENTS (IN LBS.) OF NAVEL ORANGES, TURNIPS,AND BELL PEPPERS FOR THE DURATION OF THE EXPERIMENT
S M T W T F S WEEKLY
Control weekNavel Oranges 210 20 70 175 140 105 140 860Turnips 30 25 35 62 40 20 90 302Bell Peppers 30 15 25 15 25 20 40 170
2nd weekNavel Oranges 80 90 100 210 105 210 175 970Turnips 25 20 50 68 40 59 80 342Bell Peppers 45 25 30 20 30 40 32 222
3rd weekNavel Oranges 175 140 210 140 105 175 140 1085Turnips 30 62 30 50 25 75 40 312Bell Peppers 25 55 45 21 35 35 40 256
4th weekNavel Oranges 140 175 105 140 120 140 175 995Turnips 35 45 50 30 50 55 70 335Bell Peppers 30 30 45 30 30 30 45 240
5th weekBell Peppers (12¢ apiece) 20 20 30 30 20 20 30 170Bell Peppers (49¢/lb.) 10 10 15 15 10 15 25 100
VIo
(Lbs.250
200
CD
~ 150as
tJ)
Il-loQl
~r-l~ 100
50
SMTWTFSSMTWTFSSMTWTFS SMTWTF S(Days)
FIGURE 11. POUNDS OF NAVEL ORANGES SOLD DAILY FOR DURATION OF EXPERIMENT lJ1....
(Lbs100
75
IIIQl
.-lCIStil
'H0 50
~.-l0:>
25
S M T W T F SSM T W T F SSM T W T . F-SSM T W T F(Days)
FIGURE 12. POUNDS OF TURNIPS SOLD FOR DURATION OF EXPERIMENT
V1N
(Lbs. )60
45tllQ)
.....a3en~
o
]30~
15
SM TWTFS SMT WTF S SMTWTF SSM TWTFS(Days)
FIGURE 13. POUNDS OF BELL PEPPERS SOLD FOR DURATION OF EXPERIMENT \.IIUJ
Priced per Lb.
Priced per unit
P49
P12
10
25
(Lbs. )30
§.-I~ 15
rt.l
~ 20~
tr.l
\4-1o
Su I I iii I •
S M T w T F S(Days)
FIGURE 14. COMPARISON OF TWO METHODS OF SELLING BELL PEPPERS V1~
55
F-test is the mechanism for testing the s.ignificance of the differences
among sample means.
For a two-way classification, with one observation per cell, a
sample of N values of X, the variate, can be classified according to
some factor A into m classes; and according to some factor B into n
classes; (thus N = mn). The sample variate value in the i th A-class and
jth B-c1ass would be Xij. Consequently:
m nE E (Xij - x.. )2 =
i=l j=l
m _ _ 2 n 2E n(xi. - x •• ) + E m(x.j - x.. ) +
i=l j=l
m n- - 2E E (XiJ" - xi. - X. J" x •• )
i=l j=l
The left hand member of this equation is the total sum of squared
deviations of the sample values of the variate from the general mean or
the total variation.
The first term on the right hand side of the equation is the sum of
the squared deviations from the general mean if all variation within the
A-classes is eliminated.
The second term is the sum of squared deviations from the general
mean if all variation within B-c1asses is eliminated.
The third term, the residual term, measures the variation in x
remaining after the variation due to that between A-classes and that
between B-c1asses has been separated out. 1
lRichard Goodman, Modern Statistics (New York, 1964), pp. 166-173.
56
In short,
Total Variation = Variation between A-Classes +
Variation between B-C1asses + Residual variation
The analysis of variance table is therefore:
ANALYSIS OF VARIANCE FOR TWO CRITERIAOF CLASSIFICATION (GENERAL FORM)
Source ofVariation
BetweenA-Classes
BetweenB-C1asses
Residual(A x B)
TOTAL
Sum of Squares
n- - 2E m(xi - x •• )
j=l
m nE E (xij - xi.
i=l j=l
- - )2- x. j x ••
m n- 2E E (Xij - x •• )
i=l j=l
Degreesof Freedom
m-1
n-1
(m-1)-(n-1)
mn-1
Estimate of VariancemE n(xi - i .. )2/(m-1)
i=l •
nE m(x.j - x •• ) 2/ (n-1)
j=l
m nE E (Xij - xi. - x.j
i=l j=l
x.. )2/(m-1)(n-1)
Source: Richard Goodman, Modern Statistics (New York, 1964), p. 173.
57
To test the effectiveness of the two variables, multiple pricing
and increased display space, the analysis of variance was employed to
see if weekly volume movements were significantly different. The null
hypothesis states that there is no difference between the first week,
the norm, and the week with special treatment. By using the analysis of
variance test one can readily compute the probability of the null
hypothesis, Ho ' that there is no difference.
The volume movements of the navel oranges were:
Control Week 860 pounds
2nd Week 970 pounds
3rd Week 1,085 pounds
4th Week 995 pounds
It can be seen that in the second week the F value was .53. This
meant that the difference in sales of navel oranges between test week
two and the control week was hardly significant. The difference
attributed to the test variable, multiple pricing, was small. The
change from 30¢/lb. to 3 lbs. for 89¢ did not seem great enough to
increase the sales of navel oranges significantly. The decision was to
accept the null hypothesis that there was no difference.
In the third week the F-value was 1.23. This meant that the
difference in the sales of navel oranges between test week three and the
control week was significant between .25 and .50 levels. In other words,
a difference equal to or greater than this magnitude would occur 25 to
50 percent of the time merely by chance, rather than because of the
nest variable, increased display space, being used in the third week.
It appears that increased display space was a more effective method to
TABLE XXIII. ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSIFICATION, DAYS AND WEKKS, FOR NAVEL ORANGES
58
Source of Sum of Degree of EstimateVariations Squares Freedom of Variance F
Between all 3675.00 3 1225.00 .53C&2 864.29 1 846.29 .29
Treatments C&3 3616.07 1 3616.07 1.23(weeks) C&4 1301. 79 1 1301. 79 .49-----------------------------------Between all 13696.43 6 2282.74 .98
C&2 26692.86 6 4448.81 1.50Days C&3 13921.43 6 2320.50 .79
C&4 12825.00 6 2137.50 .81- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -Residuals all 41875.00 18 2326.38
C&2 17835.71 6 2972.61C&3 17671.43 6 2945.23C&4 15810.71 6 2635.11- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -all 59246.43 27
. Total C&2 45392.86 13C&3 35208.93 13C&4 29937.50 13
302 pounds
342 pounds
312 pounds
335 pounds
59
increase the sales of navel oranges when the price per pound is as high
as thirty cents, but the increase in sales was still not highly
significant.
The fourth week essentially yielded the same result as the second
week. Again the difference in the sales of navel oranges between test
week four and the control week was hardly significant. The difference
attributed to the test variables, multiple pricing and increased
display space, was small. It appeared that, collectively, these two
variables did not increase sales of navel oranges significantly.
Note that added display space increased sales more than did added
display space plus multiple pricing. This is an interesting point
because one would expect the opposite to occur. A possible rationale
for this was the fact that shoppers became accustomed to the added
display space. Note that in the third week, navel oranges were sold at
the normal price with added display space, and on the following week,
the navel oranges were sold with multiple pricing and added display
space. The change from normal to multiple pricing was a small change
30¢/lb. to 3 lbs. for 89¢. It appears that the novelty of added display
space wore off in the fourth week. As mentioned previously, multiple
pricing in the second week did not affect the sales of navel oranges
significantly.
The volume movements of turnips exhibited a rather constant trend.
They were as follows:
Control Week
~d~~
3rd Week
4th Week
60
From Table XXIV the F-value in the second week was 8.05. This
meant that the differences in sales of turnips between test week two
and the control week was significant at the 0.05 level. The test
variable in this week was multiple pricing. A difference in the sales
of turnips equal to or greater than this magnitude would occur 5 out of
100 times merely by Chance rather than because of the test variable,
multiple pricing, being used in the second week. The price Change from
19¢ per pound to 2 pounds for 39¢ appears to be successful in increasing
the sales of turnips to a significant amount. The decision was to
reject the null hypothesis that there was no difference.
By similar reasoning, test variables in weeks three and four were
not successful in increasing the sales of turnips to a significant
amount. The test variable in week three was increased display space and
the test variables in week four were multiple pricing and increased
display space. The decisions in both cases were to accept the null
hypothesis that there was no difference.
Sales of some food products are not always responsive to Changes
in shelf space. 2 Turnips may very well be one such item. Or perhaps
the sales of turnips were not responsive to the given range of display
space.
2Keith Cox, "The Responsiveness of Food Sales to Shelf SpaceChanges in Supermarkets," Journal of Marketing ResearCh, I (May, 1964),p. 66.
TABLE XXIV. ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSIFICATION, DAYS AND WEEKS, FOR TURNIPS
61
Source of Sum of Degree of EstimateVariations Squares Freedom of Variance F
Between all 152.38 3 50.79 .23C&2 114.28 1 114.28 8.05
Treatments C&3 7.25 1 7.25 .01(weeks) C&4 77.79 1 77.79 .28- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Between all 4209.18 6 701.53 3.11C&2 5734.00 6 955.66 67.34
Days C&3 2115.71 6 352.62 .58C&4 3086.00 6 514.33 1.90
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Residual all 4056.79 18 225.37C&2 851. 72 6 14.19C&3 3636.75 6 606.12C&4 1621. 71 6 270.68
Total allC&2C&3C&4
4868.686700.005759.714785.50
27131313
62
The volume movements of bell peppers were as follows:
Control Week 170 pounds
2nd Week 222 pounds
3rd Week 256 pounds
4th Week 240 pounds
The F-values in weeks two and three were 4.88 and 4.77. This
meant that the difference in the sales of bell peppers in test week two
and control, and test week three and control were significant at the
0.10 level. The decision in both cases was to reject the null
hypothesis that there was no difference. Or, simply stated, a
difference equal to or greater than this magnitude would occur by
chance only 10 out of 100 times. The test variables in week two and
week three were multiple pricing and increased display space
respectively. It appears that either variable is effective in
increasing the sales of bell peppers to a significant amount.
Using the same type of reasoning, the decision was to reject the
null hypothesis in week four (a = 0.25). The test variables in this
week were multiple pricing and increased display space. The results in
Table XXV show that all variables affected the sales of bell peppers
significantly. The best result was obtained when either multiple pricing
or increased display space was used alone. The use of both variables
together did not produce as significant a result as the individual use
of the variables. This may be because the novelty of increased display
space wore off in the fourth week.
TABLE xxv. ANALYSIS OF VARIANCE TABLE FOR TWO CRITERIA OFCLASSIFICATION, DAYS AND WEEKS, FOR BELL PEPPERS
63
Source of Sum of Degree of EstimateVariations Squares Freedom of Variance F
Between all 476.98 3 158.99 2.03C&2 193.86 1 193.86 4.88
Treatments C&3 528.29 1 528.29 4.77(weeks) C&4 771.43 1 171.43 3.13- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -Between all 336.27 6 56.07 .72
C&2 666.00 6 111.00 2.80Days C&3 610.43 6 101. 74 .92
C&4 642.86 6 107.14 1.96- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -Residual all 1407.82 18 78.21
C&2 238.14 6 39.69C&3 664.71 6 110.78C&4 328.57 6 54.76
Total allC&2C&3C&4
2221.071098.001803.431142.86
27131313
64
The summary of the findings disclosed that at the given ranges of
pricing and increased display space: 1) The sales of navel oranges could
be increased by increasing the display space at the rather high cost of
30¢ per pound; 2) the sales of turnips was increased by multiple pricing;
3) the sales of bell peppers could be increased best by using each
variable individually. The summation of findings are shown in Table
XXVI.
TABLE XXVI. SIGNIFICANCE TESTS FOR NAVEL ORANGES, TURNIPS, AND BELL PEPPERS
Week Variable Navel Oranges Turnips Bell Peppers
2 Multiple Pricing Accept Ho Reject no Reject noor: = 0.05 or: = 0.10
3 Increased display Reject no Accept Ho Reject nospace or: = .25 - .50 or: = 0.10
4 Multiple pricing and Accept Ho Accept Ho Reject noIncreased display or: = 0.25space
0\\JI
66
WEEKLY PURCHASES COMPARED
The chi-square test was used to see if the number of unplanned and
planned purchases of the three selected produce items in the test week
was significantly different from the control week. The null hypothesis,
Ho ' that there is no difference, was tested.
2 x 2 Contingency Table(General Form)
ControlWeek
TestWeek
Total
UnplannedPurchases
A
C
A+C
PlannedPurchases
B
D
B+D
Total
A+B
C + D
N 2X2 = ~--=-_N~(_AD-....-_B~C_-_2~)---:-_----,
(A+b) (C + D) (A + C) (B + D)df = 1
the X2 test is the test of whether an observed breakdown offrequencies in a 2 x 2 contingency table could have occurredunder Ho .3
3Sidney Siegel, Nonparametric Statistics for Behavioral Sciences,New York, 1956, p. 107.
67
Applications of the X2 test were generally nonsignificant. See the
Appendix for computations. Unplanned sales were actually lower than in
the control week in a number of instances. Evidently such increases in
volume of sales as were noted cannot be attributed to increased numbers
of unplanned purchases being induced by the test variables.
COST PER-UNIT VERSUS COST PER-POUND(TWO SALES METHODS FOR BELL PEPPERS)
In this portion of the experiment, bell peppers were displayed in
two separate displays. Some interesting results were obtained. The
two bell pepper displays were separated by a small display of packaged
carrots (approximately a one-foot display). One bell pepper display
was priced at 49¢ per pound. The other display sold uniform, pre-sized
bell peppers for l2¢ each, based on the price of 49¢ per pound.
A total of 140 customers was randomly chosen and interviewed.
It was found that approximately two-thirds of the customers
interviewed who had purchased bell peppers in the fifth week chose the
pre-sized bell peppers priced at l2¢ each. Of this group, 67 percent
thought that the bell peppers priced at l2¢ each were cheaper but of the
same quality as those priced at 49¢ per pound, while 23 percent thought
the quality was better. Five and one-half percent thought that the bell
peppers priced at l2¢ apiece were of inferior quality and were sold at
a "special price." They had erroneously assumed that at l2¢ each the
peppers were cheaper in price and inferior in quality to those priced at
49¢ per pound. Amazingly, 4.5 percent of the group that purchased the
bell peppers at the unit price of l2¢ didn't see the other display of
TABLE XXVII. REASONS GIVEN FOR BUYING BELL PEPPERSPRICED AT 12¢ APIECE
68
Reasons
Thought P12 was cheaperthan P49, but of equalquality
Thought the quality ofP12 was better
Thought P12 was a "special"(cheaper in price), but ofinferior quality to P49
Didn't see the P49 display
Number of shoppers interviewedwho bought bell peppers atl2¢ apiece
Number ofShoppers
61
21
5
4
91
Percentageof Shoppers
67
23
5.5
4.5
100
TABLE XXVIII. REASONS GIVEN FOR BUYING BELL PEPPERSPRICED AT 49¢/LB.
69
Reasons
Thought P49 was superiorin quality because P12was a "special"
Thought P49 was cheaper,but of equal quality
Thought no difference betweenP12 and P49 but needed a verysmall one
Didn't see the P12 display
Number of shoppers interviewedwho bought bell peppers at49¢/lb.
Numberof Shoppers
30
17
1
2
49
Percentageof Shoppers
61.2'
34.7
2.0
4.1
100
70
peppers priced at 49¢ per pound, that was only approximately a foot
away. Thirty-three percent of all purchases from this group were
unplanned. Ten and seven-tenths percent of the 140 persons interviewed
bought more than 'they had intended and they all bought these additional
peppers from the group priced at 12¢ each.
The remaining 35 percent of the 140 shoppers interviewed offered
similar explanations for their purchases of the bell peppers priced at
49¢ per pound. Of this group 61.2 percent had erroneously reasoned
that the quality of the peppers priced at 49¢ per pound was superior
because those priced at 12¢ each were "a special." Thirty-four and
six-tenths percent thought that the peppers priced at 49¢ were cheaper.
Again, 4 percent of the shoppers interviewed didn't see the other
display. Las t1y, there was one shopper out of the 140 interviewed who
reasoned correctly -- and bought only one pepper priced at 49¢ per
pound because he needed a very small one. On1y.6 percent of all
purchases in this group were unplanned.
It can be concluded that there was a difference between the two
methods of selling bell peppers. Selling the peppers by cost per unit
was better than selling them by the pound.
CHAPTER VI
UNPLANNED PURCHASES AND CONSUMER AWARENESS OFMERCHANDISING TECHNIQUES
This Chapter is divided into four parts. Part one discusses
unplanned purchases while part two examines consumer awareness of
merchandising teChniques. Parts three and four evaluate the reasons
for purchasing and reasons for not purchasing the three selected
produce items.
UNPLANNED PURCHASES
The effectiveness of the merchandising teChniques of multiple
pricing and increased display space can be identified with the number
of unplanned purchases the marketing techniques can induce. It is
reasonable to assume that the marketing teChnique was successful if it
could induce a significant number of unplanned purchases and thereby
generate a significant increase in the level of sales, all other
variables being held constant. The following table shows unplanned
purchases as a percentage of purChases.
Summarizing the results of the table showed that most of the
unplanned purChases of navel oranges occurred in the third week where
the test variable was increased display space. This finding was
consistent with the result of the previously discussed statistical test
of significance of weekly sales. Both concurred that increased display
space was the most effective test variable in increasing the sales of
navel oranges at these given ranges of prices and display spaces. About
72
TABLE XXIX. UNPLANNED PURCHASES AS APERCENTAGE OF TOTAL PURCHASES
Week Variable Navel Oranges Turnips Bell Peppers
Percent
Control 32.5 35.9 35.9
2 Multiple pricing 28.2 21.1 44.7
3 Increased display 35.9 16.7 38.5space
4 Multiple pricing 30.8 18.4 37.1and
increased displayspace
36 percent of all purchases of navel oranges were unplanned in this
week.
Most of the unplanned purchases of turnips occurred in the second
week, but sales were not significantly different from the control week.
About 21 percent of all purchases of turnips were unplanned in this
week. The test variable, multiple pricing, could not induce enough
unplanned purchases to make sales significantly different from the
control week. It appears that at the given ranges of multiple pricing
and increased space, neither was effective in increasing sales
significantly.
Most of the unplanned purchases of bell peppers occurred in the
second week where the test variable was multiple pricing. About 45
percent of all purchases of bell peppers were unplanned in this week.
73
This finding was consistent with the result of the previously discussed
statistical test of significance of weekly sales. Both concurred that
multiple pricing was the most effective way of increasing the sales of
bell peppers at these given ranges of prices and display spaces.
Most of the unplanned purchases of the three selected produce
items were made by shoppers whose ages ranged between 35 - 49 years old.
This finding is not too surprising because as years of marriage increase
and the family grows, the consumption of food increases and the need
for variety and quantity of food rises. Pre-planning in this age group,
35 - 49 years old, becomes more time consuming and difficult. Thus,
the result is that the shopper relies more on in-store stimuli. Also,
shoppers in this age group generally have more experience in shopping
and feel better qualified to evaluate purchase alternatives within the
store. The under 20 age group's percent of unplanned purchases were
unusually high because of the extremely small number of shoppers
interviewed belong to this age group. Table XXX summarizes unplanned
purchases as a percent of all purchases, by age group.
Most unplanned purchases of the three selected produce items made
by the shoppers were on major shopping trips as opposed to fill-in
trips. This is logical because on major shopping trips the needs of
the shoppers are many and less definite as compared to a fill-in trip
where specific items to purchase are pre-planned. Also, during a major
shopping trip a greater amount of time is spent in the store.
Consequently, the shopper would be more receptive to in-store stimuli.
Table XXXI summarizes unplanned purchases as a percent of all purchases
and the major or fill-in shopping trip.
TABLE XXX. UNPLANNED PURCHASES AS A PERCENT OFALL PURCHASES, BY AGE GROUP
74
Week Age Group Navel Oranges Turnips Bell Peppers
Percent
2 Under 20 years 33.320-34 years old 3.3 3.335-49 years old 14.6 2.1 20.150-64 years old 6.3 31.3 25.0Over 65 years 16.7 33.3
3 Under 20 years 33.320-34 years old 6.7 3.335-49 years old 16.7 13.9 18.850-64 years old 12.5 6.3 18.8Over 65 years.
4 Under 20 years 33.320-34 years old 3.3 6.735-49 years old 16.7 13.2 14.650-64 years old 6.3 12.5 12.5Over 65 years
. I
TABLE XXXI. UNPLANNED PURCHASES AS A PERCENT OF ALL PURCHASESAND THE MAJOR OR FILL-IN SHOPPING TRIP
75
Type ofWeek Shopping Trip Navel Oranges Turnips Bell Peppers
Percent2 Major 23.1 10.5 23.7
Fill-in 5.1 10.5 13.2
3 Major 20.5 11.1 25.6Fill-in 15.4 5.6 17.9
4 Major 17.9 10.5 20.0Fill-in 12.8 7.9 17.1
76
From the classifications of shoppers, women shoppers who shopped
with their children made the greatest amount of unplanned purchases.
The second largest group was the man and woman classification. The
results are plausible because the more people involved in one shopping
trip the greater the chances are of seeing more items in the store.
For example, children tend to make their mothers more aware of many
possible in-the-store decision purchases.
Table XXXII shows unplanned purchases as a percent of total
purchases by who does the family shopping.
CONSUMER AWARENESS OF APPLIED MERCHANDISING TECHNIQUES
Most studies dealing with multiple pricing or other merchandising
practices designed to increase impulse buying have assumed that because
the sales were increased, the shoppers were quite aware of the applied
practices. In this section the assumption that the shoppers were aware
of multiple pricing and increased display space as applied to navel
oranges, turnips, and bell peppers will be discussed.
The shopper was asked a series of questions dealing with this
assumption. She was asked if she had planned to purchase the item, and
if so, how much of it. Then she was asked to reveal the actual quantity
of her purchase. If an unplanned purchase occurred she was asked to
give a reason for her purchase. This question gave the shopper a wide
latitude of choices among the possible answers. At this point of the
interview, a very specific question was asked of the shopper about her
awareness of the applied marketing techniques. The shopper was asked,
"Did you notice any difference between purchasing the item this week
77
TABLE XXXII. UNPLANNED PURCHASES OR A PERCENT OF TOTAL PURCHASESBY WHO DOES THE FAMILY SHOPPING
Who Does theWeek Family Shopping Navel Oranges Turnips Bell Peppers
Percent
2 Woman alone 6.2 6.2 9.2Woman with child(ren) 13.6 27.2Man alone 6.7 6.7 13.4Woman and Man 10.0 15.0 10.0Other combinations 5.6
3 Woman alone 9.2 4.6 9.2Woman with child(ren) 22.7 22.7Man alone 6.7 6.7 6.7Woman and man 5.0 5.0 15.0Other combinations 5.6
4 Woman alone 7.7 4.6 5.0Woman with child(ren) 18.2 4.5 18.2Man alone 6.7 6.7 6.7Woman and man 5.0 5.0 15.0Other combinations 5.6 5.6
78
and last week; for example, price, display, size, etc.?" This question
was asked only after the shopper was given the freedom to s tate why she
had purchased the item. Thus the opportunity to reinforce her reason
for her purchase was afforded the customer. Indirectly the shopper was
made to reveal her awareness of merchandising techniques as applied to
her purchase.
Table XXXIII shows the schedule of the prices and the display
spaces connected with each item for the experiment. The results .of the
shoppers' awareness of merchandising techniques as applied to navel
oranges, turnips and bell peppers are shown in Table XXXIV.
In the purchase of navel oranges, 15/140 or 10.7 percent of the
shoppers indicated that they noticed some difference between the weekly
display from which they had purchased and the preceding week's display.
But only 1/140 or .7 percent gave the correct answer. Fourteen out of
140 or 10 percent gave incorrect answers.
Similarly, thirteen or 9.3 percent of the shoppers who purchased
turnips indicated that they noticed a difference in its display. Only
two or 1.4 percent gave the correct answer while eleven or 7.9 percent
gave incorrect answers as to what was different about the display.
Lastly, twenty-nine or 20.7 percent of the shoppers who purchased
bell peppers indicated that they noticed a difference in its display.
Five or 3.6 percent gave correct answers and twenty-four or 17.1 percent
gave incorrect answers.
In any given week the shoppers who gave incorrect answers had a
net cancelling effect. In other words, the number of shoppers who
thought the quality of the item was better was about equal to the number
TABLE XXXIII. TEST SCHEDULE
Week Variable Navel Oranges Turnips Bell PeppersPRICE DISPLAY PRICE DISPLAY PRICE DISPLAY
CONTROL -- 30¢/lb. 54" 19¢/lb. 19" 59¢/lb. 19"
2nd Multiple pricing 3 1bs.-89¢ 54" 2 1bs.-39¢ 19" 4 pcs.-59¢ 19"
3rd Increased display 30¢/lb. 86" 19¢/lb. 24" 59¢/lb. 24"space
4th Multiple pricing and 3 1bs.-89¢ 86" 2 1bs.-39¢ 24" 4 pcs.-59¢ 24"Increased displayspace
......\0
TABLE XXXIV. SHOPPERS' AWARENESS OF DISPLAYS
80
Navel Oranges Turnips Bell Peppers
Number or Percent
Number ofShoppers Correctly 1 or .7 2 or 1.4 5 or 3.6Who NoticedDifferences Incorrectly 14 or 10 11 or 7.9 24 or 17.1in WeeklyDisplays Total 15 or 10.7 13 or 9.3 29 or 20.7
81
of shoppers who thought the quality of the item was poorer.
It was concluded that at these ranges of prices and display spaces
for navel oranges, turnips and bell peppers, the shoppers' awareness of
price and display space changes is minimal.
REASONS FOR PURCHASES
The shopper was asked questions dealing with her purchasing
activities within the store. In this section the reasons given by the
shoppers for their purchases of selected produce items will be
discussed.
Sixty out of 140, or 42.9 percent of the shoppers who purchased
navel oranges made in-the-store decisions. Of this group 29/60 or 48.3
percent purchased navel oranges because they "remembered it" (always
buy); 20/60 or 33.3 percent of the group purchased the navel oranges
because of a change in menu.
Note that only 15/140 or 10.7 percent of all shoppers who purchased
navel oranges noticed any difference between the current display and the
preceding week's display. At best they could only indicate the
direction of change with regard to price. Usually they would indicate
that they suspected a change in price. The shoppers indicated readily
their lack of knowledge of the normal price of navel oranges. This
point is supported by a study made by Morris and Firch who indicated
the same resu1t. I The shoppers' opinions of quality differences tended
IJames L. Morris and Robert J. Firch, "The Analysis of SeparatePrice and Advertising Responses to Retail Grocery Specials," Proceedings1967 Western Farm Economics Association, XXXX (July, 1967), 47-52.
TABLE XXXV. REASONS FOR PURCHASING NAVEL ORANGES
1st Week 2nd Week 3rd Week 4th Week
Intended to Buy 27 28 25 27Any Differencefrom Last Week? (NO) 25 25 24 26
(YES) 2 Q'" 3 Pt 1 Qt 1 I::.P
4
3
5
4
5
5
2
4
5
2
6
5
Not Intendingto Buy andBought 13 11 14 12R Remembered ite (always buy it)as Bought Moreon Change in MenusAny Differencefrom Last Week? (NO)
(YES)10 10 _ ~__ ~2 10
3 Qt 1 I::.P 2 Qt 2 Qt
00N
83
to balance out in any given week. Because of the shoppers' inability
to recognize the difference between successive weekly displays, it
appeared that the carry-over effect from the preceding week was minimal.
Thirty-five out of 140 or 25 percent of all shoppers who purchased
turnips made in-the-store decisions. Of this group 17/35 or 49 percent
of the shoppers purchased turnips upon remembering it (always buy) and
10/35 or 29.6 percent of this group made purchases because of a change
in menu decision. Only 13/140 or 9.3 percent of all purchasers of
turnips indicated a recognition of a difference between the current
display and the preceding week's display.
The results were similar to the ones obtained for navel oranges.
The customers' lack of knowledge of the normal price for turnips was
demonstrated. The carry~over effect from one week to the next was
probably minimal.
Fifty-nine out of 140 or 42.1 percent of all shoppers who purchased
bell peppers made in-the-store decisions. Of this group 21/59 or 35.6
percent purchased bell peppers upon remembering it (always buy) and
22/59 or 37.3 percent of the group purchased bell peppers because of a
Change in menu decision. Only 29/140 or 20.7 percent indicated a
recognition of any difference between the current week's display and
the preceding week's display of bell peppers. Again they only
suspected something was changed. The results were similar to those
obtained for navel oranges. The lack of knowledge of the normal price
for bell peppers was revealed. Again, the carry-over effect from one
week to the next was probably minimal.
TABLE XXXVI. REASONS FOR PURCHASING TURNIPS
1st Week 2nd Week 3rd Week 4th Week
Intended to Buy 25 30 30 31Any Differencefrom Las t Week? (NO) 24 27 29 31
(YES) 1 Qt 3 Qt 1 Qt 0
Not Intendingto Buy andBought 14 8 6 7R Rememb ered ite (Always buy it) 9 3 3 2as Bought More 4 1 1 2an Change in Menu 1 4 2 3sAny Differencefrom Last Week? (NO) 13 6 5 3
(YES) 1 Qt 2 Qt 1 Qt 4 Qt
00.po
TABLE XXXVII. REASONS FOR PURCHASING BELL PEPPERS
1st Week 2nd Week 3rd Week 4th Week
Intended to Buy 25 21 24 22Any Differencefrom las t Week? (NO) 22 20 21 20
(YES) 3 Qt 1 Qt 3 Qt 2
Not Intendingto Buy andBought 14 17 15 13Re Remembered ita (Always buy it) 5 5 6 5s0 Bought More 4 3 2 7ns Change in Menu 5 7 5 5Any Differencefrom Las t Week? (NO) 9 11 12 7
(YES) 5 Qt 6 Pot 3 Qt 6 tQ
00VI
86
To summarize, increased display and price changes of the three
selected products produced different results. For bell peppers, the
two merchandising techniques proved highly effective in producing
increased sales and unplanned purchases. Navel oranges showed significant
increases, though not as substantial as bell peppers. However, it seems
that the original control display for turnips was the most successful in
inducing unplanned purchases. For all three products, although consumers
were aware of their unplanned purchase, they were not aware of the
merchandising techniques which induced it.
REASONS FOR NOT PURCHASING
As in the previous section, the shopper was asked questions that
indicated the reasons for not purchasing the selected produce items.
It is interesting to note that of the group of sixty-nine shoppers
who did not purchase navel oranges because they thought the price was
high, thirty-six of them, or slightly over half, purchases orange juice
instead. Of the sixty-six who did not purchase navel oranges because
they thought the quality was poor, about two out of three purchased
other fruits. Note also that sixty-eight shoppers did not purchase
navel oranges because other commodities influenced them to make a change
in their menu -- not because of the influence of navel oranges.
One hundred and eighty-four shoppers who did not purchase turnips
stated that they "never buy turnips." Of this group thirty-two (nearly
one out of five) were not familiar with turnips. Seventy-two shoppers
stated that they "seldom buy turnips." Only five (1.2 percent) of all
TABLE XXXVIII. REASONS FOR NOT PURCHASINGNAVEL ORANGES
Reasons Number Percentage
Never buy 21 5
Seldom buy 38 9
Price is high 69 16.5
Poor quality 66 15.8
Didn't see display 17 4.1
Have at home 70 16.7
Change in menu 68 16.2
Not on shopping list 44 10.5
Miscellaneous 26 6.2
Total 419 100
87
88
shoppers who didn't buy turnips stated that the price was high. There
appeared to be a strong dislike for the turnip as a food source among a
large segment of the population.
One hundred twenty-one shoppers stated that they "never buy bell
peppers." Another eighty-two shoppers stated that they "seldom buy
bell peppers." These two classifications constituted about 48.7 percent
of all shoppers who did not purchase bell peppers. Forty-four or 20
percent of the shoppers did not purchase bell peppers because they
thought the price was high. Fifty-seven shoppers did not purchase bell
peppers because they thought the quality was poor. The bell peppers,
like turnips, seemingly lacked the general appeal of navel oranges.
In this chapter an effort was made to discover in what areas
unplanned purchases occurred and also whether consumers were aware that
the merchandising techniques were causing them to make unplanned
purchases. The findings were:
1. The greates t amount of unplanned purchases for the three
selected products occurred in the 35-49 year old age group.
2. The greatest amount of unplanned purchases for the three
selected products occurred during major shopping trips as
opposed to fill-in trips.
3. The greatest amount of unplanned purchases among shoppers for
the three selected products occurred among women shopping with
children.
4. At the ranges of prices and display spaces for the three
selected products, the shoppers' awareness of price and display
space changes was minimal.
TABLE XXXIX. REASONS FOR NOT PURCHASING TURNIPS
Reasons Number Percentage
Never buy 184 44
Seldom buy 72 17.2
Price is high 5 1.2
Poor quality 37 8.9
Didn't see display 11 2.6
Have at home 53 12.7
Not on shopping list 17 4.1
Change in menu 17 4.1
Miscellaneous 22 5.2
Total 418 100,
89
TABLE XL. REASONS FOR NOT PURCHASING BELL PEPPERS
Reasons Number Percentage
Never buy 121 29
Seldom buy 82 19.7
Price is high 44 10.6
Poor quality 57 13.7
Didn't see display 17 4.1
Have at home 48 11.5
Change in menu 17 4.1
Not on shopping list 13 3.1
Miscellaneous 18 4.2
Total 417 100
90
t .".
91
Such results are useful to the retailer who is interested in knowing
what type of consumer buys a certain product, thus enabling him to choose
the proper marketing strategy -- not only to increase sales through
unplanned purchases but also to create a better store atmosphere for the
consumers' satisfaction and enjoyment of shopping.
CHAPTER VII
SUMMARY AND IMPLICATIONS OF STUDY
Food retailing is an important segment of Hawaii's economy. The
national trend of consumer expenditures is toward an increasing number of
purchases without advance planning. The present study was based on
empirical information obtained from a group of 700 shoppers in a
representative supermarket in Honolulu, Hawaii, during a five-week test
period. It was hypothesized that consumers are aware of marketing
techniques; namely, multiple pricing and increased display space on
selected items.
The primary objectives of this study were to find the behavior
patterns of supermarket shoppers in Hawaii and their awareness of
merchandising techniques and to ascertain how multiple pricing and
increased display space influence the shopper in Hawaii to purchase
unplanned selected items among the many that are available, and to
determine the impact these merchandising practices have on sales. The
selections of the test items were based on Boyer's study and the sales
of the supermarket chain and they were representative of Hawaii. The
high demand item was navel oranges; the semi-demand item was turnips;
and the low demand item was bell peppers.
Answers gi.ven to the following questions showed similarities
between shoppers in Hawaii and on the Mainland.
1. Who does the family shopping?
2. How many marketing trips are taken per week?
3. How many stores are frequented per week?
4. How many food ads are read per week?
93
Some findings that showed a marked deviation from Mainland findings
were:
1. A greater number of Island shoppers used a written shopping
list.
2. More women shoppers were fully employed.
3. Household incomes of Island shoppers are higher than on the
Mainland.
4. Island shoppers come from more varied ethnic groups.
Aside from these few deviations, Island shoppers are very similar to
their counterparts on the Mainland.
The effectiveness of the marketing techniques, multiple pricing and
increased display space, was identified with increased sales and with
the number of unplanned purchases the marketing techniques induced. The
results with respect to unplanned purchases were:
1. The greatest number of unplanned purchases of navel oranges
occurred in the week where the test variable was increased
display space (35.9 percent of all purchases unplanned).
2. The greatest number of unplanned purchases of turnips occurred
in the week where the test variable was multiple pricing. (21.1
percent of all purchases were unplanned.)
3. The greatest number of unplanned purchases of bell peppers
occurred in the week where the test variable was multiple
pricing. (44.7 percent of all purchases were unplanned.)
There was no indication that increased display space or multiple
pricing added significantly to the number of unplanned purchases.
It was concluded that the shoppers' awareness of price and display
94
changes is minimal at the ranges of prices and display spaces for navel
oranges, turnips, and bell peppers.
The statistical test of analysis of variance was used to see if
weekly volume movements were significantly different. The findings were:
1. The sale of navel oranges could be increased by increasing the
display space even at the rather high cost of 30¢ per pound.
2. The sale of turnips was increased somewhat by multiple pricing.
3. Each of the variables, multiple pricing and increased display
space, increased the sale of bell peppers slightly.
It was also concluded that of the two sales methods for bell
peppers (cost per unit versus cost per pound), the selling of bell
peppers by the unit was definitely superior. Bell peppers are very
light in weight, so their unit cost is low in comparison to their cost
per pound. The shoppers in this store responded very favorably to unit
cost pricing of this item, though their rationalizations of their buying
decisions were for the most part erroneous.
SUGGESTIONS FOR FURTHER RESEARCH
The findings of this study suggest that further research would be
useful in answering some of the questions that were beyond the scope of
this study. For a future study, not only should these marketing
techniques be applied for different food products, but greater differences
in the ranges of display spaces and prices should also be experimented
with. The study could also be expanded on other aspects of marketing
research; for example, observing how the time of day affects unplanned
purchases.
95
Food retailing is a big business; and because of this, a
comprehensive study of Island shoppers' characteristics and shopping
patterns would be extremely beneficial -- to the retailer interested in
increased profits and better service, and to the customer interested in
ease of shopping and satisfaction in purchases.
APPENDIX
CONSUMER QUESTIONNAIRE
1. Who does the family shopping?
Alone--------Spouse;.-. _Parents, _Relatives------Friends
-~-----Child (ren) _
97
No.Date:...- _Time:...- _
2. Do store specials or promotions cause you to buy more?
Yes. _No----
3. Do you have a written shopping list?
Yes---No _ If no, Do you have a mental list?Yes, _No _
4. How often do you shop at a supermarket per week?
Once. _Twice. _
Three~----Four-----Five or more----
5. Is this a major or a filler shopping trip?
Regular _Filler _
6. How many supermarkets do you shop at per week?
One. _Two. _Three. _Four----Five or more'._---
98
7. Why do you shop at this store?
Convenient. __Near, _On the way. __Quality. _Service
~------
More specials _Low prices __Friends---------Know where everything is _Others'-----------
8. How far do you live from the store?
1 mile or less'-----1-2 miles2-3 miles·-------3-4 miles, _5 or more miles _
9. Do you have a food budget?
Yes-------No-------10. Do you usually stay within the budget?
Yes. _No _
11. How many grocery ads do you read per week?
l~ _2. _3 _4: _5 or more _
12. Do you use the ads to make a shopping list?
Yes----No-----13. Do trading stamps cause you to buy more?
Yes------No, _Indifferent __
14. Had you planned to buy oranges?
yes _No - If no, skip #15, go to #16.
99
15. If yes, how much?
lbs. or units-----.; ------16. Did you buy oranges?
Yes, _No----- If no, skip #17, go to #18.
17. If yes, how many did you buy?
____...;lbs. or units
What prompted you to purchase oranges (unplanned) or to purchaseadditional (or less) oranges after you were inside the store?
Display _Posters, sign, card~ __Price, _Qualityo:--~ _Tag, label:..- _Clerk suggestion~ _Always buy it (remembered it) _Other--------Skip #18, go to #19.
18. If no, why?
Never buy _Seldom buy _Price is high~ _Poor quality _Didn't see it'---Have at home._---Change in menu__~_Not on shopping list. _Miscellaneous---
19. Did you notice any difference between purchasing oranges this weekand last week? (for example, price, display, size, quality, etc.)
Yes ' No _
If yes, what?
P+-----; P"" _ DS+ ; DS"" Q+ Q"'__
Other _
100
20. Had you planned to buy turnips?
YesNo If no, skip 1121, go to 1122.
21. If yes, how much?
1bs. or units
22. Did you buy turnips?
Yes'No If no, skip 1123, go to 1124.
23. If yes, how much?
1bs. or units
What prompted you to purchase turnips (unplanned) or to purchaseadditional (or less) turnips after you were inside the store?
Disp1a.y _Posters, sign, card~' __Price'--------Quality~~ _Tag, 1abe1:..-~ _Clerk suggestion~ __Always buy it (remembered it) _Other--------Skip #24, go to #25.
24. If no, why?
Never buy. __Seldom buy. _Price is high __Poor Qua1ity _Didn't see it _Have at home. _Change in menu, __Not on shopping 1ist, _Miscel1aneous _
25. Did you notice any difference between purchasing turnips this weekand last week? (for example, price, display, size, quality, etc.)
Yes . No _
If yes, what?
P+ ; P-l- _ DS+ ; DS...-l- ; Q+ _ Q-l- _
101
Other' ....
26. Had you pla~ed to buy bell peppers?
YeS----No. _ If no, skip 1127, go to 1128.
27. If yes, how much?
_______lbs. or un.its
28. Did you buy bell peppers?
Yes, _No _ If no, skip 1129 and go to 1130.
29. If yes, how much?
____....;lbs. or un.its.
What prompted you to purchase bell peppers (unplanned) or topurchase additional (or less) bell peppers after you were insidethe store?
Display _Posters, sign, card, ___Price. _
Qua1ity~::__-----Tag, label~ _Clerk suggestion~__~Always buy it (remembered it) _Other--------
30. If no, why?
Never buy _Seldom buy _Price is high~ _Poor quality _Didn't see it'-----Have at home. _Change in menu~ _Not on shopping list~ _Miscellaneous _
31. Did you notice any difference between purchasing bell peppers thisweek and last week? (for example, price, display, size, quality,etc. ) Yes No' .
pt_' ; P"' _
Other, _
DSt _ DS"' ; Qt _ Q"'---
32. Sex:M~' _
33. Mar±ta1 Status
Married. _Sing1e, _Other, _
34. Age
Under 20 _20-34 _35-49 _50-64, _Over 65, _
35. Race
Caucasian~ _JapaneSe _Chinese._----Fi1ipino, _Hawaiian-------Portugese. _Korean _Other'------
36. Education
F _
102
1-7 grade schoo1~_~ _Finished grade s choo1~ _1-3 years high schoo1~ __Graduated high schoo1 _1-3 years co11ege ___Graduated co11ege~ __Beyond co11ege. _
37. Occupation
Housewife, _Professiona1 __Se1f-emp1oyed _Ski11ed~~ __Semi-ski11ed~ _C1erical __Service~ _StudentOther -------------
38. Members of the househo1d __
39. Gross household income
a. $365 or over $19,000 or overb. 307-365 16,000-18,999c. 249-307 13,000-15,999d. 192-249 10,000-12,999e. 134-192 7,000-9,999f. 77-134 4,000-6,999g. 77 or under Under 4,000
40. Number of magazines subscribed to or purchased regularly?
103
1~__; 2:.-..__; 3. ; 4, _ 5 or more:.-..__
41. Number of newspapers subscribed to or purchased regularly?
1~__; 2:.-..__ 3 or more:.-..__
42. If so, what are the names?
Honolulu Advertiser __Honolulu Star Bu11etin:.-.. _Others:.-.. _
104
APPENDIX TABLE I. AGE VS. OCCUPATIONS OFSHOPPERS INTERVIEWED
UnderOccupation 20 20-34 35-49 50-64 Over 65
Housewife 2 100 90 46 17
Professional 62 47 8 1
Self-employed 3 4 3 2
Skilled 11 20 1 2
Semi-skilled 8 7 6
Clerical 2 36 28 8 4
Service 2 48 24 4 1
Student 6 50 11
Other 1 1 10 5 2
105
APPENDIX TABLE II • ETHNIC GROUPS VS. HOUSEHOLD INCOME. ~,.
OF SHOPPERS INTERVIEWED
Ethnic Under $4000- $7000- $10,000- $13,000- $16,000- OverGroup $4000· 6999 9999 12,999 15,999 18,999 19,000
Caucasian 32 19 69 65 40 10 30
Japanese 49 6 54 124 44 24 40
Filipino 1 14 1 2
Hawaiian 1 4 3
Korean 1 4 1
Portugese 1 4
Chinese 2 2 2 5 5 10 10
Other 1 1 3 1 12 1 2
APPENDIX TABLE III. HOUSEHOLD INCOME VS. OCCUPATIONS OF SHOPPERS INTERVIEWED
Under $4000- $7000- $10,000- $13,000 $16,000 OverOccupation $4000 6999 9999 12,999 15,999 18,999 19,000
Housewife 4 30 64 79 39 10 29
Professional - 3 10 10 40 15 40
Self-employed - - 2 2 2 1 5
Skilled - - - 10 15 1 8
Semi-skilled 2 - 9 10
Clerical 10 2 17 46 3
Service 9 - 20 29 5 16
Student 60 7
Other 1 1 15 15 1 2 1
I-'o0'
107
APPENDIX TABLE IV. HOUSEHOLD INCOME VS. AGE OFSHOPPERS INTERVIEWED
Under $4000- $7000- $10,000- $13,000 $16,000 OverAge $4000 6999 9999 12,999 15,999 18,999 19,000
Under 20 10 2 1
20-34 50 20 79 125 40 7 15
35-49 6 7 50 44 54 33 47
50-64 1 6 6 32 11 5 20
Over 65 19 7 1 1 1
APPENDIX TABLE V. THE NUMBER OF SHOPPErS IN EACH HOUSEHOLDINCOME LEVEL AND THE NUMBER OF SHOPPERS WHO BOUGHT NAVEL
ORANGES, TURNIPS, OR BELL PEPPERS
Navel BellIncome Level Oranges Turnips Peppers
$19,000 or more (83) 17 17 15
16,000-18,999 (45) 7 16 15
13,000-15,999 (105) 24 23 28
10,000-12,999 (201) 48 23 31
7,000-9,999 (137) 26 24 27
4,000-6,999 (43) 12 20 8
4,000 or less (86) 6 17 26
SUMMARY OF THE RESULTS OF THE ANALYSIS OF VARIANCE
Navel Oranges (all) - Not significant
Navel oranges (C & 2) - Not significant
Navel oranges (C & 3) - a = (0.25 and 0.50)
Navel oranges (C & 4) - Not significant
108
THE ANALYSIS OF VARIANCE TABLEFOR NAVEL ORANGES (all)
Source of Sum of Degrees EstimateVariation Square of Freedom of Variance F
BetweenTreatments 3675.00 3 1225.00 .53(week)
Betweendays 13696.43 6 2282.74 .98
Residual 41875.00 18 2326.38
Total 59246.43 27
VI = 18, we find the 5% and 1% points of F to be 3.16
and 5.09 respectively. We conclude, therefore, that
the difference between treatments and days are hardly
significant.
109
ANALYSIS OF VARIANCE TABLE
C 2 3 4 Ti T.2 2J. ~j
S 4900 (70) 210 3600 (-60) 80 1225 (85) 175 (0) (0) 140 45 2025 9725
M 14400 (-129) 20 2500 (-50) 90 (0) (0) 140 1225 (35) 175 -135 18225 18125
T 4900 (-70) 70 1600 (-40) 100 4900 (70) 210 1225 (-35) 105 -75 5625 12625
W 1225 (35) 175 4900 (70) 210 (0) (0) 140 (0) 140 105 11025 6125
T (0) (0) 140 1225 (-35) 105 1225 (-35) 105 400 (-20) 120 -90 8100 2850
F 1225 (-35) 105 4900 (70) 210 1225 (35) 175 (0) (0) 140 70 4900 7350
S (0) 140 1225 (35) 175 (0) (0) 140 1225 (35) 175 70 4900 2450
Tj -120 -10 105 15 T = 10 L Ti259250i
= 54800
T2 14400 100 11025 225 r1j 2
= 25750LjLjXij2
Xij2 26650 19950 8575 4075 59250
Transfer to a new origin at 140.
............o
111
Consequently:
(i) Total sum of squared deviation t j Xij - T2
/N =
59250 - (10)2 = 59259 - 3.57 = 59246.4328
(ii) Sum of squares for treatments, j (Tj 2/nj ) - T2/N =
25750 - 3.57 = 3678.57 - 3.57 = 3675.007
(iii) Sum of squares for day E (Ti2/ni) - T2/N =i
54800 - 3.57 = 13700 - 3.57 = 13696.434
(iv) Residual sum of square = 59246.43 - 3675.00 - 13696.43 = 41875.00
112
THE ANALYSIS OF VARIANCE TABLE FOR NAVEL ORANGES (C&2)
Source of Sum of Degree EstimateVariance Squares of Freedom of Value F
BetweenTreatments 864.29 1 864.29 .29(weeks)
BetweenDays 26692.86 6 4448.81 1.50
Residual 1783.71 6 2972.61
Total 45392.86 13
Not Significant
ANALYSIS OF VARIANCE TABLE
C 2 TiTi 2 X:i.j 2
S 4900 (70) 210 3600 (-60) 80 10 100 8500
M 14400 (-120) 20 2500 (-50) 90 -170 28900 16900
T 4900 (-70) 70 1600 (-40) 100 -110 12100 6500
W 1225 (35) 75 4900 (70) 210 105 11025 6125
T 0 (0) 140 1225 (-35) 145 -35 1225 1225
F 1225 (-35) 105 4900 (70) 210 35 1225 1225
S 0 (0) 140 1225 (35) 175 35 1225 1225
Ti -120 -10 T = -130 L Ti2 46600i= 55800
T2 14400 100L T;2 =
L L Xij2i J14500 i j
X:i.j 2 26650 19950 46600
Transfer to a new origin at 140.~~w
114
Consequently,
(i) Total sum of squared deviation, r j ~j2 - T2/N
= 46600 - (-130)2 = 46600 - 16900 =14 14
4660 - 1207.14 = ~5392.86
(ii) Sum of squares for treatments, j (Tj 2/Nj ) - T2/N =
14500 - 1207.14 = 2071.14 = 864.297
55800 - 1207.14 = 27900 - 1207.14 = 26692.86,2
(iv) Residual sum of squares, 45392.86 - 864.29 - 26692.86 = 17835.71
115
THE ANALYSIS OF VARIANCE TABLE FOR NAVEL ORANGES (C&3)
Source of Sum of Degree EstimateVariance Squares of Freedom of Value F
Source ofVariance 3616.07 1 3616.07 1.23
BetweenTreatments 13921.43 6 2320.24 .79(weeks)
Residual 17671.43 6 2945.23
Total 35208.93 13
ANALYSIS OF VARIANCE TABLE
C 3 Ti T·2 xtj2. ~
S 4900 (70) 210 1225 (35) 175 105 11025 6125
M 14400 (-120) 20 0 (0) 140 -120 14400 14400
T 4900 (-70) 70 4900 (70) 210 0 0 9800
W 1225 (35) 175 0 (0) 140 35 1225 1225
T 0 (0) 140 1225 (-35) 105 -35 1225 1225
F 1225 (-35) 105 1225 (75) 175 0 0 2450
S 0 (0) 140 0 (0) 140 0 0 0
Ti -120 105 T = -15E Ti 2 = 35225
i27875
T2 14400 11025 ~ Tj2 =25425 E E xt 2
i j j
xtj2 26650 8575 35225
Transfer to new origin at 140. ........0\
117
Consequently,
(i) Total sum of squared deviati.ons, E E Xij2 -T2/N =i j
35225 - 225 = 35225 - 16.07 = 35208.9314
(ii) Total sum of squares for treatments, 1 (Tj 2/nj > -T2/N =
25425 - 16.07 = 3632.14 - 16.07 = 3616.077
27875 - 16.07 = 13937.50 = 13921.432
(iv) Residual sum of squares, 35208.93 - 3616.07 - 13921.43 = 1761.43
118
THE ANALYSIS OF VARIANCE TABLE FOR NAVEL ORANGES (C&4)
Source of Sum of Degree EstimateVariance Squares of Freedom of Value F
BetweenTreatments 1301.79 1 1301. 79 .49(weeks)
BetweenDays 12825.00 6 2137.50 .81
Residual
Total
15810.71
29937.50
6
13
2635.11
ANALYSIS OF VARIANCE TABLE
C 4 Ti T.2 Xij2J.
S 4900 (70) 210 0 (0) 140 70 4900 4900
M 14400 (-12) 20 1225 (35) 175 -85 7225 15625
T 4900 (-20) 20 1225 (-35) 105 -105 11025 6125
W 1225 (35) 175 0 (0) 140 35 1225 1225
T 0 (0) 140 400 (-20) 120 -20 400 400
F 1225 (-35) 105 0 (0) 140 -35 1225 1225
S 0 (0) 140 1225 (35) 175 35 1225 1225
Ti 120 15 T = -105 I: T. 230725• J.
J.= 27225
T2 14400 225I: T.2 =i J14625 I: I: Xij2
2 26650 4075 30725i j
xij
Transfer to new origin at 140.~~\0
120
Consequently,
(i) Total sum of squared deviations, r f Xij2 - T2/N =
30725 ~ (105)2 = 30725 ~11025 = 30725 - 787.50 = 29937.5014 14
(ii) Total sum of squares for treatments, 1:. (Tj 2/nj ) - T2/N =j
14625 - 787.50 = 2089.29 - 787.50 = 1301.797
"(iii) Total sum of squares for days, 1: (Ti2/ni) - T2/n =i
27225 - 787.50 = 13612.50 - 787.50 = 12815.002
(iv) Residual sum of squares = 29937.50 - 1301.79 - 12825.00 = 15810.71
SUMMARY OF THE RESULTS OF THE ANALYSIS OF VARIANCE
Turnips (all) - Not significant
Turnips (C & 2) - Significant at 75% = 0.25
Turnips (C & 3) - Not significant
Turnips (C & 4) - Not significant
121
122
THE ANALYSIS OF VARIANCE TABLE FOR TURNIPS (all)
Source of Sum of Degrees EstimateVariation Squares of Freedom of Variance F
BetweenTreatments 152.39 3 50.79 .23(weeks)
BetweenDays 4209.18 6 701.53 3.11
Residual 4056.79 18 225.37
Total 9868.68 27
VI = 3, v2 = 18, we find the 5 percent and 1 percent points of
F to be 3.16 and 5.09 respectively. We conclude, therefore, that
the difference between treatment is hardly significant (at the
5 percent level), while that between days is almost so (almost
significant at the 5 percent level).
ANALYSIS OF VARIANCE TABLE
C 2 3 4 Tj Ti Xij2
S 0 (0) 30 25 (5) 25 0 (0) 30 25 (-5) 35 0 0 50
M 25 (5) 25 100 (10) 20 1024 (-32) 62 225 (-15) 45 -32 1024 1374
T 25 (-5) 35 400 (-20) 50 0 (0) 30 400 (-20) 50 -45 2024 825
w 1024 (-32) 62 1444 (-38) 68 400 (-20) 50 0 (0) 30 -90 8100 2868
T 100 (-10) 40 100 (-10) 40 25 (5) 25 400 (-20) 50 -35 1225 625
F 100 (10) 20 841 (-29) 59 2025 (-45) 75 625 (-25) 55 -89 7921 3591
S 3600 (-60) 90 2500 (-50) 80 100 (-10) 40 1600 (-40) 70 -160 25600 7800
Tj -92 -132 -102 -125 T = 451 E Ti2 = 17133i51917
T2 8464 17424 10404E T.2 =
15625 i J51917 E E Xfj2
i j
~j2 4874 5410 3574 3275 17133
Transfer to new origin at 30.....r-.>UJ
Consequently,
(i) Total sum of squared deviations, E E Xij2 - T2/N =i j
17133 - (451)2/28 = 17133 - 203401/28 =
17133 - 7264.32 = 9868.68
(ii) Sum of squares for treatments, E (Tj 2/nj ) - T2/N =j
51917 - 7264.32 = 152.397
124
(iii)
45894 - 7264.32 = 11,473.5 - 7264.32 = 4209.184
(iv) Residual sum of squares, 9868.68 - 4465.26 = 5403.42
125
THE ANALYSIS OF VARIANCE TABLE FOR TURNIPS (C&2)
Source of Sum of Degrees EstimateVariation Squares of Freedom of Value F
BetweenTreatments 114.29 1 114.28 8.05(weeks)
BetweenDays 5734 6 955.66 67.34
Residual 851. 72 6 14.19
Total 6700 13
VI = 1, v2 = 6 we find the 5 percent and 1 percent points of F to
be 5.99 and 13.74 respectively. We conclude therefore that the
difference between treatments is significant (at the 5 percent
level) and that between days is highly significant (at the
1 percent level).
ANALYSIS OF VARIANCE TABLE
C 2 Ti T.2 21 xtj
S 0 (0) 30 25 (5) 25 5 25 25
M 25 (5) 25 100 (10) 20 15 225 ·125
T 25 (-5) 400 (-20) 50 -25 625 425
W 1024 (-32) 62 1444 (-38) 68 -70 4900 2468
T 100 (-10) 40 100 (-10) 40 -20 400 200
F 100 (10) 20 841 (-29) 59 -19 361 941
S 3600 (-60) 90 2500 (-50) 80 -110 12100 6100
Tj -92 -132 T = -224 ~ Ti 2 = 102841
18636
T2 8464 17424L T.2 =i J
25888 L L xt· 2i j J
2 4874 5410 10284xtj
Transfer to new origin at 30. I-'N0\
Consequently,
(~) Total sum of squared deviations, E E xtj2 - T2/N =i j
10284 ~ (224)2 = 10284 -50176 = 10284 - 3584.00 = 6700.0014 14
(~~) Sum of squares for treatments, j (Tj2/nj ) - T2/N =
25888 - 3584 = 3698.29 = 114.297
18636 - 3584 = 9318 - 3584 = 57342 -
(~v) Res~dua1 sum of squares, 6700 - 114.29 - 5734 = 851.71
127
128
THE ANALYS~S OF VARIANCE TEST FOR TURNIPS (C&3)
Source of Sum of Degrees EstimateVariation Squares of Freedom of Variance F
BetweenTreatments 7.25 1 7.25 .01(weeks)
BetweenDays 2115.71 6 352.62 .58
Residual
Total
3636.75
5759.71
6
13
606.12
ANALYSIS OF VARIANCE TABLE
C 3 Ti T.2 2l. ~j
S 0 (0) 30 0 (0) 30 0 0 0
M 25 (5) 25 1024 (-32) 62 -27 729 1049
T 25 (-5) 35 0 (0) 30 -5 25 25
W 1024 (-32) 62 400 (-20) 30 -52 2704 1424
T 100 (-10) 40 25 (5) 25 -5 25 125
F 100 (10) 20 2025 (-45) 75 -35 1225 2125
S 3600 (-60) 90 100 (-10) 40 -70 4900 3700
Tj -92 -102 T = 194 L T. 2 = 8448i i9608
T2L T.2 =
8464 10404 i J18868 L L xtj2
i j
xtj2 4874 3574 8448
Transfer to a new origin at 30. ....N\0
Consequently,
(i) Total sum of squared deviations, rf X±j 2 - T2/N =
8448 -'(194)2 = 8448 ~ 37636 = 8448 - 2688.29 = 5759.7114 14
(ii) Sum of squares for treatments, j (Tj2/nj) - T2jN =
"18868 - 1344.14 = 2695.54 - 2688.29 = 7.257 --
130
(iii)
9608 - 1344.14 = 4804 - 2688.20 = 2115.712
(iv) Residual sum of squares, 5759.71 - 7.25 - 2115.71 = 3636.75
131
THE ANALYS~S OF VARIANCE TABLE FOR TURNIPS (C&4)
Source of Sum of Degrees EstimateVariation Squares of'Freedom of Variance F
BetweenTreatment 77.79 1 77.79 .28(weeks)
BetweenDays 3086.00 6 514.33 1.90
Residual
Total
1621. 71
4785.50
6
13
270.28
ANALYSIS OF VARIANCE TABLE
C 4 Ti T.2. ~j2.J..
S 0 (0) 30 25 (-5) 35 -5 25 25
M 25 (5) 25 225 (-15) 45 -10 100 250
T 25 (-5) 35 400 (-20) 50 -25 625 425
W 1024 (-32) 62 0 (0) 30 -32 1024 1024
T 100 (-10) 40 400 (-20) 50 -30 900 500
F 100 (10) 20 625 (-25) 55 -15 225 725
S 3600 (-60) 90 1600 (-40) 70 -100 10000 5200
Tj -92 -125 T = 217L Ti 2. = 8149i12899
T2. 8464 15625L T.2. =
Ji24089 L L ~j2.
i j
~j2. 4874 3275 8149
Transfer to a new origin at 30.
....WN
Consequently,
133
(i) Total sum of squared deviations, E E Xfj2 - T2/N =i j
8149 ~ (217)2 = 8149 ='47089 = 8149 - 3363.50 = 4785.5014 14
(ii) Sum of squares for treatments, j (Tj2/nj) - T2/N =
24089 - 3363.50 = 3441.29 - 3363.50 = 77.797
(iii)
'12899 - 3363.50 = 6449.50 - 3363.50 = 3086.002
(iv) Residual sum of squares, 4785.50 - 77.79 - 3086.00 = 1621.71
134
SUMMARY OF THE RESULTS OF THE ANALYSIS OF VARIANCE. .
Bell Peppers (all)
Bell Peppers C & 2 01P)
Bell Peppers C & 3 (IDS)
- Difference between weeks
Significant at 75% (~ = 0.25)
- Difference between weeks
Significant at 90% (~ = 0.10)
- Difference between weeks
Significant at 90% (~ = 0.10)
Bell Peppers C & 4 (MP & IDS) - Difference between weeks
Significant at 75% (~ = 0.25)
135
ANALYS~S OF VARIANCE TABLE FOR BELL PEPPERS (all)
Source of Sum of Degrees EstimateVariation Squares of Freedom of Variance F
BetweenTreatments 476.98 3 158.99 2.03*(weeks)
BetweenDays 336.27 6 56.05 .72
Residual
Total
1407.82
2221.07
18
27
78.21
Significant at 75 percent.
ANALYSIS OF VARIANCE TABLE
C 2 3 4 Ti Ti
2 ~j2
S 0 (0) 30 225 (15) 45 25 (-5) 25 0 (0) 30 10 100 250
M 225 (-15) 15 25 (-5) 25 625 (25) 55 0 (0) 30 5 25 875
T 25 (-5) 25 0 (0) 30 225 (15) 45 225 (15) 45 25 625 475
W 225 (-15) 15 100 (-10) 20 81 (-9) 21 0 (0) 30 -34 1156 406
T 25 (-5) 25 0 (0) 30 25 (5) 35 0 (0) 30 0 0 50
F 100 (-10) 20 100 (10) 40 25 (5) 35 0 (0) 30 5 25 225
S 100 (10) 40 4 (2) 32 100 (10) 40 225 (15) 45 37 1369 429
Tj -60 12 46 30 T = 48 E Ti2 = 2710i 3300
T2 3600 144 2116 900 ~ Tj2 =
J.6760 E E ~.2
i j J
~j2 700 454 1106 450 2710
Transfer to new origin at 30.~W0\
Consequently,
137
(i) Total sum of squared deviations, E E ~.2 - T2/N =i j J
2710 ~ (37)2 = 2710 -'1369 = 2710 - 488.93 = 2221.0728 28
6760 - 488.93 = 965.71 - 488.73 = 476.987
(iii)
3300 - 488.73 = 825.00 - 488.73 = 336.27--r(iv) Residual sums of square, 2221.07 - 476.98 - 336.27 = 1407.82
138
ANALYSIS OF VARIANCE TABLE FOR BELL PEPPERS (C&2)
Source of Sum of Degrees EstimateVariation Squares of Freedom of Variance F
BetweenTreatments 193.86 1 193.86 4.88*(weeks)
BetweenDays 666.00 6 111.00 2.80
Residual
Total
238.14
1098.00
6
13
39.69
ANALYSIS. OF VARIANCE TABLE
C 2 Ti Ti 2 Xij2
S 0 (0) 30 225 (15) 45 15 225 225
M 225 (-15) 15 25 (-5) 25 -20 400 250
T 25 (-5) 25 0 (0) 30 -5 25 25
W 225 (-15) 15 100 (-10) 20 -25 625 325
T 25 (-5) 25 0 (0) 30 -5 25 25
F 100 (-10) 20 100 (10) 40 0 0 200
S 100 (10) 40 4 (2) 32 12 144 104
Tj -40 12 T = -28 ~ Ti2 = 11541.
1444
T2 1600 144I: Tj 2 =i
1744 I: I: x·j2i j 1.
X.. 2 700 454 11541.J
Transfer to new origin at 30.I-'IJ,)\0
Consequently,
(i) Total sum of squared deviations, f j Xij2 -T2/N =
1154 -'(25)2 = 1154 ~'784 = 1154 - 56 = 1098~0014 14
1744 - 56 = 249.14 - 56 = 193.86--r
~ - 56= 722 - 56 = 666.002
(iv) Residual sum of squares, 1098.00 - 193.86 - 666.00 = 238.14
140
141
ANALYSIS OF VARIANCE TABLE FOR BELL PEPPERS (C&3)
Source of Sum of Degrees EstimateVariance Squares of Freedom of Variance F
BetweenTreatments 528.29 1 528.29 4.77(weeks)
BetweenDays 610.43 6 101. 74 .92
Residual
Total
664.71
1803.43
6
13
110.78
ANALYSIS OF VARIANCE TABLE
C 3 Ti Ti2 Xij2
S 0 (0) 30 25 (-5) 25 -5 25 25
M 225 (-15) 15 625 (25) 55 10 100 850
T 25 (-5) 25 225 (15) 45 10 100 250
W 225 (-15) 15 81 (-9) 21 -24 576 306
T 25 (-5) 25 25 (5) 35 0 0 50
F 100 (-10) 20 25 (5) 35 -5 25 125
S 100 (10) 40 100 (10) 40 20 400 200
Tj -40 46 T = 6~ Ti 2 = 1806J.1226
T2 1600 2116I: Tj 2 =i
3716 ~ ~ ~j2J. J
~j2 700 1106 1806
Transfer to new origin at 30.
......~N
Consequently,
143
(i) Total sum of squared deviations, E E Xij2 - T2/N =i j
1806 -'~ = 1806 - 2.57 ='1803.4314
'~ - 2.57 = 530.86 - 2.57 = 528.297
.~ - 2.57 = 613 - 2.57 = 610.432
(iv) Residual sum of squares, 1803.43 - 528.29 - 610.43 = 664.71
144
ANALYSIS OF VARIANCE TABLE FOR BELL PEPPERS (C&4)
Source of Sum of Degrees EstimateVariance Squares of Freedom of Variance F
BetweenTreatments 171.43 1 171.43 3.13(weeks)
BetweenDays 642.86 6 107.14 1.96
Residual
Total
328.57
1142.86
6
13
54.76
ANALYSIS OF VARIANCE TABLE
C 4 Ti T·2 ~j21.
S 0 (0) 30 0 (0) 30 0 0 0
M 25 (-5) 25 0 (0) 30 -15 225 225
T 25 (-5) 25 225 (15) 45 10 100 250
W 225 (-15) 15 0 (0) 30 -15 225 225
T 25 (-5) 25 0 (0) 30 -5 25 25
F 100 (-10) 20 0 (0) 30 -10 100 100
S 100 (10) 40 225 (15) 45 25 625 325
Tj -40 30 T = -10L Ti2 =
1150i1300
T2 1600 900~ Tj 2 =1.
2500 L L xt 2i j j
~j2 700 450 1150
Transfer to new origin at 30. ....~VI
Consequently,
(i) Total sum of squared deviations, f j ~j2 -T2/N =
1150 ~ (10022 = 1150 - 7.14 = 1142.8614
(ii) Sum of squares for treatments, j (Tj 2/nj ) - T2/N =
2500 - 7.14 = 178.57 - 7.14 = 171.43""1"4"
1300 - 7.14 = 650 - 7.14 = 642.862
(iv) Residual sum of squares, 114.86 - 171.43 - 642.86 = 328.57
146
CHI-SQUARE TESTS
148
Chi-square Test(Navel Oranges (C&2)
Week Variable Unplanned PlannedPurchases Purchases
Week C (12.2>13 (27.8>27 40
Week 2 Multiple (11.8>11 (27.2)28 39pricing
24 55 79
FrequenciesObserved Theoretical
f o f t
13112728
D.F. 1ex: = 0.10
12.211.827.827.2
.8-.8-.8
.8
.64
.64
.64
.64
.052
.054
.023
.024
x2 = .158
Ho - No difference in the number of unplanned purchases in controlweek and week two.
Decision: Accept Ho
Chi-square Test(Navel Oranges C&3)
Week Variable Unplanned PlannedPurchases Purchases
Week C (13.8h3 (26.2)2740
Week 3 In,creased (13.2)14 (25.8)25 39Display Space
27 52 79
149
FrequenciesObserved Theoretical
f o f t
13142725
D.F. = 10:: = 0.10
13.813.226.225.8
-.8.8.8
-.8
.64
.64
.64
.64
.046
.048
.024
.025
x2 = .143
Decision: Accept Ho
Chi-square Test(Navel Oranges C&4)
Week Variable Unplanned PlannedPurchases Purchases
Week C (12.8)13 (27.2)27 40
Multiple PricingWeek 4 "& Increased (12.3>12 (26.7)27 39
Display Space
25 54 79
150
FrequenciesObserved Theoretical
fo ft
13122727
D.F. = 1ex = 0.10
12.812.327.226.7
.2-.3-.2
.3
.04
.09
.04
.09
.0031
.0073
.0014
.0033
x2 = .0151
Decision: Accept Ho
Chi;"'square Test(Turnips C&2)
Week Variable Unplanned PlannedPurchases Purchases
Week C (11. 2)14 (27.8)25 39
Week 2Multiple
(10.8) 8 (27.2)30Pricing 38
22 55 77
151
FrequenciesObserved Theoretical
f o f t
148
2530
D.F. = 1a: = 0.10
11.210.827.827.2
2.8-2.8-2.8
2.8
7.847.847.847.84
.700
.726
.282
.288
x2 = 1.996
Decision: Accept Ho
Chi-square Test(TurnipsC&3)
Week Variable Unplanned PlannedPurchases Purchases
Week C (10.4)14 (28.6)25 39
Week 3Increased
(9.6) (26.4)30Display Space 6 36
20 55 75
152
FrequenciesObserved Theoretical
f o f t
146
2530
D.F. = 1ex: = 0.10
10.49.6
28.626.4
3.6-3.6-3.6
3.6
12.9612.9612.9612.96
1.251.350.450.49
x2 = 3.54
Decision: Reject Ho
Chi;"square Test(Turnips C&4)
Week Variable Unplanned PlannedPurchases Purchases
Week C (10.7)14 (28.3)25 39
Multiple PricingWeek 4 & Increased (10.3) 7 (27·7)31 38
Display Space
21 56 77
153
FrequenciesObserved Theoretical
f o f t (fo-f t )
147
2531
D.F. = 1ex = 0.10
10.710.328.327.7
3.3-3.3-3.3
3.5
10.8910.8910.8910.89
1.021.06
.38
.39
x2 = 2.85
Decision: Reject Ho
Chi-square Test(Bell PeppersC&2)
Week Variable Unplanned PlannedPurchases Purchases
Week C (15.8>14 (23.2)25 39
Week 2 Multiple(15.2>17 (22.8)21Pricing 38
31 46 77
154
FrequenciesObserved Theoretical
£0 £t ( £ _£ )2o t
14172521
D.F. = 1ex: = 0.10
15.815.223.222.8
-1.81.81.8
-1.8
3.243.243.243.24
.21
.21
.14
.14
x2 = .70
Decision: Accept EO
Ch:l-square Test(Bell PeppersC&3)
Week Variable Unplanned PlannedPurchases Purchases
Week C (14.5h4 (24.5)25 39
Week 3 Increased (14.5)15 (24.5)24Display Space 39
29 49 78
155
FrequenciesObserved Theoretical
f o f t
14152524
D.F. = 1ex = 0.10
14.514.524.524.5
-.5.5.5.5
.25
.25
.25
.25
.017
.017
.010
.010
x2 = .054
Decision: Accept Ho
Chi-square Test(Bell Peppers C&4)
Week Variable Unplanned PlannedPurchases Purchases
. . . . . . .
Week C (14.3)14 (24·7)25 39
Multiple PricingWeek 4 & Increased (12·7)13 (22.3)22 35
Display Space
27 47 74
156
FrequenciesObserved Theoretical
£0 £t
14132522
D.F. = 1ex: = 0.10
14.3·,12.724.722.3
-.3.3.3
-.3
.09
.09
.09
.09
.0062
.0070
.0036
.0040
x2 = .0208
Decision: Accept Ho
BIBLIOGRAPHY
A. Periodicals and Magazines
Blankenship, A. B. "Creativity in Consumer Research. 1I Journal ofMarketing, XXV (October, 1961), 34-38.
Cox, Keith. "The Responsiveness of Food Sales to Shelf Space Changesin Supermarkets. 1I Journal of Marketing Research, I (May, 1964),63-67.
Kollat, David T. and Ronald P. Willett. "Customer Impulse PurchasingBehavior."· Journal of Marketing Research, IV (February, 1967),21-31.
Morris, James L. and Robert J. Firch. liThe Analysis of SeparatePrice and Advertising Responses to Retail Grocery Specials. 1I
Proceedings 1967 Western Farm Economics Association, XL (July, 1967),47-52.
Namias, Jean. "Intentions to Purchase Compared with Actual Purchasesof Household Durables." Journal of Marketing, XXIV (July, 1959),60-64.
"Intentions to Purchase Related to ConsumerCharacteristics." Journal of Marketing, XXIV (July, 1960), 32-35.
Schapker, Ben L. "Behavior Patterns of Supermarket Shoppers."Journal of Marketing, XXX (Oct., 1966), 46-49.
Shaffer, James Duncan. "The Influence of 'Impulse Buying' or In-theStore Decisions on Consumers' Food Purchases." Journal of FarmEconomics, XLII (May, 1960), 317-323.
Stern, Hawkins. "The Significance of Impulse Buying Today." Journalof Marketing, XXXVI (April, 1962), 59-62.
B. Monographs and Special Studies
Dean, Gerald N. and Norman R. Collins. "World Trade in FreshOranges: Analysis of the Effect of European Economic CommunityPolicies." Giannini Foundation Monograph, XVIII (January, 1967), 9.
E. 1. DuPont DeNemours and Co. liThe Family Shopper." The 7thDuPont Consumer Buying Habits Study (1964).
IIToday's Purchases in Supermarkets."The 7th DuPont Consumer Buying Habits Study (1964).
158
Hawaii State Department of Health. Estimated Civilian Populationand Dwelling Units in Oahu Census Tracts '(July 1, 1967).
C. General Works
Bliss, Perry. ' Marketing and the Behavioral Sciences. Boston, 1967.
Brunk, H. D.' An Introduction to Mathematica1·Statistics. Boston,1960.
Dixon, Wilfred J. and Frank J. Massey, Jr. Introduction toStatistical Analysis. New York, 1957.
Kraft, Charles H. and Constance Van Eeden. A NonParametricIntroduction to Statistics. New York, 1965.
Massy, William F., Ronald E. Frunk and Thomas Lodah1. PurchasingBehavior and Personal Attributes. Philadelphia, 1968.
Ost1e, Bernard.Second Edition.
Statistics in Research.Ames, Iowa, 1964.
First Edition, 1954.
Siegel, Sidney. NonParametric Statistics for the Behavioral Sciences.New York, 1956.
Wessel, Robert H. and Edward R. Willett. Statistics as Applied toEconomics and Business. New York, 1961.
Top Related