See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/246926654

A Scale for Measuring Attitude Toward Private

Label Products and an Examination of Its
Psychological and...

Article in Journal of the Academy of Marketing Science October 1998

DOI: 10.1177/0092070398264003

CITATIONS READS

263 1,679

4 authors, including:

Scot Burton Richard G. Netemeyer

University of Arkansas University of Virginia
115 PUBLICATIONS 5,217 CITATIONS 66 PUBLICATIONS 8,552 CITATIONS

SEE PROFILE SEE PROFILE

Judith Anne Garretson Folse

Louisiana State University
56 PUBLICATIONS 1,413 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Research on Creativity View project

All content following this page was uploaded by Judith Anne Garretson Folse on 28 July 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
A Scale for Measuring Attitude
Toward Private Label Products and
an Examination of Its Psychological
and Behavioral Correlates
Scot Burton
University of Arkansas
Donald R. Lichtenstein
University of Colorado
Richard G. Netemeyer
Louisiana State University
Judith A, Garretson
University of Arkansas
A measure of consumers' atfftude toward private label
brands is developed, and its psychometric properties are
assessed. Predictions are then tested regarding relation-
ships between private label attitude and (1) latent percep-
tuat and sales promotion constructs, and (2) purchase
behaviors measured in afield setting. The measure is posi-
tively related to value consciousness, deat proneness, and
smart-shopper self-perceptions, and negatively related to
the propensity to be brand loyal and hold price-quality
perceptions. Predictive validity of the private label meas-
ure is supported by a positive relationship with private ta-
bet purchases from a grocery store shopping trip. Despite
a positive relationship between the latent constructs of pri-
vate label attitude and deal proneness, the consumer seg-
ment that allocated a high percentage of total purchases to
private label products made fewer purchases on sate or with
a coupon. Thesefindings suggest that consumers may choose
between price-related deals and private label purchases.
Journal of the Academy of Marketing Science.
Volume 26, No. 4, pages 293-306.
Copyright 9 1998 by Academy of Marketing Science.
Private label products, or store brands, currently repre-
sent more than 25 billion dollars in supermarket sales and
now exceed 20 percent of unit volume (HeUer 1997; Lie-
beck 1996). While accounting for less than one fifth of
dollar sales volume, private label brands were responsible
for one half of the increase in total supermarket dollar vol-
ume in 1996 (Wellman 1997). Some experts project that
the overall share of private label brands may eventually
grow to 40 percent or more of supermarket sales (DeNitto
1993a). The level of concern about private labels among
national brand manufacturers is summarized in a recent
Value Line article assessing the financial prospects of
firms in the food-processing industry: "Private label
brands remain the bane of companies in the food process-
ing industry, especially as store brands have limited the
shelf space for, and prevented more-than-modest price in-
creases of, nationally branded items" (Seligman
1995:1461). Such performance, coupled with the interest
in private label business of large discount retailers, sup-
ports optimistic predictions for further share increases for
private label brands within the United States (Kahn and
McAlister 1997; Ortega and Stern 1993).
Despite the substantial increase in private label activity,
the recognition of private label concerns by food manufac-
294 JOURNALOF THE ACADEMYOF MARKETINGSCIENCE
turers and retailers as a critical strategic issue, and a bur-
geoning literature on private label products in the business
press, there have been few scholarly studies that address
consumers' attitudes toward private label products. To
help address this void, the primary objective of this article
is to develop a multi-item measure that assesses the latent
construct of private label attitude and to offer various tests
associated with this measure. After development and ini-
tial testing in a pretest, the measure is evaluated by using
advocated tests of internal consistency and validity. Pre-
dictions pertaining to relationships between consumers'
attitudes toward private label brands and several latent
value-perception constructs (e.g., value consciousness,
price-quality perceptions), deal proneness, and other la-
tent constructs from within the consumer marketing do-
main (e.g., brand loyalty, risk averseness) are then tested.
Next, we assess predictive validity through an examina-
tion of relationships between private label attitude and ac-
tual private label and national brand purchase behaviors
assessed unobtrusively in the marketplace. We also com-
pare deal-related behavior (coupon usage, response to ad-
vertised sale items) observed in the marketplace for
consumer segments scoring both high and low on the pri-
vate label scale measure, as well as for those purchasing a
large or small percentage of private label brands. Finally,
relationships between the private label measure and demo-
graphic variables are assessed, and implications of find-
ings and possible directions for future research are offered.
BACKGROUND AND PREDICTIONS ABOUT
CONSUMERS' PRIVATE LABEL ATTITUDES
Private label brands are those sold under retailers' (or
wholesalers') own labels rather than the brand name of a
national manufacturer (cf. Boone and Kurtz 1995; Kotler
and Armstrong 1996). The focus of this article is on con-
sumers' attitude toward private label grocery products as a
whole, rather than any particular private label brand or pri-
vate labels for any specific grocery product category.
Thus, the private label attitude measure is viewed as a rela-
tively enduring construct that is sufficiently broad to be of
9 1
use in a general sense across grocery product categories.
As such, it is consistent with construct measures that as-
sess general consumer attitudes or tendencies (e.g., atti-
tude toward advertising in general rather than attitude
toward the ad).
Recent literature in the business press suggests three
broad sets of reasons that potentially underlie consumer
receptivity to private label products. Figure 1 offers a con-
ceptual overview of predicted relationships between pri-
vate label attitudes and latent constructs from within each
of these three domains. First, attitudes toward private label
brands are associated with consumers' orientation toward,
and perception of, price. For example, consumers may
view private label products positively due to a desire to pay
low prices (i.e., price consciousness), or a strong desire to
maximize the ratio of quality received to the price paid
(i.e., value consciousness), or view them negatively be-
FALL1998
cause they believe that price is a strong indicator of quality
(i.e., a price-quality schema). Second, articles in the busi-
ness press also allude to possible relationships with other
common marketing constructs such as (lower) consumer
brand loyalty and risk averseness. Third, the literature also
speculates that there is some overlap between consumers'
utility derived from acting on promotional deals and pri-
vate label attitude. A review of relevant literature and justi-
fication for these respective relationships are presented in
the three sections that follow.2
Relationships Associated With
Price-Perception Constructs
Figure 1 first addresses linkages between consumers'
price perceptions and their attitudes toward private label
products. While increases in the market share of private la-
bel brands have generally been linked to issues associated
with price, speculations about how the various aspects of
price may relate to private label attitude have been ex-
pressed in various ways (e.g., Deveny 1993; Kirk 1992).
For example, the absolute (low) level of the price of private
label brands sometimes is viewed as the key factor. That is,
consumers with favorable attitudes toward private label
brands are extremely price conscious and tend to focus al-
most exclusively on paying low prices, essentially mini-
mizing or disregarding other factors in brand evaluations.
Various findings from the studies on the aggregate share of
private labels offer some support for an explanation of this
price consciousness. For instance, analyses show that as
aggregate disposable income goes down, private label
share increases, even though there is no change in overall
levels of perceived product quality (Hoch and Banerji
1993). Thus, although the quality levels of alternative
brands are unchanged, the loss of income alters consumer
purchase behavior in favor of more purchases of private la-
bel products, presumably because of increased price con-
sciousness. Furthermore, differences in private label share
across product classes have led others to argue that private
labels do well in product classes in which consumers are
particularly price conscious (Raju, Sethuraman, and Dhar
1995). Finally, in a survey asking consumers why they buy
store brands rather than national brands, 67 percent rated
low price as "very important" (Kirk 1992).
Price may also be perceived at a broader level in that
consumers consider the ratio of quality received to price
paid in a purchase, and they are thus "value conscious" (cf.
Lichtenstein, Netemeyer, and Burton 1990; Thaler 1985;
Zeithaml 1988). Several articles in the popular press sug-
gest that a balancing of price and quality issues are related
to attitudes toward private label products (e.g., Deveny
1993; Liesse 1993). For example, in the survey cited
above, a slightly larger proportion of respondents (75%)
listed quality as a very important reason for buying store
brands than listed price (67%) (Kirk 1992). Moreover,
much of the recent increase in private label share has been
attributed to increases in their overall levels of quality (cf.
"Battle of the Brands" 1993; DeNitto 1993b). Hoch and
B anerji (1993) found that product quality had a significant
 Burtonet al. / PRIVATELABELPRODUCTS 295

FIGURE 1
Predicted Relationships Between Consumers' Attitude
Toward Private Label Products and Other Constructs

CONSUMER PRICE PERCEPTIONS

9 PRICE CONSCIOUSNESS(+)
9VALUE CONSCIOUSNESS(+)
9 p R I C E - Q U A L I T Y PERCEPTION(-)
1
MARKETING CONSTRUCTS
~ LOYALTY (-) PRIVATE PURCHASE
9RISK AVERSENESS (-) LABEL OF
9 IMPULSIVENESS (-) ATTITUDE PRIVATELABEL
9 SMART SHOPPER.SELF- (+) PRODUCTS
PERCEPTIONS (+)

D E A L PRONENESS CONSTRUCTS
9 G E N E R A L D E A L PRONENESS (+)
9 PRICE-RELATED D E A L S (+)
9 NON-PRICE DEALS (+)
9 I N T E R N A L REFERENCE PRICE
(IRP) RELIANCE (+)

NOTE:Whiletheimplieddirectionofcausalityinthe figureis leftto right,arrowsare notincludedto be consistentwiththenoncausalhypothesesproposed

in the text.

effect on private label market share for various product
categories, while the average level of price discount for
private brands did not. In sum, both value consciousness
and price consciousness are expected to be positively re-
lated to consumers' attitude toward private labels.
Another correlate of consumers' private label attitude
may be consumers' belief in price-quality relationships
across product classes. This belief, reflecting a "price-
quality schema" suggests that low prices are associated
with low quality, and thus lower priced products within a
product class are viewed less favorably (Lichtenstein and
Burton 1989; Peterson and Wilson 1985). Following this
perspective, Wolinsky (1987) argued that private label atti-
tudes are directly affected by the degree to which consum-
ers draw inferences from price and brand names to assess
quality. He viewed the existence of national and private la-
bel brand markets as evidence of segments that perceive
strong price and brand relationships with quality and those
who do not perceive such relationships. Wolinsky's con-
tention is consistent with results of cluster analyses that
suggest some groups of consumers are more likely to per-
ceive positive price-quality relationships across all prod-
uct classes, while another group is less likely to perceive a
relationship between price and quality for any product
category (Lichtenstein and Burton 1989). Wolinsky's ar-
gument and prior research suggest that a schema indicat-
ing a positive relationship between price and quality will
be negatively related with consumers' private label atti-
tude. On the basis of the above rationale, the following re-
lationships are proposed:
Hypothesis I: Attitude toward private label products is
related to consumer price perceptions. Specifically,
private label attitude is (a) positively related to price
consciousness, (b) positively related to value con-
sciousness, and (c) negatively related to price-
quality perceptions.
Relationships With Other
Marketing Constructs
The business press offers several indications of con-
sumer behavior constructs that are believed to be associ-
ated with attitudes toward private label brands. For
example, the recent gains made by private labels have of-
ten been linked to (lower) levels of brand loyalty. While
brand loyalty can be product category-specific, some con-
sumers have a propensity to be brand loyal across product
categories, indicating a general tendency or consumer
trait. Survey results show that in 1976, 76 percent of con-
296 JOURNALOF THE ACADEMYOF MARKETINGSCIENCE
sumers considered themselves to be "brand loyal" com-
pared with 23 percent in I993 ("Battle of the Brands"
1993). As loyalty to national brands decreases, consumers
may switch from one national brand to another or from
some specific national brand to a private label brand.
Therefore, this offers an increased opportunity for private
label brands to attract a share from national manufacturer
brands and suggests that attitude toward private label
brands will be negatively related to the general propensity
to be brand loyal across product categories.
One of the consumer benefits generally attributed to na-
tional manufacturer brands is that they reduce consumer
risk because, as a group, national brands are perceived to
have less variability in product quality than are private la-
bel brands (Kotler 199I; Montgomery and Wernerfelt
1992). The greater variability in quality across the private
label products increases the risk that a given private label
brand will not perform in a satisfactory manner. This sug-
gests that consumers who are more risk averse would be
less receptive to private label brands. Consistent with this,
research from the price-perceived quality research stream
provides evidence that risk-averse consumers are more
likely to purchase higher-priced brands within a product
category as a means of reducing the risk of purchasing a
brand of inferior quality (e.g., Peterson and Wilson 1985).
On the basis of this rationale, and consistent with the pre-
diction of a negative relationship between private label at-
titude and price-quality perceptions (Hypothesis lc), the
attitude toward private label grocery products should be
negatively related to risk averseness.
Transaction utility is defined as the pleasure (displeas-
ure) arising from a purchase transaction based on the con-
sumer paying a price lower (higher) than his or her internal
reference price (Thaler 1985). To the extent that internal
reference prices are related to prices of national brands, it
is proposed that transaction utility underlies, at least to
some degree, private label attitude. An orientation toward
transaction utility implies some level of the thoughtfulness
and effort associated with performing the "mental arith-
metic" required in evaluating the merits of a purchase
transaction. This thoughtfulness and effort suggest a nega-
tive relationship with impulsive decision-making by the
consumer. Comparisons between national and private la-
bel brands that result in a positive private label attitude
suggest thoughtful shoppers who take pride in their
decision-making ability. Consumers with such orienta-
tions may view themselves as sophisticated "smart" shop-
pers who are not easily influenced by national ad
campaigns, who are less impulsive in their decision-
making, and make rational choices among brand alterna-
fives without being influenced by national brand product
images ("Battle of the Brands" 1993; Blattberg and Neslin
1990). Conventional views also have described at least
one segment of buyers of private label brands as "upscale
smart shoppers" who actively make comparisons across
brands (Walker 1991). Consistent with the above ra-
tionale, consumers oriented toward private labels believe
that there are advantageous purchase opportunities avail-
able in the marketplace, act less impulsively in their
FALL 1998
decision-making, and take pride in capitalizing on these
opportunities.
Hypothesis 2: The attitude toward private label products
is (a) negatively related to brand loyalty, (b) nega-
tively related to risk averseness, (c) negatively re-
lated to impulsiveness, and (d) positively related to
smart-shopper self-perceptions.
Relationships With Deal Proneness
For some consumers who have a positive attitude to-
ward private label products, beyond seeking low prices per
se, utility may be derived by obtaining favorable deals
through sales promotions, and thus, these consumers can
also be described as being "deal prone" (cf. Lichtenstein
et al. 1990; Thaler 1985). The theoretical mechanism un-
derlying deal proneness is viewed as not the low deal price
per se but the transaction utility that is created when a con-
sumer pays a price below his or her internal reference price
(i.e., a mentally stored price against which other prices are
judged) (Rosch I975; Thaler I985). For example, the use
of a coupon (e.g., 50 cents off) for a product priced at $1.75
would likely be more attractive to the deal-prone consumer
than the purchase of the same product offered at an identi-
cal everyday price (e.g., $1.25). While both situations
yield the same net price of $1.25 (creating the same value
for the money or "acquisition utility") (cf. Thaler 1985), to
the degree the consumer accepts the everyday normal
price as a valid internal reference price, the form of the deal
in the first situation creates positive transaction utility.
One basis that many consumers may use for their inter-
nal reference prices for grocery products is the everyday
price of a leading national brand or some average price of
national brands. Retail merchants readily reinforce this
perception by frequently using semantics such as "Com-
pare at $XX, Our Price $YY" in advertisements and
shelf-talkers where the "XX" is the advertised reference
price corresponding to some national brand and "YY" is
the lower price of the store brand. Thus, by the same theo-
retical mechanism, to the extent consumers base their in-
ternal reference price on the price of national brands,
transaction utility would be created by the purchase of a
lower-priced private label brand. Consumers who are deal
prone also may be motivated to seek transaction utility and
may be similar in this regard to consumers with a favorable
attitude toward private label products. This common con-
tern with transaction utility suggests a positive relation-
ship between deal proneness and private label attitude.
Inherent in the notion that there are consumers who
seek transaction utility from purchases is the assumption
that these consumers have some level of awareness for
prices they pay (or expect to pay), which serve as a basis
for internal reference prices necessary for the mental cal-
culation of transaction utility. However, much research
evidence indicates that many consumers are not aware of
prices, even for products they have just purchased (cf.
Dickson and Sawyer 1990), suggesting corresponding
variance in the degree to which consumers have and rely

on mentally stored reference prices. Because a positive at-
titude toward private label brands is hypothesized to be
based, in part, on transaction utility, it follows that the atti-
tude toward private label products should also be posi-
tively related to the reliance on internal reference prices.
To this point, deal proneness has been discussed in a
broad-based, general manner. However, some of the recent
literature notes that consumer sales promotions differ in
their orientation and specific type of appeal to consumers
(Blattberg and Neslin 1990; Shimp 1990). One suggested
means of classifying consumer promotions or"deals" is by
whether the particular promotion results in a lower pur-
chase price (e.g., sales, coupons) versus those that offer no
direct price incentive (e.g., contests and sweepstakes, free
gifts). Given the above rationale concerning the common-
ality of the role of price in transaction utility for private la-
bel products and consumer price promotions, it may be
predicted that the private label attitude will be more
strongly related to price-based deal proneness than non-
price-based deal proneness. On the basis of the above dis-
cussion, we offer the following predictions.
Hypothesis 3: Private label attitude is (a) positively re-
lated to the general deal proneness level of the con-
sumer; (b) more strongly related to price-related
types of deal proneness (sale proneness, cents-off
proneness, coupon proneness, rebate proneness) than
non-price-related types of deal proneness (contest/
sweepstakes, display proneness, free gift, and buy-
1-get-1-free); and (c) positively related to consumer
reliance on internal reference prices.
Relationships With In-Store
Purchase Behaviors
Several relationships relating to actual consumer pur-
chase behavior obtained from a shopping trip were ad-
dressed to assess the validity of the private label attitudinal
measure and to profile consumers with positive attitudes
and behavior toward private label products. To support
predictive validity, the latent measure of private label atti-
tude should be positively related to the percentage of pri-
vate label purchases. Furthermore, as a more rigorous test
of the validity of the private label attitude scale, private la-
bel attitude should be the strongest predictor of the per-
centage of actual private label purchases, capable of
explaining this behavior after accounting for variance ex-
pinned by the predicted psychological correlates noted in
previous hypotheses. Hypothesis 4 follows.
Hypothesis 4: Private label attitude is positively related
to percentage of private label purchases made on a
shopping trip and is able to explain variance in pri-
vate label purchases beyond that explained by price
perceptions, deal perceptions, and other marketing-
related constructs.
While the relationship between private label proneness
and deal proneness was addressed in a prior hypothesis
Burtonet al. / PRIVATELABELPRODUCTS 297
(H3a), the relationships concerning the direct behavioral
manifestations of each of these pronenesses, private label
purchases, and deal purchases, respectively, are also of in-
terest. Consistent with the transaction utility rationale un-
derlying prior hypotheses concerning deals (i.e., H3a,
H3b), a consumer segment that purchases a high percent-
age of private label products may strongly value transac-
tion utility and thus be above average in the number of sale
products purchased and coupons redeemed (i.e., a positive
relationship between the purchase of private label prod-
ucts and purchases on deal).
Alternatively, the literature suggests that as consumers
purchase more items that are on sale or items using cou-
pons, they may purchase fewer private label brands. Al-
though transaction utility may underlie pronenesses
toward both price-oriented promotions and private label
products, when it comes to the direct behavioral manifes-
tations of these pronenesses, because consumers only
need and can use so much of a product, acting on both
types of pronenesses becomes less plausible. (That is, con-
sumers will rarely choose to buy both a large percentage of
national brands on promotion and a large percentage of
private label brands on any given shopping trip.) This ra-
tionale suggests that consumer segments that purchase a
high percentage of private label brands will not be above
average purchasers of sale items and items with coupons.
Thus, consumers acting on one type of proneness do so at
the expense of acting on the other type of proneness, and at
the behavioral level, there may be a negative relationship
between purchasing private label products and acting on
price-oriented promotions. Consistent with this rationale,
it has been noted that price-oriented sales promotions are
often used by national brand manufacturers to attract a
share from lower-priced private label brands (Stern 1993).
However, in research that has examined the effect of prod-
uct category promotion activity for national brands on ag-
gregated private label market share, mixed results have
been obtained (Hoch and Banerji 1993; Sethuraman and
Mittelstaedt 1992). Given the conflicting rationale and
mixed findings, no explicit hypotheses are offered for the
relationship between private label purchases and purchase
behaviors involving coupon redemptions and acting on ad-
vertised sales, but findings relevant to this important ques-
tion are addressed as part of our study results.
METHOD
Pretest
A pretest was conducted to develop the scale to meas-
ure consumer attitude toward private label brands and to
pretest measures of several of the postulated correlates for
this attitude measure. Items to assess constructs were ad-
ministered to a convenience sample consisting of 140 non-
student respondents who reported that they were the
primary shoppers for their households. The median age of
these respondents was 35 years, and the median level of
298 JOURNALOF THE ACADEMYOF MARKETINGSCIENCE
education was some college. Sixty-seven percent of the
pretest respondents were female.
As part of the scale development process, the following
definition of private label attitude was formulated--"a
predisposition to respond in a favorable or unfavorable
manner due to product evaluations, purchase evaluations,
and/or self-evaluations associated with private label gro-
cery products" On the basis of a review of the literature,
this definition appeared broad enough to encompass a sin-
gle dimension attitude that reflected a favorable or unfa-
vorable predisposition toward private labels based on both
economic (e.g., quality, price) and noneconomic (e.g., joy
associated with getting a "good deal") factors. An initial
pool of items consistent with this conceptual domain was
then generated by the researchers. Before responding to
these items, respondents read survey instructions that de-
fined what was meant by a private label or store brand gro-
cery product, and examples across several major grocery
stores in the area were provided for clarification. The in-
structions emphasized that the scale items referred to pri-
vate labels sold in grocery stores in general, and not to
brands sold in any specific store. Twelve items were pre-
tested to obtain preliminary estimates of dimensionality
and internal consistency. All items were 7-point Likert-
type scales, and four of the items were negatively worded.
The responses to the private label items were initially
subjected to a series of principal components analyses. In
the first analysis, no restrictions were placed on the
number of components to be extracted. On the basis of the
"eigenvalue greater than 1" heuristic, two principal com-
ponents were extracted. However, the first principal com-
ponent accounted for 62 percent of the total variance, and
there was a large break in the scree plot between the first
and second principal components (eigenvalues of 7.28 and
1.23, respectively), suggesting a strong single component.
Items with loadings less than .60 were next deleted (elimi-
nating three negatively worded items), and a second prin-
cipal component analysis was performed on the remaining
items. In this second analysis, only one principal compo-
nent with an eigenvalue greater than 1 was extracted. This
result was consistent with a single component measure,
and this single component explained 67 percent of the total
variance. To further trim the number of items to produce a
more parsimonious scale that would be easier to adminis-
ter, we deleted items with loadings of.70 or less (one item)
or .85 or more (two items deleted for reasons of "empirical
redundancy" [cf. Bagozzi and Yi 1988]). We performed
another analysis on the remaining six items. Again, only
one principal component was extracted, and it accounted
for 65 percent of the variance in the data. These final six
items comprising the private label scale are presented in
the top of the appendix, and the six items that were deleted
based on pretest analyses are shown in the bottom of the
appendix.
The six items in the private label attitude scale were
next subjected to confirmatory factor analyses using LIS-
REL VIII (Jrreskog and Srrbom 1993). A one-factor
model was estimated to confirm the structure indicated in
the principal component analyses. Fit statistics from the
FALL1998
pretest were supportive of internal consistency and dimen-
sionality of the items comprising the scale. 3Coefficient al-
pha was .89. The pretest also assessed items that were
generated by the authors for constructs hypothesized to be
related to private label attitude (e.g., risk averseness,
smart-shopper self-perception) but for which no estab-
lished measures were available.
Procedures and Measures
Used in the Main Study
Consumers were initially recruited for the study at two
different grocery stores belonging to the same chain. Inter-
viewers were trained on how to initially approach shop-
pers just before they exited the grocery stores. These
trained interviewers were positioned behind a table near
the exits of the two supermarkets over a 5-day period
(Wednesday-Sunday) from 10:00 a.m. to 8:00 p.m. The
two grocery stores were located in the midwestern region
of the country in a small city of about 90,000 people. Con-
sumers who agreed to participate in a university-
sponsored project were asked to relinquish their cash reg-
ister receipts. To assess various in-store behaviors, infor-
mation was coded directly from the retained grocery
receipts. The number of total products purchased, and the
number of private label, national brand, and generic prod-
ucts purchased were coded. These data were used to calcu-
late the percentage of private label and national brand
purchases. Other purchase information coded included (1)
the number and dollar amounts saved from purchasing sale
items advertised (all national brands) in the weekly flyer
and (2) the quantity and face value of coupons redeemed.
Data collection was scheduled so that it coincided with a
single sale period for the store. During this time period,
there were no advertised sale offers or coupons available
for private label brands.
Shoppers agreeing to participate in the study were
asked several questions in the store (e.g., "About how long
did you spend in the grocery store today?") and then were
given a take-home survey to be answered at a later date and
a postage-paid return envelope. The survey contained the
private label attitude measures, measures of constructs hy-
pothesized to be related to private label attitude, and demo-
graphic measures. Of 896 surveys distributed to those
contacted in the stores, 333 (37%) complete, usable sur-
veys were returned. Two thirds (67%) of the respondents
were married, and more than three fourths (78%) were fe-
male. The median annual household income was between
$35,000 and $49,999, and the median age category was
35-44.
The number of scale items, coefficient alpha, and sam-
ple items are provided in Table 1 for the multi-item meas-
ures of the constructs predicted to be related to private
label attitude. The price-related construct measures (i.e.,
price consciousness, value consciousness, and price-
quality perception) were taken from Lichtenstein, Ridg-
way, and Netemeyer (1993), and the general and deal-
specific deal proneness measures were taken from
Lichtenstein, Netemeyer, and Burton (1995). Impulsive-
 Burton et al. / PRIVATELABEL PRODUCTS 299

Multi-Item Measure
Value consciousness
Price consciousness
Price-quality perception
Brand loyalty
Impulsiveness
Risk averseness
Smart-shopper self-perception
General deal proneness
Reliance on internal reference price
Deal-specific proneness
Coupon proneness
Sale proneness
Cents-off proneness
Rebate proneness
Free gift proneness
One-free proneness
Contest proneness
Display proneness
TABLE 1
Measure Descriptions and Reliabilities
Number of ltems Coefficient ot Sample Itema
7 .86 "I always check prices at the grocery store to be sure I get the best value for
the money I spend"
5 .86 "The time it takes to find low prices is usually not worth the effort" (reverse
coded).
4 .85 "Generally speaking, the higher the price of a product, the higher the quality"
5 .92 "Once I get used to a brand, I hate to switch?'
9 .83 "Generally speaking, I would consider myself an impulsive shopper."
4 .78 "I don't like to take risks?'
4 .94 "Making smart purchases makes me feel good about myself?'
8 .90 "Compared to other people, I am very likely to purchase brands that come
with promotional offers?'
4 .85 "In deciding if the price of a grocery product is a 'good' price, I think about
what a fair price for the brand would be?'
5 .86 "I enjoy using coupons, regardless of the amount I save by doing so?'
6 .86 "I am more likely to buy brands that are on sale?'
7 .90 "Compared to most people, I would say I have a positive attitude toward
cents-off deals?'
6 .83 "Receiving cash rebates makes me feel good?'
6 .91 "Beyond the money I save, buying a brand that comes with a free gift gives
me a sense of joy?'
6 .84 "When I take advantage of a 'buy-one-get-one-free' offer, I feel good?'
6 .91 "Manufacturers' contests and sweepstakes are fun to enter, even if I know
I'll never win?'
7 .90 "End-of-aisle displays have influenced me to buy brands I normally would
not buy?'

a. All items for each of the multi-item measures are available from the first author on request.

ness was assessed via items from an expanded version of
the scale of Martin, Weun, and Beatty (1994), and included
four additional items that were assessed in the pretest. On
the basis o f the work o f Jacoby and Chestnut (1978) and
Raju (1980), brand loyalty was measured with five items
that were previously used by Lichtenstein et al. (1990). For
smart-shopper self-perception, risk averseness, and reli-
ance on internal reference price, items consistent with the
conceptual domains were generated by the authors and as-
sessed in the pretest. On the basis of reliability estimates in
the pretest, less-reliable items were deleted prior to use of
these construct measures in the main study.
RESULTS
Measurement Properties of
the Private Label Attitude Scale
In the main study, the six-item measure of private label
attitude developed in the pretest was subjected to a confir-
matory analysis using L I S R E L VIII to assess psychomet-
ric properties. The Z 2 associated with a one-factor model
was 65.4 ( d f = 9, p < .01), and the Goodness o f Fit Index
(GFI) and Adjusted Goodness of Fit Index (AGFI) were
.94 and .86, respectively. Because these values are influ-
enced by sampling considerations and deviations from
normality, the Comparative Fit Index (CFI) and Tucker-
Lewis Index (TLI) are often recommended because they
are considered robust to these variations (cf. Bollen 1989).
Advocated fit levels for these indexes are in the .90 region,
and for these data, both indexes exceeded this criterion
(CFI = .94, TLI = .91). Estimates of internal consistency
also were supportive o f the scale. Item loadings were all
highly significant (t-values associated with the loadings
ranging from 10.8 to 18.2), coefficient alpha was .873, and
the variance extracted estimate was .56. Variance ex-
tracted estimates of .50 and above indicate convergent va-
lidity among items in a given scale (Bagozzi and Yi 1988;
Fornell and Larcker 1981), and the alpha meets levels rec-
ommended in the literature (Nunnally 1978). The mean
for the private label scale was 25.7 (an average item score
of 4.3); the median equaled 26.0; the standard deviation
was 7.5; scores ranged between 6 and 42; and quartile
splits occurred at 21, 26, and 31.
Given the common focus on price between private label
attitude, coupon proneness, and value consciousness, the
common theoretical base of transaction utility postulated,
and predicted positive correlations, confirmatory models
were performed to test discriminant validity. These mod-
els included (1) a hypothesized three-factor model in
which the private label, coupon proneness, and value con-
sciousness constructs are modeled as separate, but corre-
lated, constructs; (2) a one-factor model in which all three
constructs' items were specified as loading on a single fac-
tor; and (3) three different two-factor models specifying
300 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 1998

TABLE 2
Bivariate Relationships Between Private Label Attitude and Price Perception,
General Consumer Marketplace, and Deal Proneness-Related Constructs
1 2 3 4 5 6 7 8 9 10 11 12

1. Private label attitude

2. Value consciousness .29** --
3. Price consciousness .26** .46** --
4. Price-quality perceptions --.25** -.23** -.37** --
5. Brand loyalty --.27** -.25** -.39** .28** --
6. Risk averseness .00 .00 .01 -.04 -.14' --
7. Impulsiveness -.20** -.26** -.14" .23** .09 .11
8. Smart shopper .27** .28** .32** -.09 -.08 -.03 .12"
9. General deal proneness .18"* .29** .25** .02 -.I0 .02 .13' .59* --
10. Price-related deals .30** .35** .40** -.14"* -.16 -.07 .01 .72** .61'*
11. Nonprice deals .15" .11 .19'* .06 .00 -.03 .27** .62** .51"* .60** B

12. IRP a reliance .14" .43** .44** -.16" -.22** -.02 -.14" -.25** .13' .17" .10

NOTE: Correlations shown in bold and italics provide tests of relationships with private label attitude proposed in Hypotheses 1, 2, and 3.
a. IRP = internalreferenceprice.
*p < .05. **p < .01.

the items of two scales into a single factor while allowing
the items of the third construct to load on the second factor.
(For example, combining the private label and value items
into one factor, and allowing the coupon proneness (CP)
items to load on a CP factor constitutes a two-factor
model). From these models, three tests of discriminant va-
lidity recommended in the literature were conducted (see
Anderson and Gerbing 1988; Fornell and Larcker 1981).
First, the difference in the chi-square value for the hy-
pothesized three-factor model was compared to the one-
factor and two-factor models. The chi-square for the hy-
pothesized three-factor model was 324.5 (df= 132; GFI
and AGFI = .90 and .87, respectively). The chi-square for
the one-factor model was 1,758.0 (df= 135; GFI and AGFI
= .53 and .49, respectively), and the chi-squares for the
three two-factor models were 920.2, 1,041.8, and 1,105.0
(df= 134; GFIs of .65, .67, and .68 and AGFIs of .55, .58,
and .59, respectively). The chi-square difference between
the one-factor and postulated three-factor model is 1,433.5
(df= 3, p < .01), and the differences for the three- and two-
factor models range between 595.7 and 780.5 (df= 2, p <
.01). All these comparisons show the hypothesized three-
factor model to be better fitted to the data than the one- or
two-factor models. Also, the correlations between the
three constructs (completely standardized phi estimates)
for the three-factor model ranged from. 19 to .32, and all
phi estimates were significantly less than 1.
A more rigorous test of discriminant validity assesses
whether the square of the phi estimate between two con-
structs is less than the average variance extracted estimates
for the two constructs. The average variance extracted was
.56, .47, and .56 for private label, coupon proneness, and
value consciousness measures, respectively. The square of
the phi estimates ranged from .04 to. 10, and all were less
than the average variance extracted for the corresponding
construct pairs. In sum, across all three tests, discriminant
validity was supported.
Relationships With Price-Related and
Consumer Marketplace Constructs
As shown in Figure 1, Hypothesis 1 predicted that pri-
vate label attitude is positively related to value conscious-
ness and price consciousness and negatively related to
price-quality perceptions. Hypothesis 2 pertained to the
relationships between private label attitude and the general
marketplace constructs of brand loyalty, risk averseness,
impulsiveness, and smart-shopper self-perceptions. Hy-
pothesis 3 concerned relationships with deal proneness-
related constructs. To test all of these predictions, bivariate
correlations were examined, and results relevant to Hy-
pothesis 1 to Hypothesis 3 are shown in the first column
of correlations in Table 2. The bivariate correlations be-
tween private label attitude and the price perception vari-
ables (see correlations in the first column for measures 2
through 4) are all in the predicted direction and significant
(p < .01). These findings support the predicted relation-
ships between private label attitude and consumer price
perceptions.
Results pertaining to the relationship between private
label attitude and general marketplace constructs postu-
lated in Hypothesis 2 are shown for Measures 5 through 8
in the first column of Table 2. Three of the four correla-
tions (i.e., brand loyalty, impulsiveness, and smart-
shopper self-perceptions) are in the postulated direction
and significant (p < .01 for each). Risk averseness is not re-
lated to private label attitude.4 Taken in sum, these results
offer general support for Hypothesis 2, and specific sup-
port for Hypotheses 2a, 2c, and 2d.
Relationships With Deal Proneness
It was predicted that private label attitude is positively
related to consumer deal proneness in general (Hypothesis
3a) and that the relationship is stronger for price than for

non-price-oriented deals (Hypothesis 3b). As shown in
Table 2, the correlation between private label attitude and
the general measure of deal proneness is .18 (p < .01),
thereby providing support for Hypothesis 3a. The correla-
tions of private label attitude with the price-oriented deals
of sale proneness, rebate proneness, cents-off proneness,
and coupon proneness are .30, .23, .21, and. 18 (p < .01 for
all), respectively. The correlations with the non-price-
oriented deals (buy-l-get-l-free, free gift with purchase,
display proneness, and contest/sweepstakes proneness)
are. 19 (p < .01),. 15 (p < .05), .09 (ns), and .05 (ns), respec-
tively. To test Hypothesis 3b formally, the specific prone-
ness measures were combined to form more general price
and nonprice deal proneness measures (coefficient alphas
for these price and nonprice measures were .89 and .91, re-
spectively), and the difference in correlations for these
price and nonprice deal measures with the private label at-
titude scale was assessed. As shown in Table 2, the overall
correlation between private label attitude and the measure
associated with proneness toward price-oriented deals is
.30, and the correlation between private label attitude and
consumer proneness toward nonprice deals is.15. The dif-
ference between these two dependent correlations is sig-
nificant (t = 2.63, p < .01), thus offering support for
Hypothesis 3b (Cohen and Cohen 1983:56-57).
It was posited that transaction utility underlies private
label attitude (similar to the manner it has been proposed to
underlie deal proneness; cf. Lichtenstein et al. 1990).
Since the mental calculation of transaction utility relies on
some conception of an internal reference price, Hypothe-
sis 3c predicts that private label attitude is positively re-
lated to this reliance. The Pearson correlation, shown in
the bottom row of Table 2, is. 14 (p < .05), thus supporting
Hypothesis 3c. (The correlation between general deal
proneness and this reliance was. 13 [p < .05], suggesting
that transaction utility might underlie both constructs to a
similar degree.)
Relationship Between Private Label Attitude
and Private Label Brand Purchases
A simple test of predictive validity suggests that private
label attitude should be positively associated with the per-
centage of private label brand purchases and negatively as-
sociated with the percentage of national brand purchases
made. The Pearson correlation between private label atti-
tude and the percentages of private label purchases made
on the consumers' shopping trips is .25 (p < .01). 5
A more rigorous test of validity was proposed in
Hypothesis 4, where it was postulated that private label at-
titude explains variance in the percentage of private label
purchases made on the shopping occasion, after account-
ing for variance explained by the price perception con-
structs, general deal proneness construct, and the other
general marketplace measures addressed in Hypothesis 2.
To test Hypothesis 4, a hierarchical stepwise regression
was performed in which percentage of private label pur-
chases was the dependent variable. In the first stage of
Burton et al. / PRIVATE LABEL PRODUCTS 301
TABLE 3
Hierarchical Regression Results for Prediction
of the Percentage of Private Label Purchases
Independent Standardized
Variable Coefficient t Value Model F Adjusted R2
Stage 1 7.26*** .06
Impulsiveness -.22 -3.02***
Price-quality
perceptions -.11 -1.50"
Stage 2 9.90*** .12
Private label attitude .27 3.76***
Impulsiveness -.19 -2.63***
Price-quality
perceptions -.05 -0.70
*p < .10. ***p< .01.
analysis, nine psychological variables were eligible for en-
try, and significant predictors were entered in a stepwise
manner. Following the entry of all significant predictors,
private label attitude became eligible as a predictor in the
second stage of analysis. Results are shown in Table 3.
As shown in the upper portion of Table 3, in Stage 1 im-
pulsiveness and price-quality perceptions entered as sig-
nificant predictors. Upon entry of these two predictors, no
other variables entered the equation. These two variables
explained 6 percent of the variance in the percentage of
private label purchases. In Stage 2 of the analysis, private
label attitude became eligible for entry into the equation.
As shown in the bottom of Table 3, it entered as a signifi-
cant predictor (F change = 14.1; p < .01), increasing the
model's adjusted R 2to .12. Given a lack of common meth-
ods' variance for the independent and dependent variables,
this effect is viewed as nontrivial (cf. Peterson, Albaum,
and Beltramini 1985) and offers support for Hypothesis 4. 6
Private Label Purchases and
Deal-Responsive Behaviors
To test the relationships between private label brand
purchases and purchases of advertised sale offers and cou-
pons redeemed on the shopping trip, consumer segments
were formed by dividing respondents into terciles on the
basis of private label purchase percentage scores (cf.
Richins and Dawson 1992). Also, in an attempt to gain in-
sight into differences in the relationships between (1) pri-
vate label purchase behavior and promotion-responsive
behavior, and (2) private label attitude and promotion-
responsive behavior, similar segments were formed on the
basis of respondents' private label attitude scores. Means for
coupon and sale promotion response variables for the high
and low segments for both purchase behavior and attitude
are shown in Table 4. To obtain more stable estimates of
the percentage of private label products purchased, this
analysis was limited to consumers who purchased 20 or
more items on their shopping trip. To control for the effect
of total quantity of items purchased on the number of cou-
302 JOURNALOF THE ACADEMYOF MARKETINGSCIENCE FALL1998

TABLE 4
Purchase Behavior Results for Private Label Purchase and Attitude Segments
Percentageof PrivateLabelPurchases PrivateLabelAttitude
Low-Percentage High-Percentage F Low-Attitude High-Attitude F
Dependent Variable Segment Segment Value Segment Segment Value
Numberof couponsredeemed 4.0 1.4 10.2"** 3.2 2.5 0.5
Amountof couponsredeemed $2.27 $0.73 10.6"** $1.87 $1.35 0.8
Numberof sale itemspurchased 3.0 2.0 3.6* 2.3 3.2 3.4*
Amountsavedon saleitems $2.33 $1.45 4.2"* $1.77 $2.50 3.1"

% of privatelabelpurchases 15.I 24.8 21.6"**

% of nationalbrandpurchases 82.5 71.7 21.3.**

Wilks'slambda 0.92 3.5*** 0.84 4.6***

*p < .I0. **p < .05. ***p< .01.

pons redeemed and sale items purchased, the total number
of items purchased was used as a covariate, and one-way
multivariate analyses of covariance (MANCOVAs) with
follow-up univariate analyses were performed to assess
between group differences. Results are shown in Table 4.
Results for those purchasing a low percentage versus a
high percentage of private label brands for their store visit are
shown in the left-hand portion of Table 4. Overall, results
are significant (Wilks's lambda = 0.92; F = 3.5, p < .01).
Despite the positive correlations between deal proneness
and private label attitude (demonstrated by results for Hy-
pothesis 3), when examining the most direct manifestation
of private label attitude (i.e., actual private label pur-
chases), the mean scores in Table 4 show that those in the
high-percentage private label purchase segment redeemed
fewer coupons and redeemed coupons for a lesser total
value for this shopping trip than did the low-percentage
private label purchase group. Similarly, this high-
percentage private label purchase group bought fewer sale
items advertised in the store flyer and "saved" a lesser
amount of money from sale items purchased. At the be-
havioral level, this suggests a negative association be-
tween private label purchases and purchases using
coupons or products on sale.
The right-hand portion of Table 4 shows results pertain-
ing to the six-item private label attitude measure. The mul-
tivariate effect is significant (Wilks's lambda =0.84; F = 4.6,
p < .001). While the coupon-related findings were highly
significant across the segments formed on the basis of the
private label purchase measure, results pertaining to cou-
pon redemptions and total coupon redemption dollar
amounts are not significant for segments formed on the ba-
sis of the private label attitude measure. Also, despite the
negative relationship between percentage of private label
purchases and sale purchases shown in the left-hand col-
umn of Table 4, results for sale items and amount saved are
marginally significant in the opposite direction for the atti-
tude scale (i.e., more sale items are purchased by consum-
ers scoring high on the private label attitude scale). Also,
predictive validity of the attitude scale measure is again
strongly demonstrated by the percentages of private label
brands purchased by consumer segments higher (24.8%)
and lower (I5.1%) on the attitude scale. The overall pat-
tern of results shown across Table 4 is consistent with the
belief that although private label attitude and deal prone-
ness are positively correlated constructs, in dealing with
behavioral manifestations of these constructs on any sin-
gle shopping trip, there is some trade-off between pur-
chases of sale items and private label items.
Private Label Attitude and
Consumer Demographics
A question of interest to marketing practitioners is
whether there are differences in private label attitudes
across demographic variables. Demographic profiles for
both the measure of private label attitude and percentage of
private label purchases are shown in Table 5. Analyses of
variance results indicate significant differences between
levels of education and family income and private label at-
titude. Differences for gender and age are nonsignificant.
Interestingly, despite a positive association between edu-
cation and household income (p < .05), for these data, re-
spondents with more education have higher mean scores
on the attitude measure, while those with greater family in-
come (i.e., more than $50,000) have somewhat lower
scores. The directional patterns of relationships for these
demographic variables are reflected in the percentages of
private label purchases (e.g., higher education and lower
incomes associated with higher purchase percentages), but
the large variance in the percentage measure resulted in
nonsignificant relationships.
DISCUSSION
Despite tremendous interest in the current and future
role of private label brands by retailers, distributors, and
national brand manufacturers, there has been little schol-
arly research that has examined the private label attitude of
consumers. This research has sought to address this void
by developing a multi-item measure of private label atti-

TABLE 5
Demographic Differences for Private Label Attitude and Percentage of Private Label Purchases
Demographic Variable Means for Private Label Attitude
Gender
Male 26.3
Female 25.6
Age
< 35 25.8
35-44 25.7
45-54 25.0
> 55 26.6
Education
Some college or less 24.6
College graduate 25.4
Postgraduate degree 27.3
Family income
< $25,000 27.7
$25,000-49,999 26.4
$50,000-74,999 24.1
> $75,000 25.0
*p < .05.
tude, assessing its psychometric properties, and examin-
ing predictions about variables related to this construct.
One potential benefit of the private label attitude measure
is that it will allow product managers and retailers to track
changes in general attitudes over time (in much the same
way as constructs such as "consumer confidence" are
tracked), rather than rely only on private label sales infor-
mation, which is confounded with other variables. The po-
tential usefulness of this measure is reinforced by
predictions about changes in consumer acceptance of pri-
vate labels over the next decade, coupled with the interest
in private label opportunities by major retail chains and
product manufacturers and marketers (DeNitto 1993b).
By selecting a segment of consumers scoring high on this
measure, marketers can easily identify consumers who
constitute a primary target market for private label grocery
brands. 7In this study, the segment high on the private label
scale bought 50 percent more private label products than
did the segment low on the scale measure. This target
group is obviously of great interest to retailers marketing
private label brands, as well as to national brand manufac-
turers concerned about competition provided by private la-
bel products.
Findings from this study also provide empirical support
for several relationships suggested in the literature regard-
ing private label brands. Private label attitude was posi-
tively related to value consciousness and deal proneness,
and negatively related to brand loyalty and price-quality
perceptions. In addition, positive relationships were found
between private label attitude and reliance on internal ref-
erence prices and smart-shopper self-perceptions, and a
negative relationship was found between private label atti-
tude and impulsiveness. Taken as a group, the level of sup-
port across predictions supports the nomological validity
of the measure. The empirical evidence also supports the
Burton et al. / PRIVATE LABEL PRODUCTS 303
F Value Percentage of Private Label Purchases F Value
0.62 21.5 1.86
18.5
17.8
0.53 18.7 0.60
18.8
21.2
17,0
3.24* 20,0 1.19
19.8
19.6
3.12* 20.8 0.70
18.3
17.9
discriminant, convergent, and predictive validity of the
private label attitude scale.
Private Label Attitude and Deal Proneness
Because national manufacturers are attempting to re-
gain a share from private labels through the use of promo-
tional deals, the relationship between private label attitude
and deal proneness is of both practical and theoretical in-
terest. As hypothesized, private label attitude was posi-
tively related to deal proneness, and it was more strongly
related to a proneness to deals involving price reductions
than to other deal types. Of eight deal types, the correlation
was highest for sale proneness, a promotion type offering a
direct price reduction with little or no effort expended out-
side the store. This lower price with little out-of-store ef-
fort needed is similar to benefits offered by private labels.
Given conflicting rationale and mixed previous empiri-
cal findings for the relationship between national brand
promotion activity and private label share, we were inter-
ested in examining the level of coupon usage and sale pur-
chases made by consumer segments that purchased a high
versus a low percentage of private label brands. Despite
the positive correlation between the proneness scale meas-
ures for coupons, sales, and private label brands (i.e., sup-
port for Hypothesis 3b), the purchase of private label
products was negatively associated with the number of
sale items purchased and coupons redeemed. The consum-
ers who actually purchased a lower percentage of private
label brands bought more sale items and saved a greater
amount on the national brand products that were on sale
and redeemed more coupons for a larger amount. This pat-
tern of results is consistent with the notion that consumer
segments may similarly value the transaction utility de-
rived from promotion deals and private label brands (thus
304 JOURNALOF THE ACADEMYOF MARKETINGSCIENCE
leading to a positive correlation between deal proneness
and private label attitude as predicted in Hypothesis 3a).
However, at a behavioral level, this desire for transaction
utility underlying both of these constructs generally leads
to above-average purchases of either private label brands
or deal-based purchases, but not both on any single pur-
chase occasion.
In addition, results showed a reversal of relationships
between (1) private label attitude and sale purchases (posi-
tive) and (2) private label purchases and sale purchases
(negative). This finding again suggests that a positive atti-
tude toward private label products also relates to purchase
of goods on sale, but to the detriment of actual private label
brand purchases. This suggests that periodic sales may be
particularly effective in competing against private label
brands. However, such a strategy will be unlikely to meet
with long-term success unless consumers perceive the pro-
moted national brand to be of sufficiently higher quality to
offset its (nonpromotional) higher price once the promo-
tion is removed. To the extent that manufacturers success-
fully accomplish this, internal reference prices for national
brands may be significantly higher than for private brands,
thereby allowing for higher selling prices. However, to the
extent that promotions are run too frequently and/or na-
tional brand quality is not perceived as superior to private
brands, this sales promotion strategy could result in a
long-term decrease, rather than an increase, in the differ-
ence between internal reference prices for national and pri-
vate brands. In sum, promotional sales by national
manufacturers could enhance utility perceptions that may
attract a share from private label brands in the short run,
but frequent promotional activity may have potential detri-
mental effects for national brands over the longer term
(Liesse 1993).
An additional finding with potentially important impli-
cations is the negative relationship found between impul-
siveness and private label attitude/purchases. It appears
that consumers who have a positive attitude toward private
label products view themselves as "smart" shoppers who
are more purposeful in their behavior and are willing to
seek out private label versions of a product rather than pur-
chasing on impulse. These findings suggest that such con-
sumers are more likely to seek out private label products
even when they have less-visible locations in the store. Na-
tional brands, therefore, may benefit somewhat more from
desirable locations (e.g., end-of-aisle displays) than do
private label brands. Because national brand manufactur-
ers are often willing to compensate stores via promotional
allowances for desirable locations, retailers may be able to
benefit from these allowances while knowing that many
consumers with a positive private label attitude will seek
out private label products even if in somewhat less-
enticing store locations.
Limitations and Implications
for Future Research
While research projects that make any attempt to com-
bine measures of latent, unobservable constructs with ac-
FALL1998
tual purchase behavior are rare and advantageous because
of a lack of common methods' variance, a limitation of our
study is the use of a single shopping occasion for the be-
havioral measures. Measurement on a single occasion can
negatively affect the reliability of the measured variable
when the variable is taken as an operationalization of more
general behavior (Epstein 1979). This lower reliability and
the effect of situational variables for any single occasion
can attenuate the strength of relationships between latent
constructs and behavior. Consequently, the significant re-
lationships found in the present study may actually under-
represent the true ecological relationships. Future research
that combines latent construct measures with purchase be-
havior data for a broader time frame (e.g., diaries or scan-
ner panels) appears warranted. Such longitudinal data
would foster a more complete understanding of relation-
ships between private label attitude and private label pur-
chases. Longitudinal data also could provide greater
insight into the effects of frequent national brand promo-
tions on national brand and private label purchases, a sub-
ject of great importance to grocery product marketers.
Along these same lines, longitudinal data would also
permit an assessment of the moderating role of a broader
range of environmental and situational factors on private
label attitude/purchase relationships. Such variables
clearly are important determinants of purchase behavior
on any single shopping occasion, and comparisons across
trips would be useful in enhancing our understanding of
this broader model. In addition, to extend the generaliz-
ability of findings, research should be conducted in other
stores using larger and more geographically diverse samples.
The findings of this study suggest several other avenues
for future research. One opportunity is extending the use
of the private label attitude measure to the international do-
main to examine relationships between consumers' per-
ceptions and purchase behaviors in markets in which
private label brands capture greater (or lesser) overall mar-
ket shares. It would be interesting to see to what degree the
magnitude of relationships between private label attitude
and price and deal-related constructs examined here will
generalize to these (other) international markets. Also, an
examination could be made between the attitudinal meas-
ure used here and other measures of consumers' responses
to store brands that have been offered recently. Specifi-
cally, Richardson, Jain, and Dick (1996) constructed a
store brand proneness index by asking consumers to esti-
mate the frequency of purchase of store brands (e.g.,
rarely, sometimes, always) for 28 product categories.
Given the significant relationship between the attitude
measure and actual in-store purchase behavior in this
study, one would anticipate a strong positive relationship
between our attitude measure and this self-report purchase
index. Also, while our focus in this study was on grocery
products, future research may address consumers' private
label attitudes using this scale in studies examining dura-
ble products. Results of this study may be used as a spring-
board for future research of interest to both marketing
academics and industry researchers who are interested in
private label brands.

APPENDIX
Private label attitude scale items
1. Buying private label brands makes me feel good.
2. I love it when private label brands are available for the
product categories I purchase.
3. For most product categories, the best buy is usually the pri-
vate label brand.
4. In general, private label brands are poor-quality products.
(reverse coded)
5. Considering value for the money, I prefer private label
brands to national brands.
6. When I buy a private label brand, I always feel that I am
getting a good deal.
Items that were deleted on the basis of pretest results
The lower price on private label brands is due to the poor quality
of the product.
The quality of advertised national brands is always better than
that of private label brands.
Private label brands are less expensive than national brands be-
cause they offer such poor quality.
I think private label brands offer great value for the money.
I feel like a smart shopper when I buy private label brands.
When I buy private label brands, I feel like I am getting the most
for my money.
NOTES
1. In our validation, we provide support for the predictive validity of
the measure by demonstrating its positive relationship to across-category
purchase behavior. Also, following previous conceptualizations, various
product-category-related and situational variables are not viewed as af-
fecting an individual consumer's attitude toward private label products.
Such variables, including the availability of private label brands within
specific product classes orthe purchase for self or others, are viewed as
moderators of a consumer's propensity to act on his or her relatively en-
during attitude toward private label products. This treatment is consistent
with the attitude-behavior literature, which shows that attitude-behavior
relationships may be moderated by many situational variables that affect
the strength of the attitude-behavior relationship but do not influence the
level of the attitude itself.
2. A reviewer noted that it would be interesting to test relationships
between all of these constructs in Figure I by using structural equation
modeling. While such results would be of interest, it is difficult to justify
conceptually any specific "causal" ordering for the predicted correlates
of attitude toward private label brands. Given the scale development ob-
jectives of the article and cross-sectional data, relationships indicated in
Figure I are assessed in the Results section using bivariate and multiple
correlation techniques.
3. More specifically, the g2 associated with a one-factor model for the
pretest was 22.1 (dr= 9,p < .01), and the Goodness of Fit Index (GFI) and
Adjusted Goodness of Fit Index (AGFI) were .95 and .88, respectively.
The comparative fit and Tucker-Lewis indexes, which are generally con-
sidered to be robust to sampling variations (cf. Bollen 1989),were .97 and
.95, respectively.
4. As noted by a reviewer, the low correlations between risk averse-
ness and other constructs may be due to its more general operationaliza-
tion (e.g., "I don't like to take risks") that is not closely tied to grocery
shopping attitudes or behavior.
5. The Pearson correlation between private label attitude and the per-
centage of national brands purchased is-.27 (p < .01), while the correla-
tion between private label attitude and the percentage of generic brands
Burton et al. / PRIVATELABEL PRODUCTS 305
purchased is +.13 (p < .05). We also performed a tercile split on the pri-
vate label attitude measure and examined the number of private label
product purchases across the groups. The highest segment on the attitude
scale bought 50 percent more private label products than did the lowest
segment (F = 10.7,p < .01).
6. At the suggestion of one of the reviewers, we also performed an
even more rigorous test in which all of the independent variables were
forced into the equation at the first stage, and then the private label meas-
ure became eligible for entry in the second stage. Private label attitude
was again a significant predictor (t = 3.59, p < .01), the F change value
was significant (p < .01), and the adjusted R" (which takes into account
the larger number of predictors in the model) increased from .04 to .10
when private label attitude was added to the model.
Also, as noted by another reviewer, labelingthe relationship between
private label attitude and the percentage of private label purchases as a
test of"predictive validity" appears unusualbecause operationally the at-
titude measure was obtained after the shopping episode in which the be-
havior occurred. Here, we have assumed that the private label attitude is
relatively enduring (e.g., Ajzen and Fishbein 1980), that it was present
before the shopping episode, and given our results, the attitude appears
useful in explaining private label purchase behavior.
7. While behavioral data gleaned from scanner data are also useful for
targeting efforts, the magnitude and difficulty of the processing task may
make the use of a simple self-report measure more efficient for many
firms.
REFERENCES
Ajzen, leek and Martin Fishbein. 1980. Understanding Attitudes and
Predicting Social Behavior. Hillsdale, NJ: Lawrence Erlbaum.
Anderson, James C. and David W. Gerbing. 1988. "Structural Equation
Modeling in Practice: A Review and Recommended Two-Step Ap-
proach." PsychologicalBulletin 103 (November): 411-423.
Bagozzi, Richard and Youjae Yi. 1988. "On the Evaluation of Structural
Equation Models" Journal of the Academy of Marketing Science 16
(Spring): 74-94.
"Battle of the Brands,' 1993. Consumer Reports 58 (September): 565-
566.
Blattberg, Robert C. and Scott A. Neslin. 1990. Sales Promotion: Con-
cepts, Methods, and Strategies. 1st ed. Englewood Cliffs, NJ: Pren-
tice Hall.
Bollen, Kenneth A. 1989. Structural Equations With Latent Variables.
New York: John Wiley.
Boone, Louis E. and David L. Kurtz. 1995. Contemporary Marketing.
Fort Worth, TX: Dryden.
Cohen, Jacob and Patricia Cohen. 1983. Applied Multiple Regression/
Correlation Analysis for the Behavioral Sciences. Hillsdale, NJ:
Lawrence Erlbaum.
Denitto, Emily. 1993a. "No End to March of Private Label?' AdAge, No-
vember 1, pp. 5-6.
: 1993b. 'q'bey Aren't PrivateLabels Anymore--They'reBrands"
AdAge, September 13, p. 8.
Deveny, Kathleen. 1993. "Bargain Hunters Bag More Store Brands?'
Wall Street Journal, April 15, pp. B1, B10.
Dickson, Peter R. and Alan G. Sawyer. 1990. "The Price Knowledge and
Search of Supermarket Shoppers?' Journal of Marketing 54 (July):
42-53.
Epstein, Seymour. 1979. "The Stability of Behavior: I. On Predicting
Most of the People Much of the Time." Journal ofPersonalityand So-
cial Psychology 37 (July): 1097-1126.
Foraeli, Claes and David E Larcker. 1981. "Evaluating Structural Equa-
tion Models With Unobservable Variables and Measurement Error"
Journal of Marketing Research 18 (February): 39-50.
Heller, Al. 1997. "Moving to Sawier Marketing?' Supermarket Business
52 (May): 17-18.
Hoch, Stephen J. and Shumeet Banerji. 1993. "When Do Private Labels
SucceedT' Sloan Management Review 34 (Summer): 57-67.
Jacoby, Jacob and Robert W. Chestnut. 1978. Brand Loyalty: Measure.
ment and Management. New York: John Wiley.
JiSreskog, Karl G. and Dag Sfrbom. 1993. LISREL 8: User's Reference
Guide. Chicago: Scientific Software International.
 306 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE
Kahn, Barbara E., and Leigh McAlister. 1997. GroceryRevolution: The
New Focus on the Consumer.Reading, MA: Addison-Wesley.
Kirk, Jim. 1992. "The New Status Symbols"Adweek, October 5, pp. 38-39.
Kotler, Philip. 1991. Marketing Management. Englewood Cliffs, NJ:
Prentice Hall.
--and Gary Armstrong. 1996. Principlesof Marketing. Englewood
Cliffs9 NJ: Prentice Hall.
Lichtenstein, Donald R. and Scot Burton. 1989. "The Relationship Be-
tween Perceived and Objective Price-Quality?' Journal of Marketing
Research 26 (November): 429-443.
9Richard G. Netemeyer, and Scot Burton. 1990. "Distinguishing
Coupon Proneness From Value Consciousness: An Acquisition-
Transaction Utility Theory Perspective?' Journal of Marketing 54
(July): 54-67.
--, and .1995. "Assessing the Domain Specificity
of Deal Proneness: A Field Study." Journal of ConsumerResearch22
(December): 314-326.
-, Nancy M. Ridgway, and Richard G. Netemeyer. 1993. "Price
Perceptions and Consumer Shopping Behavior: A Field Study?' Jour-
nal of Marketing Research 30 (May): 234-245.
Liebeck, Laura. 1996. "Private Label Goes Premium" Discount Store
News 3 (November): F38.
Liesse, Julie. 1993. "Big Name Marketers Are Being Stalked by Strong,
High Quality Store Brands?' Advertising Age, April 12, pp. 1,4.
Martin, Wendy, Seungoog Weun, and Sharon Beatty. 1994. "Validation
of an Impulse Buying Tendency Scale,' Working Paper. University of
Alabama, Tuscaloosa.
Montgomery, C. A. and B. Wernerfelt. 1992. "Risk Reduction and Um-
brella Branding,' Journal of Business 65 (January): 31-50.
Nunnally, Jum C. 1978. PsychometricTheory.New York: McGraw-Hill.
Ortega, Bob and Gabriella Stem. 1993. "Retailers' Private Labels Strain
Old Ties?' WallStreet Journal, September 9, pp. B 1, B8.
Peterson, Robert A. and William R. Wilson. 1985. "Perceived Risk and
Price Reliance Schema as Price-Perceived Quality Mediators?' In
Perceived Quality: How Consumers View Stores and Merchandise.
Eds. Jacob Jacoby and Jerry C. Olson. Lexington, MA: D.C. Heath,
247-268.
, Gerald Albanm, and Richard E Beltramini. 1985. "A Meta-
Analysis of Effects Sizes in Consumer Behavior Experiments?' Jour-
nal of ConsumerResearch 12 (June): 97-103.
Raju, P. S. 1980. "Optimum Stimulation Level: Its Relationship to Per-
sonality, Demographics, and Exploratory Behavior." Journal of Con-
sumer Research 7 (December): 97-103.
Raju, J. S., R. Sethuraman, and S. Dhar. 1995. "The Introduction and Per-
formance of Store Brands?' ManagementScience41 (June): 957-973.
Richardson, Paul S., Arun K. Jaln, and Alan Dick. 1996. "Household
Store Brand Proneness: A Framework." Journal of Retailing 72 (2):
159-185.
Richins, Marsha and Scott Dawson. 1992. "A Consumer Values Orienta-
tion for Materialism and Its Measurement: Scale Development and
Validation." Journal of Consumer Research 19 (December): 303-
316.
Rosch, Elenor. 1975. "Cognitive Reference Points,' Cognitive Psychol-
ogy 7 (October): 532-547.
Seligman, Phillip M. 1995. "Food Processing,' ValueLine, August 18,
p. 1461.
Sethuraman, Raj and John Mittelstaedt. 1992. "Coupons and Private La-
bels: A Cross-Category Analysis of Grocery Products?' Psychology
& Marketing (November-December): 487-500.
Shimp, Terence A. 1990. PromotionManagementand Marketing Com-
munications. 2nd ed. Hinsdale, IL: Dryden.
Stem, Gabdella. 1993. "Brand Names Are Getting Steamed Up to Peel
Off Their Private Label Rivals." Wall Street Journal, April 21,
pp. B l-B2.
View publication stats
FALL 1998
Thaler, Richard. 1985. "Mental Accounting and Consumer Choice,'
Marketing Science 4 (Summer): 199-214.
Walker, Chip. 1991. "What's in a Name?' American Demographics 13
(February): 54.
Wellman, David. 1997. "Souping Up Private Label?' SupermarketBusi-
ness 52 (October): 13-20.
Wolinsky, Asher. 1987. "Brand Names and Price Discrimination" Jour-
nal of lndustrial Economics 35 (March): 255-268.
Zeithaml, Valarie A. 1988. "Consumer Perceptions of Price, Quality, and
Value: A Means-End Model and Synthesis of Evidence" Journal of
Marketing 52 (July): 2-22.
ABOUT THE AUTHORS
Scot Burton is professor and Wal-Mart chairholder in the De-
partment of Marketing and Transportation at the University of
Arkansas. His research interests include public policy and con-
sumer welfare concerns, consumer price and promotion percep-
tions, and survey research measurement issues. His work has
been published in the Journal of Marketing, the Journal of Mar-
keting Research, the Journal of Consumer Research, the Journal
of the Academy of Marketing Science, Public Opinion Quarterly,
the Journal of Applied Psychology, and others.
Donald R. Lichtenstein is a professor in the Department of Mar-
keting at the University of Colorado. His research interests in-
clude consumer processing of price and sales promotion
information and consumer welfare issues related to marketplace
choice. His work has been published in the Journal of Marketing,
the Journal of Marketing Research, the Journal of Consumer Re-
search, and others.
Richard G. Netemeyer is a professor in the Department of Mar-
keting, E. J. Ourso College of Business Administration at Louisi-
ana State University (LSU). He received his Ph.D. in business
administration from the University of South Carolina in 1986.
Since then, he has been a member of the marketing faculty at
LSU. His research interests include measurement and scaling,
public policy, maladaptive behaviors, and consumer behavior in
general. His research has been published in the Journal of Con-
sumer Research, the Journal of Marketing Research, the Journal
of Marketing, the Journal of Applied Psychology, the Journal of
Public Policy & Marketing, and others. He is also a member of
the editorial review boards of the Journal of Consumer Research
and the Journal of Public Policy & Marketing.
Judith A. Garretson is a Ph.D. candidate in the Department of
Marketing and Transportation at the University of Arkansas. Her
research interests include promotion issues and public policy and
consumer welfare. Her work has appeared in the Journal of Pub-
lic Policy & Marketing, the Journal of Professional Services
Marketing, and proceedings of the American Marketing Asso-
ciation and Association for Consumer Research.