ORIGINAL ARTICLE The impact of age and shopping experiences on the classification of search, experience, and credence goods in online shopping Yun Wan • Makoto Nakayama • Norma Sutcliffe Received: 21 December 2009 / Revised: 19 April 2010 / Accepted: 14 May 2010 / Published online: 23 November 2010 Ó Springer-Verlag 2010 Abstract This study explores how age and consumers’ Web shopping experience influence the search, experience, and credence (SEC) ratings of products and ser- vices in online shopping. Using the survey data collected from 549 consumers, we investigated how they perceived the uncertainty of product quality on six search, experience, and credence goods. The ANOVA results show that age and the Web consumers’ shopping experience are significant factors. A generation gap is iden- tified for all but one experience good. Web shopping experience is not a significant factor for search goods but is for experience and credence goods. There is an interaction effect between age and Web shopping experience for one credence good. Implications of these results are discussed. Keywords Online shopping Á Generation gap Á Quality perceptions Á Search goods Á Experience goods Á Credence goods 1 Introduction Since the popularity of World Wide Web was facilitated by the first multimedia Web browser, Mosaic, in 1994, online shopping has transformed from a trendy emerging shopping channel into the shopping mainstream. Most consumers have overcome their initial concerns about risk in online transactions. The general public have developed trust with major shopping portals though there are still issues to be investigated (Holsapple and Sasidharan 2005; Yan et al. 2008). In addition, mobile commerce Y. Wan (&) University of Houston, Victoria, 14000 University Blvd, Sugar Land, TX 77479, USA e-mail: [email protected]M. Nakayama Á N. Sutcliffe DePaul University, 243 South Wabash Avenue, Chicago, IL 60604, USA e-mail: [email protected]123 Inf Syst E-Bus Manage (2012) 10:135–148 DOI 10.1007/s10257-010-0156-y
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ORI GIN AL ARTICLE
The impact of age and shopping experienceson the classification of search, experience,and credence goods in online shopping
Yun Wan • Makoto Nakayama • Norma Sutcliffe
Received: 21 December 2009 / Revised: 19 April 2010 / Accepted: 14 May 2010 /
Published online: 23 November 2010
� Springer-Verlag 2010
Abstract This study explores how age and consumers’ Web shopping experience
influence the search, experience, and credence (SEC) ratings of products and ser-
vices in online shopping. Using the survey data collected from 549 consumers, we
investigated how they perceived the uncertainty of product quality on six search,
experience, and credence goods. The ANOVA results show that age and the Web
consumers’ shopping experience are significant factors. A generation gap is iden-
tified for all but one experience good. Web shopping experience is not a significant
factor for search goods but is for experience and credence goods. There is an
interaction effect between age and Web shopping experience for one credence good.
(Gebauer et al. 2008) and collaborative commerce (Hartono and Holsapple 2004) are
building upon the existing B2C infrastructure and adding new dynamics to online
shopping. However, among all these B2C opportunities and challenges, there is a
fundamental issue that has not been thoroughly addressed: the impact of age and
shopping experience on consumers’ perception of online products and services.
Conventional wisdom is that teens and young adults have advantages in online
shopping because they are quick to learn and adapt to the new online shopping
environment. The older generation probably does not shop online as much because
they are less familiar with and slower to adapt to the new environment. Thus, the older
generation shies away from online shopping more than the younger generation.
But, several recent surveys indicate that this may not be the case. For example,
according to a survey by Pew Research (Jones 2009), although older generations use
the Internet less for socializing and entertainment, they do use it more as a tool for
searching for information, emailing, and buying products. In addition, now both young
and old equally pursue video downloads, online travel reservations, and work-related
research. Another survey conducted by the University of Southern California found
that older Americans have equal or even more enthusiasm to Web 2.0 than their
younger, more tech-savvy counterparts (USC 2008). The same survey indicates that
while instant messaging and video downloading still remain more popular with the
younger generation, older Americans check the Internet more frequently for news. The
older generations are logging onto online communities, researching purchases,
becoming socially active and playing games in increasing numbers (USC 2008). A
survey by a UK-based media company in December 2008 found that there were no
significant differences between younger and older generations in terms of their general
shopping behavior and concerns about online fraud (NWA 2009).
The older generation might be slow in learning new technologies and find
themselves in a disadvantageous position in terms of keeping up with the trends on
the Web when being compared to their younger counterparts, but older generation
do have more shopping experience, even though most of such experiences are
rooted in a traditional environment. Such experiences may give them an edge in
evaluating and purchasing certain types of products or services on the Web.
Thus the reality of online shopping by different age groups may be more complex
than a simple dyad of young and fast versus old and slow. It is possible that both
groups have their advantages and disadvantages when shopping online, thus,
different needs and expectations of online vendors. Their behavior in online
shopping and reactions to shopping technologies might also be different because of
their accumulated shopping experience on Main Street as well as online. The same
survey by Pew Research found that, in terms of preference for online shopping,
instead of a downward linear trend with age, interest in online shopping is
significantly lower among both the youngest and oldest groups—‘‘38% of online
teens buy products online, as do 56% of Internet users ages 64–72 and 47% of
Internet users age 73 and older’’—and significantly higher among those in the
medium range, with 80% for age 33–44 and 71% for age 18–32 (Jones 2009).
Therefore, the conventional wisdom may underestimate the online shopping
tendency and ability of older generation online shoppers. They may not only be
more active in online shopping than expected but also have more experience
136 Y. Wan
123
because of their accumulated shopping experienced in a traditional environment. If
we stick with the conventional wisdom, it may lead to bad strategies for online
vendors and ignore potential business opportunities that exist in older generation
online shoppers. Thus it is critical for us to explore this research question in a more
rigorous way.
In this study, we examined how the age and shopping experience influence
consumer’s classification of search, experience, and credence goods. Through the
lens of SEC, we found that age, Web shopping and search knowledge, as well as
prior purchase experiences all have a significant influence on a consumer’s quality
evaluation as reflected in an SEC rating, which leads to different SEC ratings for the
same product.
The remainder of this paper is arranged as follows. First, we review previous
studies of factors that have impact on online shopping as well as the SEC
framework. Then we explain our survey-based empirical experiment design as well
as the outcomes. Finally, we analyze the results and conclude the study.
2 Previous studies
2.1 Search, experience, and credence goods perception
Consumers have been used to conducting shopping in an environment where they
can inspect the goods directly and converse with the sellers or service providers in a
face-to-face environment to assess the quality of the goods to be purchased. We can
classify consumer goods and service into three categories—search goods, experi-
ence goods, and credence goods—based on the point in time consumers evaluate the
quality of the goods they have purchased, hence the SEC framework (Nelson 1970;
Darby and Kami 1973; Nelson 1974). Specifically, search goods are those that
consumers can confidently evaluate the quality of before the purchase. Experience
goods are those that consumers can evaluate the quality of once they are consumed
or serviced. Credence goods are those that consumers cannot evaluate the quality of
even a long time after the purchase.
Because of the differences in times of which consumers can evaluate a product
confidently, they have to use different product evaluation strategies when making
shopping decisions. In a traditional brick-and-mortar market, consumers may use a
direct inspection method to evaluate search goods and a sampling strategy for
experience goods. However, for credence goods, they largely depend on the brand
name and recommendations to make decisions because they have difficulty
evaluating the quality directly. Different shopping and product evaluation behaviors
by consumers lead to different advertising and promotion strategies by vendors for
different SEC categories. The SEC framework has been widely adopted in the
advertising industry and used in consumer behavior research (Ekelund et al. 1995).
The SEC framework provides a relatively objective classification schema for
commodities. The fact that it is being widely adopted in the advertising industry
indicates the classification method applies to most populations. However, we also
realized that for a specific product or service, depending on the consumers’
The impact of age and shopping experiences 137
123
familiarity with it, one consumer may rate it in a different SEC category from
another consumer. For example, a Dell laptop is a search product for computer
geeks, but it could be an experience product for computer rookies.
Such differences in individual perception originate from two sources.
Firstly, any product or service has multiple attribute dimensions and these
attributes can range from search and experience to credence categories—that is,
some attributes’ quality can be evaluated prior to purchase, others after purchase,
while still others cannot be evaluated even after long-term use. In other words, all
goods can have search, experience, and credence attributes simultaneously. We
classify a product or service into a certain SEC category because the attributes in
that category are those we are most concerned with. Though the generally accepted
SEC category for a particular product indicates the most important attributes for the
majority of consumers, it is possible that for a specific consumer, such attributes are
less important than a different set of attributes in other SEC categories. So the SEC
rating for this product is different for this consumer compared with others.
Secondly, consumers may have the capability to infer the attributes information
in one category, say credence attributes, by evaluating related attributes in other
categories, say search attributes. Since such capability varies among consumer
groups, their perception of SEC rating for a product or service also varies.
For example, a computer geek may rate a Dell laptop as a search product because
based on his prior experience of using laptops, he can infer the quality of experience
and credence attributes of the laptop by observing its search attributes. He can check
the smoothness of program launch, listen to the noise of the hard disk and CPU fan,
observe the screen display quality, and touch the keyboard to infer those experience
or credence attributes like the overall integration soundness and efficiency of heat
dissipation. The latter may greatly influence the life span of a laptop, which is
largely a credence attribute. Since a rookie user of laptops does not have such
experience, he or she may either regard it as an experience product or even a
credence product.
Thus, though the SEC rating for the same product and service could be the same
in the traditional environment for the majority of consumers, it could be different
when the evaluation environment becomes the Web, as well as when the product or
service being evaluated by consumers with different sets of characteristics.
2.2 Goods evaluation in Web shopping environment
With the growing popularity of World Wide Web since 1994, online shopping has
become part of our daily life. In this new online environment, no goods can be
inspected directly, and only limited interactions with service providers are possible
(Alba et al. 1997). Thus, by default it seems, goods in online shopping automatically
become either experience or credence goods. However, because online retail site
designers have used various methods to help consumers evaluate physical products
(Smith et al. 2005), goods in online shopping could also be search goods—as our
experiments indicated.
Regarding how to evaluate goods in an online environment, the behaviors and
strategies employed by shoppers are influenced by several key factors. Bhatnagar
138 Y. Wan
123
et al. (2000) found that age, years of using the Web, and gender affect purchase risk
perceptions differently, thus leading to different behaviors and strategies. A more
recent study that examines the impacts of demographic factors on online shopping
found that age, gender, income, and location influence online purchase frequency
and expenditures (Chang and Samuel 2004).
Age is essentially a rough indicator of Main Street shopping experiences
accumulated through time. The younger generation has limited shopping experience
but could cleverly use those shopping tools to leverage their existing shopping
experience. For older generation, we speculate that, while the adaptation process
may take longer, their rich Main Street shopping experiences give them advantages.
For example, on the correlation between the online search and the actual purchase of
products, it was found that although young people are likelier to purchase online the
longer they searched for the product, older generations are comparatively more
likely to purchase because they spend less time searching (Sorce et al. 2005).
Like age, gender is another important factor in explaining many differences in
consumers’ shopping behaviors and perception of goods. However, it seems gender
is not as significant factor as age in predicting online shopping behavior. A recent
study for online shopping behaviors from international and cross-cultural perspec-
tives found gender has no significant influence on shopping behavior (Stafford et al.
2004). But it found that the 25–34 age group was the most active online shopping
group. Another e-commerce study found gender and social class were not significant
factors for mobile commerce adoption though it found that younger consumers are
more predisposed to use mobile equipment as a shopping channel (Bigne et al.
2005).
Web shopping experience, including using various Web-based decision support
tools for searching, comparing, and analyzing products and services in the online
environment, has positive influence on the perception and evaluation of goods on
the Web. Dennis et al. (2002) found that younger people are ‘‘more Web-literate
than older age groups’’ and those young consumers with more Web shopping
experience have a more positive attitude towards Web shopping than those without
it (Dillon and Reif 2004). It was also found that Web shopping experience has a
positive influence on m-commerce adoption (Bigne et al. 2005).
In summary, in the online environment, we identify the following factors that
may influence an online shopper’s SEC classifications. They are age, gender, Webshopping experiences, Web search experience, and prior purchase experience forthe same goods.
3 Hypothesis
As mentioned previously, goods are classified as search, experience, and credence
goods depending on when a consumer can confidently evaluate their quality.
However, such classification is mostly dependent on an individual’s previous
purchase and usage experience, especially for experience and credence goods. A
consumer from the older generation, because of greater shopping experience, may
be less likely to categorize a specific product or good as an experience or credence
The impact of age and shopping experiences 139
123
good than a younger consumer would. We expect such differences to also exist in
the online shopping environment. Thus, we have our first hypothesis:
H1 Online shoppers assess the SEC ratings of the same goods differently
depending on their age group, with the older generation rating the same goods as
less credence than the younger generation.
Based on existing research of the gender impact on goods perception in online
shopping, we have the following hypothesis:
H2 Gender has no significant influence on online shoppers’ SEC classification of
the same goods.
Since online shopping is a relatively new shopping mode and is still less than
15% of the US retail market, we expect that the extent to which an online shopper
evaluates goods online is highly influenced by that shopper’s Web purchasing
experience and online search skills. A Web-savvy shopper may rate credence goods
more like experience goods and experience goods more like search goods in the
online environment. Thus, we have hypothesis 3:
H3 Online Shoppers assess the SEC classification of the same goods differently
based on their level of Web shopping experience, and those shoppers who have
more Web shopping experience tend to rate the same goods as less credence than
those with less Web shopping experience.
Since evaluating a product or service in an online environment also depends on
searching and processing information from the Web, an online shopper’s search
engine experience may influence his or her SEC rating for the product or service. If
an online shopper is familiar with using a search engine to find product or service
quality information, he or she may tend to evaluate such a product or service more
as a search good compared with those consumers who are less familiar with using a
search engine, with all other conditions being the same. Thus, we have hypothesis 4:
H4 Online shoppers assess the SEC classification of the same goods differently
based on their level of online search engine experience, and those shoppers who
have more search engine experience tend to rate the same goods as less credence
than those with less search engine experience.
In addition to general shopping experience accumulated with age, when someone
has prior purchase experience of a particular product or service, he or she may feel
less uncertain about this product or service in later purchases, and such confidence
may be accumulated or enhanced with each additional purchase of the same product
or service. Thus, consumers may change their perception of this product or service
from credence to experience or from experience to search category. Such experience
accumulation can come from either online or Main Street shopping. Thus we have
hypothesis 5:
H5 Online shoppers assess the SEC classification of the same goods differently
based on whether or not they have prior purchasing experience of the goods, and
140 Y. Wan
123
those shoppers who have prior purchase experience of the goods tend to rate the
same goods as less credence than those who do not have such experience.
Now we have explained all our hypotheses. In the next section, we explain the
design of our experiment to verify these hypotheses.
4 Research design
Since the most direct method of obtaining a product or service’s SEC rating is
asking the shopper to rate it (Iacobucci 1992; Ekelund et al. 1995; Girard et al.
2002), we use similar survey-based questionnaires by asking subjects for their SEC
ratings of a pre-selected set of goods.
The selection of the candidate goods was a critical step in this research. To avoid
reinventing the wheel, we conducted a comprehensive review of previous literature
to identify candidate goods that had been repeatedly identified in the same SEC
category by at least two prior studies, though all the contexts of such ratings were in
the traditional shopping environment. As a result, we got a relatively long list of
more than ten products and services. Then we reduced the list by using only those
common goods whose purchases are relatively neutral to age, gender, income and
ethnic groups. This process led to six goods as representative SEC goods, two in
each SEC category:
Search goods include PCs and bestselling books (Ekelund et al. 1995; Girard
et al. 2002, 2003; Hoskins et al. 2004).
Experience goods include cell phones and cars (Nelson 1970; Iacobucci 1992;
Girard et al. 2002).
Credence goods include vitamins and auto insurance (Girard et al. 2002; von
Ungern-Sternberg 2004; Chiu et al. 2005).
We created three scenarios to examine the effect of shopping contexts. In the first
scenario, shoppers can shop only online for the above six items (‘‘Web Only’’). In
the second scenario, they can only evaluate the six items in a traditional shopping
environment and cannot use the Web at all (‘‘No Web’’). In the third scenario,
consumers can shop for these six items by using any means—whether using the
Web or not (‘‘No Restriction’’).
In each scenario, there are two sections. The first section collected subjects’ age,
gender, Web shopping experience, and Web information search experience.
In the second section, subjects were asked to identify the SEC category for the six
items we selected. We used the same survey instrument as Iacobucci (1992) and
asked respondents to rate items in their respective SEC category by using a 7-point
Likert scale on a single item construct. That is, we asked the respondent to evaluate
if the quality of an item ‘‘could be assessed prior to purchase’’ (search), ‘‘could be
evaluated only after purchase’’ (experience), or ‘‘would be difficult evaluate even
after trial’’ (credence). And similar ratings were conducted in all three scenarios.
For each of the six items, we asked if the subject had previously purchased it
from the Web or Main Street, the frequency of purchases, and the ratio of purchases
frequency between online and Main Street.
The impact of age and shopping experiences 141
123
5 Pilot study and data collection
To examine the effectiveness of the experiment, we conducted a pilot study with
students from two Midwest and Southwest universities. Based on the feedback of
the students, we made improvements on the readability of several statements. The
overall structure of the experiment proved effective.
We then recruited subjects from the public by using online forums and sites like
Craigslist. A modest $5 Amazon.com gift certificate was used as an incentive for
participation. We removed those responses that were incomplete or invalid (e.g.,
entering the first choice for all the questions).
Altogether this study received 549 valid completed responses, 52.4% were male,
and 47.6% were female. This indicates a largely balanced sample of the population.
6 Data analysis and findings
We first calculated the mean SEC ratings for each item and the summary is listed in
Table 1. The overall mean was calculated based on all cases. The subsequent three sets
of means were calculated based on three scenarios as indicated in the research design.
As indicated in the pattern, the overall order of SEC ratings of these six items is
in accordance with the ratings collected from the previous literatures in the
traditional shopping environment. That is, PC and Book, as search goods, have
lower SEC ratings than experience goods, such as cell phones and cars. Cell phones
and cars have lower SEC ratings than vitamins, the credence goods. The only
exception is auto insurance. Although auto insurance was considered a credence
good and was much more difficult to evaluate than search and experience goods in
previous literature, in this study, its SEC rating was between cell phones and cars,
the two experience goods (Fig. 1).
In addition, through t-test, we found that there were no statistically significant
differences between the mean ratings of the same item among three scenarios. This
indicated that those selected items are neutral to shopping environmental factors.
Further analysis indicates the skewness and kurtosis statistics have a range of
0.185–0.55 and -0.212 to -0.722 respectively. This indicates a cluster to the low