Nghiên cứu khoa học - Sưu tầm

Longitudinal comparison of
Finnish and US online shopping

behaviour among university
students: The five-stage buying
decision process
Received (in revised form): 10th August, 2006
Charles Comegys
is the Ciejek Chair of Business and is Associate Professor of Marketing in the Girard School of Business & International
Commerce at Merrimack College, USA. Dr Comegys’ research interests include research methodologies and internet purchase
behaviour. His articles appear in numerous publications including the Journal of Marketing Theory and Practice, Practical
Assessment, Research & Evaluation, Journal of Internet Commerce, Journal of Business Education, Journal of Small Business
Management, Journal of the Academy of Business Education, Management Research, Atlanta Economic Review, Operations
Management Review, and Marketing News. He has served on the Board of Directors of the American Marketing Association,
the Marketing Research Association, and currently is on the Board of the Marketing Educators’ Association.
Mika Hannula
holds MSc (Eng) and DrTech in Industrial Management and Engineering. Professor Hannula is currently the Head of the
Institute of Business Information Management and the Director of the Degree Program of Information and Knowledge
Management at Tampere University of Technology, Finland.
Jaani Väisänen
received his Master’s Degree in Statistics from the University of Tampere, Finland. He currently works as a researcher in the
Institute of Business Information Management at Tampere University of Technology, where he is also completing his doctoral
studies. His current research interests are related to electronic commerce and especially in the use of search engines in the
marketing field.
Abstract This study investigates the online purchase behaviour of a key segment of
the population, the ‘Net Generation’ university-aged student, from two of the world’s
most advanced IT nations with the greatest potential in e-commerce: Finland and the
USA. Information about online shopping behaviour in 2002 is compared with 2004/2005
for the two countries. The research also answers the question whether online shoppers
from the two countries approach the consumer buying decision process differently over
time. The results provide useful guidance to consumer e-marketing companies.
INTRODUCTION devices almost throughout their entire

Charles Comegys, PhD
Girard School of Business The use of the internet has rapidly lives, which makes them highly literate
& International Commerce grown from its early years, thanks to the in online activities. This paper is focused
Merrimack College
315 Turnpike Street so-called ‘Net Generation’. The Net on the online shopping behaviour of a
North Andover, MA, USA Generation is said to consist of young key segment of two of the world’s most
Tel: +1 (978) 837 5409 people and adolescents born between advanced IT nations: Finnish and US
Fax: +1 (978) 837 5086
Email:
1977 and 1997.1 These people have been university students.2 Although there has
ccomegys@merrimack.edu using the latest information technology been ample discussion and criticism
336 Journal of Targeting, Measurement and Analysis for Marketing Vol. 14, 4, 336–356 䉷 Palgrave Macmillan Ltd 1479-1862/06 $30.00
Finnish and US online shopping behaviour among university students
concerning the use of university students that even with all these new innovations
as test subjects, they fit well for this and abilities, the internet has not
study as university students in the target changed the basic buying patterns of
countries are typically extensive users of consumers.6
the internet. This study is not intended These mixed opinions are to the
to gather information pertaining to the benefit of the present study, which aims to
proportion of the population that use the combine the aspects of flexibility with
internet, but to understand the buyer fixed purchasing patterns. Thus, the study
decision process of those that do. Thus, assumes that buying behaviour follows the
university students, although not a truly five-stage model with some modifications
representative sample, do reflect the for the online situation, and considers the
behaviour of heavy online users. changes that have occurred in two years
Information about this segment has been within the framework of the model.
gathered at two points in time: in 2002 Assuming that the model is a ‘closed
and 2004 in Finland, and in 2002 and space,’ which completely covers every
2005 in the USA. The focal point of this aspect of purchase, it is possible to draw
study is to compare the differences and conclusions on which parts of the model
relations between the two samples, and are more relevant to customers than
the two nations. Among different models others.
of consumer behaviour, perhaps the most
well known is the five-stage model,3
which describes the consumer purchase BACKGROUND
event as a process which starts long The five-stage buying decision process
before the actual purchase is model is a widely used tool for
consummated and continues after the marketers to gain a better understanding
purchase. This model is the backbone of about their customers and their
the present research, which depicts the behaviour.7 The idea of the model is that
changes in the process and the factors when a customer purchases an item, the
that underlie them between the samples. purchase event is a forward-moving
The results give marketers an insight into process, which begins long before the
how the Finnish and US online actual purchase and continues even after
marketplace has changed in the past two the purchase is made. As the name
years in the customer’s eyes, along with implies, there are five different stages in
some implications of where it might be the process, which are need recognition,
in another two years. information search, evaluation of
Electronic commerce offers many alternatives, purchase decision, and
advantages over traditional commerce postpurchase behaviour (see Figure 1).
that are useful to both marketers and Following is a review of each of these
customers. For example, customer stages in the consumer buying process.
relationship management (CRM) helps
the company to gather information and
maintain relations to the customer, who Need recognition
in turn will get precisely the information The buying process begins with need
they need.4 This flexibility is said to be recognition (sometimes referred to as
one of the internet’s greatest values for problem recognition), where the buyer
marketers when compared with senses a difference between their actual
conventional methods of commerce.5 On state and a state they desire. This need
the other hand, there are still opinions can be triggered by either an internal
䉷 Palgrave Macmillan Ltd 1479-1862/06 $30.00 Vol. 14, 4, 336–356 Journal of Targeting, Measurement and Analysis for Marketing 337
Comegys, Hannula and Väisänen
Need Information Evaluation Purchase Postpurchase
Recognition Search Decision Behavior
Figure 1: Five stage buying decision process model
(hunger, thirst) or external (passing have suggested that higher purchase

through McDonald’s stimulates hunger) frequencies stimulate repeat purchasing
stimuli.8 rather than variety-seeking behaviour,
Other factors besides differences in the which gives e-marketers a good reason
buyer’s actual and desired state influence to make the threshold for shopping as
need recognition as well. One of the low as possible.
directly observable influences is
demographic factors, including age, sex,
income, race, education, household size, Information search
and marital status.9 There are also The next stage in the buying process is
inferred influences which affect need information search, where the consumer
recognition. Psychological factors play a uses different channels to gather
major role in these inferred influences. information about available products,
Of these psychological factors, which might fulfil the needs discussed
motivation is the basis of all consumer above. Kotler15 defines two levels of
behaviour.10 Although there is no general arousal during the information search
agreement among psychologists on the phase. In the milder state, heightened
best way to classify consumer needs as attention, the consumer just becomes
far as motivation goes, Kinnear and more familiar with the different products
Bernhardt11 divide the motivational needs that might be suitable for them. In this
into physiological and psychological state, the consumer only pays attention
needs. The former include, for example, to advertisements and conversations
the need for food and shelter, while the about the subject. In the next state,
latter are generated by one’s social active information search, the consumer
environment. actively engages these conversations and
Perception reflects how the customers searches for information about different
see themselves and their surroundings, brands, models, etc.
which in turn affects not only the need Kotler16 also defines four methods by
recognition phase, but the other phases which consumers receive product/service
as well.12 Depending on the consumer’s information. Personal sources include
needs and perceptions, need recognition family and friends. Commercial sources
may take different forms. In addition to include advertising and salespersons.
purchasing something totally new or Public sources include mass media and
replacing a broken/outdated item, consumer-rating organisations.
consumers may replace a product that Experimental sources include examining
fills their needs completely with another. and using the product itself. Most of the
This phenomenon is called variety information comes from the commercial
seeking.13 Van Trijp, Hoyer and Inman14 sources, but according to Dubois,17 the
most effective information comes from increase the amount of searching, as the
personal sources. comprehension of the results would be
There are also differences in no problem. Knowledge uncertainty on
information search with different levels of the other hand seemed to lower the
expertise. Alba and Hutchinson18 report amount of search. Insights from Urbany
that experts are able to search more et al. disagree somewhat with Alba and
efficiently and therefore have more Hutchinson. The conclusions derived
knowledge of available products. They from both of these studies is that a
point out a common case, however, certain amount of knowledge about the
where experts do less actual searching desired product is necessary to increase
because they have already gathered active information searching, even
previous information about the products though potentially improved knowledge
in their memory (internal search). When may reduce it.
Beatty and Smith19 propose that product Of all the possible products and brands
class knowledge is negatively associated that might satisfy a customer’s needs,
with the search effort, it can be only a handful will be brought to the
concluded that often those with consumer’s attention. Kotler22 defines
intermediate knowledge of the desired four sets of alternatives from which the
product do most of the searching, final purchase decision will be made.
although its efficiency might not rival the The total set includes all the possible
experts’. Considering the online brands available to the customer. Derived
environment, expertise about the product from the total set, is an awareness set,
and expertise about the search process which includes the brands the customer
must be separated. As previously stated, has come to know. Brands that meet the
those with excellent knowledge about buyer’s purchasing criteria are taken from
the product might reduce the amount of the awareness set to form the
information search, and better knowledge consideration set. As more information
about the online search methods can about the products of the consideration
actually do the same, as the search set is gathered and evaluated, the final
procedure becomes more effective. In purchase decision will be made from the
addition, better knowledge about the choice set.
desired product can improve the quality Peter and Olson23 define a similar
of the consideration set (discussed later in process, which divides all the brands in
this paper), thus also reducing the search the product class into familiar and
time and improving the search unknown brands. From unknown brands,
efficiency.20 those that are found accidentally and
Another factor affecting of the amount those found through intentional search
and type of information search is the make it to the choice set. From familiar
perceived risk (uncertainty) involved brands, only those that are activated from
with the purchase. In their study, memory (evoked set) make it to the
Urbany, Dickson and Wilkie21 divide choice set. Among the familiar brands,
uncertainty into two categories. there may be some that once have been
Knowledge uncertainty includes familiar to the customer, but are lost in
uncertainty regarding information about memory. Of the evoked set, brands that
alternatives, and choice uncertainty are liked more and thought more typical
means uncertainty about which to the product class tend to emerge into
alternative to choose. Urbany et al. noted the choice set.24
that choice uncertainty appeared to Narayana and Markin25 have extended
the awareness set to consist of three Evaluation of alternatives

different subsets. They define the evoked Consumers tend to set rules, or attribute
set as a set of products of which the cut-offs for the products in their choice
consumer has a positive opinion and set. These are the minimum acceptable
from which they are likely to make their levels that an alternative must possess in
purchase. Inert set includes products for order to be considered as the final
which the consumer has neither negative purchase. Huber and Klein28 have
nor positive opinion. The consumer may showed two characteristics to these
be aware of such products but not have cut-offs. The first one states that when
formed an opinion, possibly due to lack the reliability of the attribute the
of knowledge. The final set is called the customer is considering is high (the
inept set, which includes the products for information comes from a reliable source,
which the consumer has formed a at least in the customer’s eyes), the
negative opinion, and therefore these cut-offs on that attribute are more severe
products will not to be purchased. than when reliability is low. The second
No matter how one conceptualises characteristic is that when there is a
these sets, the online environment gives positive correlation between two
the consumers a variety of different tools attributes (for example, low rent in a
for information search to form these sets. high-quality apartment), the cut-offs on
Many companies have utilised so-called those attributes are more severe than
recommendation agents, which are they would be if the correlation was
interactive tools that assist the consumers negative or of equal magnitude. With
in their screening of alternatives based on respect to price, these cut-offs may
the information they have provided. however vary between online and offline
Häubl and Trifts26 have gathered transactions. Bhatnagar and Ghose29
evidence that these recommendation found that price was not one of the
agents ‘reduce the number of alternatives major factors for online shoppers when
for which detailed product information is they evaluate different alternatives. This
viewed’, thus reducing the search time of may be because consumers might think
the consumers. One of the most the web prices are broadly similar and so
common online search methods is they do not need to pay much attention
browsing. Rowley27 asserts that browsing to the price tag.
can be either general or purposive. As consumers reduce their alternatives
Purposive browsing occurs when the to the choice set, they have first acquired
consumer has fairly specific requirements information about the products and then
of the needed product, as opposed to compared and evaluated them. As no
general browsing which is used to keep consumer has unlimited resources (time,
up to date with the latest knowledge in this case) at their disposal, a line must
concerning the product range. Rowley be drawn as to when to stop the
defines three distinctive purposes where evaluation process and make the actual
browsing is preferred to search engines: purchase decision. Hauser, Urban and
(1) the search objective cannot be clearly Weinberg30 propose that consumers
defined; (2) the cognitive burden (ie the allocate their time in such a fashion that
knowledge the consumer needs to have when the trouble of getting extra
in order to do the searching) makes information about the product outweighs
browsing easier; (3) the consumer’s the value of the additional information
system interface encourages browsing itself, information search and evaluation
over other methods of searching. stop. More precisely, consumers try to
maximise the equation 兺svs(ts) ⫹ v0(t0), technology has its advantages. By

where vs(ts) is the value of time ts spent keeping record of individual customers’
gaining information about the product, purchases and personal information,
and v0(t0) is the value of time t0 spent in companies can now target their
other activities. As online evaluations can advertising accurately to each individual
be done from a personal computer, customer,33 thus having a chance to
saving the trouble of travelling from exploit the concept of conceptual fluency
place to place, the equation changes so discussed above.
that we have an increase in t0 and a
decrease in ts.
Still, consumers do not make their Purchase decision
decisions based solely on the information After the evaluation stage, the consumer
about different brands, but also include has ranked the items in the choice set in
how easy it is to process the information some sort of order but not always will
about the brand. A brand that is easily the number one (if there is a number
recognised by its physical characteristics is one assigned) item be chosen. There are
perpetually fluent, while a brand that two factors that come between the
comes to mind easily is conceptually evaluation and purchase decision stages.34
fluent. Lee and Labroo31 demonstrated First there are the attitudes of others,
that in their experiments that a brand where best friends or community
becomes more conceptually fluent if the pressure may change a consumer’s
consumer has been exposed to related preference ranking for a certain brand
products just before they come into even if they intended to buy a different
contact with the target product. one. Secondly there might be some
Following their example, a billboard unexpected situational factors that affect
advertising ketchup produced more the purchase decision. The price of the
conceptual fluency, when it was preceded product may have suddenly gone up, or
with an advertisement for french fries some other purchase becomes more
compared with a washing detergent, thus urgent. As online shopping usually occurs
the message from the billboard was easier in a more private environment, other
to handle. In addition, with high people’s influences are reduced at the
conceptual fluency conditions, the brand point of purchase. Even when a
evaluations tend to be more favourable. consumer has decided the exact product
It is important to note, however, that if they are going to buy, there are still a
conceptual fluency creates negative few purchase sub-decisions to be made.35
associations, its effect will be negative. These sub-decisions include price range,
For example, the aforementioned point of sale, time of purchase, volume
advertisement of french fries may yield of purchase, and method of payment.
negative associations to the people that In spite of all the theories on need
dislike them, which leads to less recognition, information search, and
favourable evaluations of the ketchup evaluation as a fundamental basis for the
ad.32 purchase decision itself, there are
This produces a great challenge for situations where none of these processes
marketers in an online environment, as apply. This phenomenon is called
customers may be faced with a plethora impulse purchasing. Impulse buying
of different advertisements, prior to the occurs when the consumer purchases an
one that is needed to get their attention. item based solely on their impulses and
This is where the new customisation emotions. Baumeister36 submits that there
are some genuinely irresistible impulses, and (postpurchase) satisfaction. Typically,

mainly physiological, from which there is loyal customers tend also to be satisfied
no escape, although they do not but, according to Oliver, satisfaction does
necessarily lead to purchase. This author not necessarily produce loyalty. However,
continues to define three characteristics, studies on several areas of consumer
which form consumer’s self-control. To research such as Auh and Johnson,42 and
fail one of them might result in Ball, Coelho and Machás43 emphasise
impulsive shopping given the right how satisfaction indeed does produce
circumstances. loyalty.
The first of these characteristics are the The importance of satisfaction is as
standards the individual has set for relevant in an online environment as it is
themselves. If consumers have set certain in an offline world. Evanschitzky et al.44
goals and norms and they know exactly have replicated a study by Szymanski and
what they want, they are less likely to Hise45 and have concluded that the single
act on impulse. This also reduces the most important factor forming satisfaction
vulnerability to sales personnel and in electronic commerce is the shopping
advertisers. The second characteristic is convenience, which is a typical
monitoring. People who keep better advantage of online shopping compared
track on their relevant behaviour are less with conventional shopping. Typically, if
prone to lose their self-control. Third a customer is not satisfied with the
and most important, is the consumer’s purchase, there is a chance that they will
capacity to change. Even if the two complain about the product/service. As
previous ingredients fail, there must be the channels involving traditional and
something inside the consumer that is electronic commerce differ, it is not
willing to make the change and purchase surprising that there are noted differences
the certain item their impulses suggest.37 in the complaining threshold and the
Similar results were found by Sojka and degree of dissatisfaction of the purchase
Giese.38 in an online vs an offline environment as
reported by Cho et al.46
In addition to satisfaction and loyalty,
Postpurchase behaviour important issues in postpurchase actions
The purchase process continues even include brand preference and repurchase
after the actual purchase is made. If intentions and how they affect each
marketers and retailers want the other. Hellier et al.47 have proposed a
customers to come back, they must model where all these aspects are
understand their behaviour after the discussed and a number of hypotheses are
purchase as well. Postpurchase behaviour derived. The following conclusions were
can be divided into two subgroups: made: loyalty has a positive effect on
postpurchase satisfaction and postpurchase brand preference; satisfaction has a
actions.39 When it comes to postpurchase positive effect on loyalty; and brand
satisfaction, there is evidence offered by preference and the strength of brand
Mittal and Kamakura40 that consumers preference have a positive effect on
with different characteristics have repurchase intentions. If the
different thresholds when it comes to aforementioned assumption by
loyalty towards the store even if they Evanschitzky et al. concerning the basic
were dissatisfied with their purchase. importance of satisfaction and its
Oliver41 has further discussed the subsequent effects hold, it may be
concepts of loyalty (postpurchase action) assumed that the conclusions offered by
Hellier et al. are also valid in an online this conclusion, the following general
environment. hypothesis is derived:
H1: Gender influences the online

AREAS OF STUDY AND buying process.
PROBLEM CONCEPTUALISATION
The first hypothesis is adjusted to fit the
Gender five-stage model. It is proposed that if
The first hypotheses discuss the the whole online buying process is
differences between online shopping considered to consist of five stages, it
patterns of men and women. The basis would be reasonable to test the
of this research area is the work reported assumption that gender influences all of
by Palanisamy,48 and Rodgers and the stages and not just some of them.
Harris.49 According to Palasinamy, gender Hence, the following hypotheses:
influences the relationship between
online advertisement interpretation and H1a: Gender influences the need
the attitude toward the advertisement recognition phase of the online
which in turn reflects the attitude buying process.
towards the company. Thus, if the H1b: Gender influences the
interpretation of an online advertisement information search phase of the
is the same for both genders, but the online buying process.
attitude towards the ad is more positive H1c: Gender influences the evaluation
by one gender, this would impact the phase of the online buying
perception of the entire buying process. process.
Rodgers and Harris support this theory H1d: Gender influences the purchase
with their results. They claim that decision phase of the online
women are emotionally less gratified and buying process.
have less satisfaction from online H1e: Gender influences the
shopping than men. Men would also be postpurchase phase of the online
more trusting than women and think buying process.
that online shopping is more convenient.
Dholakia and Chiang50 suggest that
men are often considered to be more Differences in consumer online buying
technologically literate than women, a behaviour between 2002 and 2004/2005
fact that is reflected especially in Once the connection between gender and
stereotypes that portray the online the purchase process stages is established,
shopping event. There has been evidence an understanding of how the sampling
found by Garbarino and Strahilevitz51 to period affects these stages is undertaken.
support these stereotypes. Even though There is also evidence that electronic
women have been closing the gender commerce is growing rapidly in Finland53
gap considering internet usage, it is still and in the USA.54 It would therefore be
thought that women perceive higher reasonable to assume that as electronic
risks than men in online purchasing in commerce is growing, online buying
both probabilities and consequences.52 behaviour might also be changing. This
All the evidence above seems to leads to the next general hypothesis:
indicate that there is indeed a difference
in the way men and women perceive H2: The time of sampling affects the
the entire online buying process. From online buying process.
To examine this general hypothesis more following general hypothesis is

thoroughly, it is separated into five proposed:
smaller sub-hypotheses:
H3: There is a positive relationship
H2a: The time of sampling affects the between the consumer’s affiliation
amount of need recognition. to the different stages of the
H2b: The time of sampling affects the buying process and the online
amount of information search. purchase volume.
H2c: The time of sampling affects the
amount of evaluation. Next, it is assumed that the cognitive
H2d: The time of sampling affects the absorption proposed by Shang et al.
amount of purchase decisions. influences all of the components in the
buying behaviour process. Therefore, the
In this study, postpurchase behaviour is general H3 hypothesis is broken down
measured in terms of loyalty (the into five parts for each individual
likelihood of repurchase). As Oliver55 has purchase process stage, where it is stated
stated, loyal customers tend also to be that those who use the internet more for
satisfied, so the term postpurchase need recognition etc. also purchase more
satisfaction is used here. online:
H2e: The time of sampling affects the H3a: There is a positive relationship
amount of postpurchase between the amount of consumer
satisfaction. online need recognition and the
online purchase volume.
H3b: There is a positive relationship
The effects of consumer online between the amount of consumer
behaviour patterns on purchase volume online information search and the
The next point of interest is the online purchase volume.
relationship between purchase volume H3c: There is a positive relationship
and different aspects of consumer between the amount of consumer
behaviour. The following is proposed: online evaluation and the online
the stronger (more positive) the purchase volume.
consumer’s affiliation to any given H3d: There is a positive relationship
phase of the purchase process in an between the amount of consumer
online environment, the higher the online purchase decisions and the
purchase volume will be. Evidence by online purchase volume.
Shang, Chen and Shen56 suggests that H3e: There is a positive relationship
online shopping is not so much a between the amount of consumer
goal-oriented activity rather than a online postpurchase satisfaction
result from cognitive absorption and the online purchase volume.
experiences from the web. So,
according to theory by Shang et al., if If a customer is satisfied with earlier
a consumer discovers a product on the purchases, it should produce store loyalty,
internet and uses it to obtain which means more repurchases.57,58 If a
information about the said product, the customer has made several purchases
cognitive absorption obtained during from a particular website, they may have
the process from the web may well developed a routine pattern for
result in online shopping. Therefore the purchasing online, or at least the
threshold of purchasing online should be online information search and the

lower from that particular site. It is possible changes in purchase
therefore concluded that customers who volume.
are likely to make a repurchase may have H4c: There is a positive relationship
larger online purchase volume than those between the amount of consumer
who have lower likelihood for online evaluation and the possible
repurchases, which leads to a following changes in purchase volume.
hypothesis: H4d: There is a positive relationship
between the amount of consumer
H3e: There is a positive relationship online purchase decisions and the
between the amount of consumer possible changes in purchase
online postpurchase satisfaction volume.
and the online purchase volume. H4e: There is a positive relationship
between the amount of consumer
online postpurchase satisfaction
The effects of consumer online and the possible changes in
behaviour patterns to purchase purchase volume.
volume change
The former hypothesis combines certain
parts of the H2 and H3 hypotheses. The METHODOLOGY
H2 and H3 hypotheses questioned how The sampling plan required that a
the different behavioural characteristics multi-page questionnaire be
affected the possible change in purchase self-administered to randomly selected
volume between the sampling periods. samples of students. The four separate
That is to say that the possible increase in and independent samples used in this
purchase volume might be greater among study were gathered in Finland and the
those with higher amount of online need USA during 2002 and again in
recognition, etc. So if H2 and H3 are 2004/2005. The questionnaires used in
supported by the evidence, the following the USA and Finland were identical
hypotheses may be proposed: except the Finnish students responded to
the questionnaire in Finnish. The content
H4: There is a positive relationship of each question was carefully translated
between the consumer’s affiliation in order to match the English version. In
to the different stages in buying addition, the Finnish students responded
process and the possible changes to the expenditure questions in terms of
in online purchase volume. their local currency (Finnish markka).
The Finnish currency responses were
As with the preceding hypotheses, this converted to US dollars using the
general hypothesis may be separated into exchange rate for the Finnish markka at
five sub-hypotheses: the time of the data gathering.
The first two samples consisted of data
H4a: There is a positive relationship gathered in Finland and the USA during
between the amount of consumer 2002. The first sample consisted of 152
online need recognition and the students attending a college located in
possible changes in online northeast USA. The second sample of
purchase volume. 194 was drawn from students attending a
H4b: There is a positive relationship university located in Tampere, Finland.
between the amount of consumer The third sample gathered data in 2004
from 194 students attending two different Postpurchase actions have been studied
universities in Tampere, Finland. The through store loyalty. The respondents
final sampling in 2005 consisted of 162 were asked whether they were likely to
students at two institutions, one small make a repurchase from an internet site,
private college and one large public assuming they were happy with their
university located in northeast USA. earlier purchase.
In this study, the first of the five All of these questions were measured
stages, need recognition, means either on a three-point scale. The main
needs that are recognised directly online challenge of this study was to determine
as external stimulus, or possible solutions the differences between the portions of
(found online) to already existing needs. the samples in relation to the explanatory
In the latter case the phase actually variable (usually one of the buying
moves from need recognition to process stages). When dealing with
information search as the product shifts categorical variables, such as the 1-2-3
from the total set to awareness set, as scale, chi-square testing was used. Along
defined by Kotler.59 In either case, this with the chi-square statistic and a
was regarded as need recognition and corresponding p-value, the contingency
was measured by the question ‘While coefficient was also computed to find out
online, I have discovered products the strengths of the possible relationships.
and/or services that I am interested in.’ When confronted with scale variables,
Information search means information such as purchase volume, the differences
acquired only from the internet through were sorted with the analysis of variance
active searching. Therefore personal (ANOVA). If the homogeneity of
sources like instant messaging services or variances held true, the regular F-statistic
email do not apply, leaving commercial to calculate the p-value was used, but if
(advertising) or public (third-party this assumption did not hold, the p-value
websites) sources. The corresponding was calculated by using the
question is ‘I have searched websites for Welch/Brown-Forsythe statistics.60,61 As
information about products and/or ANOVA provides only results for sample
services I am interested in.’ mean values to be equal or not,
Evaluation has been measured as active regression analysis with dummy variables
evaluation. Active evaluation means that was used to find out which of the
the evaluation process is happening sample means differed from each other,
through the internet with a possible along with their direction, magnitude
intent to make a purchase at some point. and 95 per cent confidence limits. For
Measurement of evaluation was carried more information on the statistical
out with the question ‘I use websites to analyses carried out in this study, see eg
evaluate different services and/or Greenwood and Nikulin,62 Christensen,63
products, brands, prices, features, and and Draper and Smith.64
their availability.’
Perhaps the most crucial factor in the
whole online buying process is the RESULTS
purchase decision itself. By definition, the
online purchase decision means that the H1: Gender influences the online buying
actual intent to purchase online has indeed process
occurred online. The question ‘I make my The first hypothesis considered that
final product and/or service purchase gender affected the purchase process and,
decisions while online’ measures this. in particular, it was postulated that
Table 1: Hypotheses H1: Gender
Variable Chi-square value p-value Contingency coefficient
Finland
Need recognition 13.152 0.001 0.254
Information search 7.534 0.023 0.194
Evaluation 28.653 0.000 0.360
Purchase decision 17.179 0.000 0.287
Postpurchase behaviour 4.856 0.088 –
USA
Need recognition 1.844 0.398 –
Information search 1.229 0.268 –
Evaluation 1.248 0.536 –
Purchase decision 0.674 0.714 –
Postpurchase behaviour 1.227 0.541 –
gender influenced all of the five stages of relations between gender and the
purchase behaviour when these stages different purchase process stages are
were observed separately. Table 1 displays about equally strong. Although the
the chi-square and p-values along with relation between gender and information
contingency coefficients where necessary search was statistically significant, it is
for all of the five stages of purchase clearly on the weaker side (compared
process. For most up-to-date results, only with need recognition, evaluation, and
the 2004/2005 samples were used. purchase decision). In all of these cases,
Table 1 clearly show how the online men were found to be more active than
shopping process differs in terms of women.
gender in Finland and the USA. The The US sample showed that none of
table shows that with a risk level of the purchase process stages differed when
␣ ⫽ 0.05, the only variable that does not men and women were compared,
reject the null hypothesis of providing strong evidence against the
independence between genders in hypotheses. In the USA, the 2002
Finland is the postpurchase behaviour. sample yielded similar results, showing
The need recognition phase is the most that the online purchase process has
problematic, as one of the expected remained approximately the same with
frequencies was less than unity. However, both men and women over the last few
the difference is so small that the results years. In Finland, however, the earlier
would be significant even if the sample showed that the purchase process
observations were arranged so that the was much more similar between men
expected frequencies were greater than and women two years ago than it is
one. (If one female respondent had now. Should this trend continue in
changed her mind and answered ‘never’ Finland, men would continue to be
instead of ‘sometimes’ to the more active in the purchase process in
corresponding question, the p-value the following years, a fact that companies
would have been 0.003). Therefore, the and marketers need to acknowledge in
conclusion is that the difference is order to get their messages to the desired
significant even with an expected value audience. Given the trend in the USA,
less than one. this is something that should not be of
Need recognition, evaluation, and concern when targeting the US-based
purchase decision have fairly similar audience.
contingency coefficients meaning that the In summary, the Finnish results give
Table 2: Summary of results for H1 hypotheses
H1a: Gender influences the need recognition phase of online buying process. A(FIN)/R(US)
H1b: Gender influences the information search phase of online buying process. A(FIN)/R(US)
H1c: Gender influences the evaluation phase of online buying process. A(FIN)/R(US)
H1d: Gender influences the purchase decision phase of online buying process. A(FIN)/R(US)
H1e: Gender influences the postpurchase phase of online buying process. R(FIN)/R(US)
A = accepted, R = rejected
Table 3: Hypotheses H2: Time of sampling
Variable Chi-square value p-value Contingency coefficient
Finland
Need recognition 12.649 0.002 0.178
Evaluation 30.022 0.000 0.268
Postpurchase behaviour 8.644 0.013 0.187
USA
Need recognition 4.557 0.102 –
Evaluation 19.027 0.000 0.240
Postpurchase behaviour 10.569 0.005 0.199
support to four of the first five in a positive direction, it was concluded

hypotheses, while the US sample did not that there was a positive relation to the
support any of them. Table 2 gives an purchase volume, thus accepting the
overview of these results. hypotheses H2a–H2e. It is to be noted
that even though a positive relationship
between the possible increases in the
H2: Differences in consumer online buying process stages and purchase
buying behaviour between 2002 and volumes were confirmed, no conclusions
2004/2005 concerning the causal directions of these
The second hypothesis suggested that relationships may be established. It is
there have been changes in the way equally likely that a large amount of
consumers approach the general purchase online evaluation triggers the purchase
process. As expected, the amount of decision because avid online buyers tend
online shopping has increased in both to evaluate the products online before
Finland and in the USA. It was also purchase. Table 3 summarises the results
noticed that the average monetary of the chi-square analysis for different
amount of purchase has decreased in stages.
Finland, while the purchase quantities With a 5 per cent risk level, the only
have increased, giving clues about more variables that supported the hypothesis of
‘everyday’ type of shopping. This gives independence (contrary to the
reason to assume that should there be hypotheses) was need recognition in the
any changes in consumer behaviour, they USA. More importantly, ANOVA
are likely to be positive (more need showed that all the increases measured
recognition, information search etc). A were significant. Therefore, it can safely
chi-square test was first performed on all be concluded that there is a positive
of the purchase process stages and if a relationships between the increases in the
statistically significant change was found aforementioned purchase process stages
H2a: The time of sampling affects the amount of need recognition. A(FIN)/R(US)
H2b: The time of sampling affects the amount of information search. A(FIN)/A(US)
H2c: The time of sampling affects the amount of evaluation. A(FIN)/A(US)
H2d: The time of sampling affects the amount of purchase decisions. A(FIN)/A(US)
H2e: The time of sampling affects the amount of postpurchase satisfaction. A(FIN)/A(US)
and the increase in purchase quantity. surfers indicated that pop-ups interfered
When the samples were divided between with their use of webpages. Additionally,
men and women, it was found that all there may be something fundamentally
the stages that were significantly different in the Finnish marketplace.
dependent were also significantly Possibly the strategies used to trigger
dependent when men and women were need recognition are more effective for
observed separately, so neither of the the Finnish customers at this stage.
genders was ‘responsible’ for this positive Thus, it has been established that the
relationship. purchase process is dependent on gender
By definition, need recognition is and sampling time. It is no surprise that
triggered by either internal or external it has also been determined that the
stimuli. Therefore, need recognition is amount of online shopping has increased
passive in nature. Need recognition does in both countries between the two
not require any active measures from the samples. Hypotheses H3 and H4 explore
end user. Marketers could benefit from more deeply into the relationships
identifying the circumstances and between purchase volume increase and
conditions where these stimuli are purchase process behaviour. Table 4
triggered, so that they could summarises the results derived from H2.
communicate the appropriate information
to the customers as suggested by Kotler.65
In Finland, this information has been H3: The effects of consumer online
received rather successfully in the past behaviour patterns to the purchase
couple of years, while in the USA there volume
appears to be no visible change. Due to The third hypothesis suggested that there
this finding, one may question whether is a positive relationship between the
the US online markets are already too affiliation to the different factors that
saturated with information from influence the buying process and the
advertisements. US customers have purchase volume. This hypothesis was
changed their attitudes and behaviour tested with ANOVA and regression
towards such promotional information by analysis using dummy variables. Most of
not paying as much attention to it as the respondents answered that their
they did in the past. These customers affiliation to any given factor was either
feel that many forms of delivery of high or moderate, and therefore the
online promotional information such as responses where the number was equal
banner and skyscraper ads, pop-up and to or less than 5 per cent of the total
pop-under ads are a nuisance, intrusive responses were omitted. The results from
and annoying, and they should be the Finnish 2004 sample supported this
blocked from their online browsing. For theory for all of the five stages. Table 5
example, a 2001 study by nua Internet shows the means of purchased items in
Surveys66 found that 62 per cent of web the last 12 months for all the variables
Table 5: Hypotheses H3: Purchase volume
Amount
High Moderate Low
Mean p-value Mean p-value Mean p-value
Finland
Need recognition 6.54 0 1.80 0.000 – –
Information search 4.99 0 2.03 0.003 – –
Evaluation 4.98 0 2.52 0.017 – –
Purchase decision 6.24 0 2.85 0.003 0.84 0.004
Postpurchase behaviour 8.02 0 2.83 0.000 – –
USA
Need recognition 17.76 0 8.21 0.001 – –
Information search 14.62 0 7.93 0.009 – –
Evaluation 14.97 0 8.70 0.019 – –
Purchase decision 24.96 0 8.77 0.000 2.20 0.000
Postpurchase behaviour 14.74 0 8.08 0.020 – –
followed by p-values from regression the hypotheses, with the exception of

analysis classified by the affiliation (high, postpurchase behaviour on all occasions.
moderate and low) for each variable. The The results from the female sample only
p-values indicate the difference from support the cases of need recognition,
‘high’ internet usage with null hypothesis purchase decision, and postpurchase
being that this difference is 0. Given the behaviour. Table 6 summarises the results
5 per cent risk level, there is a clear from H3a–H3e.
indication of a positive relationship with
need recognition, information search,
evaluation, purchase decision and H4: The effects of consumer online
postpurchase satisfaction in Finland. behaviour patterns to purchase volume
Table 5 shows how the purchase change
decision mean values drop in Finland as It has been established that in both
the level of internet usage in different countries there is an increase in the
stages decreases with the first five purchase volume between the earlier and
variables. When men and women were the later samples, and that there is a
observed separately, few interesting positive relationship between some of the
findings were made. From the Finnish purchase process stages and purchase
sample, the results showed how online volume. The fourth hypothesis suggests
evaluation is positively related to that there is a positive relationship
purchase volume. However neither men’s between the purchase process stages and
nor women’s separate samples supported the increase in purchase volume. While
these findings. The most likely reason is the mean results support this hypothesis
that as the p-value was quite high fairly well, in reality things are not as
(0.042) to begin with, the decreases in simple as they seem. In most of the
usable observations reduced the F-values Finnish samples’ cases, the majority of
enough to raise the p-value above 5 per the respondents belong to the middle
cent. The data also suggest that there are group, leaving rather few observations for
no dependencies between the purchase the extreme ends of the spectrum. This
volume and the levels of information leads to large standard deviations and to
search or purchase decision as far as the fact that the test statistics for
women are concerned. In the USA, the ANOVA and regression analysis would
results from the male sample supported need to be much larger for the extreme
H3a: There is a positive relationship between the amount of consumer online need A(FIN)/A(US)
recognition and the online purchase volume.
H3b: There is a positive relationship between the amount of consumer online A(FIN)/A(US)
information search and the online purchase volume.
H3c: There is a positive relationship between the amount of consumer online evaluation A(FIN)/A(US)
and the online purchase volume.
H3d: There is a positive relationship between the amount of consumer online purchase A(FIN)/A(US)
decisions and the online purchase volume.
H3e: There is a positive relationship between the amount of consumer online postpurchase A(FIN)/A(US)
satisfaction and the online purchase volume.
groups than for the middle group to differences between the mean values of
make any statistical significance. In the the 2002 and 2004/2005 samples.
␤
USA, most of the respondents fell into The conclusion for this hypothesis is
the ‘high’ and ‘moderate’ groups unclear. The mean results give support in
reducing the usefulness of the statistical some form to all of hypotheses H4a–H4e,
analyses as far as the ‘low’ group is but statistical evidence only supports
concerned. hypotheses H4a and H4b in Finland, and
In Finland, the statistical differences H4a, H4b and H4d in the USA. As there
that suggest any support for the general is always the possibility that the mean
H4 hypothesis, were significant only with values were due to chance (that is,
H4a and H4b. The increases in purchase indeed, what the results from ANOVA
volume were significantly larger indicate), all hypotheses that did not
(␣ ⫽ 0.05) for people with a high have statistical support from the analyses
frequency in online need recognition and are rejected. However, it must be
information search. In the USA, recognised that there might be some
supportive evidence was found for H4a, truth in the hypotheses that were
H4b, and H4d. As was the case with rejected. Given more observations, the
Finland, the increases in purchase volume test statistics might show significant
in the USA was larger for people with a differences similar to the hypotheses.
high frequency in online need Table 8 sums up the hypotheses of H4 as
recognition and information search, and they were computed above.
no differences were found in the ‘low’
groups. With postpurchase satisfaction,
the differences were less between the CONCLUSIONS AND
high and moderate groups, but as in the LIMITATIONS
low group, there was no increase at all, From the results it seems clear that
there is enough support for H4d in the online shopping has increased in
USA. Overall the USA increases were popularity among both male and female
significantly larger than Finnish. Table 7 portions of the target groups in Finland,
lists the p-values from ANOVA and and more so in the USA. The internet
regression analysis. If the assumption of also has increased in popularity as a tool
homogenous variances failed,67,68 the used to contribute to and assist in the
statistics were computed instead of the buying process. In spite of the increase
usual F-statistic. Table 7 also lists the in both genders, it seems that in Finland,
␤-coefficients (when statistically different men tend to recognise more needs
from zero) obtained from regression online and use the internet for
analysis. These coefficients show the information search and evaluation more
Table 7: Hypotheses H4: Change in purchase volume
p-value ␤
Finland
Amount of online need recognition High 0.033 3.522
Moderate 0.000 1.072
Low –a –
Amount of online information search High 0.019 2.68
Low –a –
Amount of online evaluation High 0.069 –
Low 0.810 –
Amount of online purchase decision High 0.176 –
Low 0.008 0.685
Amount of postpurchase behaviour High 0.070 –
(likelihood of repurchase) Moderate 0.635 –
Low 0.585 –
USA
Amount of online need recognition High 0.005 12.018
Low 0.743 –
Amount of online information search High 0.002 9.049
Low –a –
Amount of online evaluation High 0.024 8.286
Low 0.015 8.449
Amount of online purchase decision High 0.082 –
Low 0.062 –
Amount of postpurchase behaviour High 0.015 8.825
(likelihood of repurchase) Moderate 0.007 4.538
Low –a –
a
Statistics could not be computed due to a zero-variance variable
H4a: There is a positive relationship between the amount of consumer online need recognition A(FIN)/A(US)
and the possible increase in online purchase volume.
H4b: There is a positive relationship between the amount of consumer online information A(FIN)/A(US)
search and the possible increase in purchase volume.
H4c: There is a positive relationship between the amount of consumer online evaluation R(FIN)/R(US)
and the possible increase in purchase volume.
H4d: There is a positive relationship between the amount of consumer online purchase R(FIN)/R(US)
decisions and the possible increase in purchase volume.
H4e: There is a positive relationship between the amount of consumer online postpurchase R(FIN)/A(US)
satisfaction and the possible changes in purchase volume.
than women. Finnish men also showed a shoppers is bound to be changing over
higher frequency in online purchase time, as new services and products
decisions and postpurchase behaviour. In emerge in the markets, along with better
the USA, there was no such gender gap, logistic solutions for delivery and
showing that the online shopping handling. The two-year span of this
orientation between men and women study, however, was not enough to
did not differ significantly. capture any large changes. Affordable
The distribution for most frequent airline tickets made available online are
still a good example of a product that different stages of the purchase process,
has gained popularity, and although not tends to buy more from the internet.
listed in the questionnaire, online This phenomenon is occurring generally
ordering of train tickets was mentioned with men and women in both of the
by many respondents in Finland. Online target countries. The need recognition
train ticket services are still a fairly new phase in particular is of interest because
phenomenon in Finland, and if the the results show that if a potential
increase in airline tickets is an indication, customer frequently discovers products
the demand for online train tickets may they need while online, they are likely
increase in the future, especially among to make more purchases than those who
the Net Generation in Finland (for use the internet frequently for only
example, students get a 50 per cent information search or evaluation. In
discount from train tickets, plus many of Finland, the results suggest that the
them live away from their hometown). frequency of need recognition has a
The effects on the general population greater influence on the number of
remain to be seen. goods bought from the internet than the
When the buying process is defined frequency of online purchase decisions
by the five-stage purchase model, it is themselves. The results also clearly show
evident that the internet is playing a that it is worthwhile for e-marketers to
continuously increasing role. When the keep their customers satisfied. If the
amount of internet usage related to e-marketer satisfies the customer, that
different phases of the model was customer is a prime candidate for a
measured, a statistical increase in both repurchase. As the total number of
countries from 2002 to 2004/2005 was online purchases continues to increase,
found with the largest increases being in the e-marketer who effectively serves,
the purchase decision itself. There is no satisfies, and delights their online buyers
reason to assume that this trend will will enjoy repeat patronage.
decrease anytime soon. The Net This study has shown that the widely
Generation in both Finland and the USA discussed five-stage model for consumer
have been enjoying affordable broadband buying behaviour has implications in the
connections for some time now, and online environment. Although online
these consumers are familiar with the shopping in general does not depend on
advantages the internet has to offer. This the geographical location (neither buyer’s
trend is also likely to spread to the rest nor vendor’s), the cultural differences,
of the population, as the number of along with the ICT infrastructure, do
internet service providers offering affect the process as was illustrated in this
inexpensive broadband connections study. Even though Finland and the USA
increases. This will certainly push the both have a very high level of
internet penetration rate even higher. No sophistication in the ICT field, the US
changes in need recognition were online shopping field could be described
measured, but a slight increase in the as somewhat more convergent, because
trust factor was noted. In light of current men and women are frequently marketed
internet connectivity technology (ICT) to as separate segments. The increase in
developments, this trend is likely to US online shopping also was significantly
continue in both countries. larger than in Finland. As there is clear
This study shows clearly that the Net growth in online purchase volume in
Generation segment of the population both countries and in fact globally,
with high internet usage relative to the e-marketers may be guided by the
findings of hypothesis H4. E-marketers would have been beneficial in obtaining

should continue to design and deliver more accurate information about the
effective online support for the need behavioural trends in online shopping,
recognition, information search, and instead of just drawing conclusions from
purchase decision stages of the consumer previous trends of development.
buying decision process. Additionally, the According to this and previous reports,
online evaluation and postpurchase however, there is no need to assume that
satisfaction stages should be prioritised for the amount of online shopping will
continued investigation, better diminish anytime soon.
understanding, and improvement.
Particularly in emerging online markets, References
during the introduction and growth 1 Alch, M. (2000) ‘Get ready for a new type of
worker in the workplace: The Net Generation’,
phases of online purchasing, gender may Supervision, Vol. 61, No. 4, pp. 3–7.
influence all but the postpurchase stage 2 World Economic Forum (2004) ‘Networked
of the consumer buying decision process, Readiness Index’; available at:
with males being more active than http://www.weforum.org/pdf/Global_
Competitiveness_Reports/Reports/GITR_2004_
females. These results provide important 2005/Networked_Readiness_Index_Rankings.pdf,
guidance as to how the five-stage accessed 26th May, 2005.
purchase process model could be utilised 3 Kotler, P. and Kelle, K. L. (2006) ‘Marketing
Management’ (12th edn), Prentice Hall, Upper
by e-vendors to improve their marketing Saddle River, NJ, pp. 191–199.
online. 4 Jobber, D. and Lancaster, G. (2003) ‘Selling and
It is essential to acknowledge the Sales Management’ (6th edn), Prentice Hall, Upper
Saddle River, NJ, p. 216.
limitations of this study that reduce the 5 Roth, R. (2001) ‘Yes, customer control is
generalisability of the findings. The main fundamental change’, InternetWeek, February, p. 23.
limitations of this study are the samples. 6 Fader, P. (2001) ‘No, it doesn’t change basic buying
University students, as part of the Net patterns’, InternetWeek, February, p. 23.
7 Kotler and Kelle (2006), see ref. 3 above.
Generation, are extensive users of the 8 Kotler, P. and Armstrong, G. (2005) ‘Marketing: An
latest information technology. This makes Introduction’, Prentice Hall, Upper Saddle River,
them ideal for this study because they all NJ, p. 165.
9 Kinnear, T. and Bernhardt, K. (1986) ‘Principles of
utilise computers frequently. The bias Marketing’ (2nd edn), Scott, Foresman and Co.,
may come from the fact that these Glenview, IL, p. 146.
samples represent only a portion of the 10 Ibid, p. 149.
11 Ibid.
population that have above average 12 Ibid, p. 150.
computer skills, as well as free access to 13 Dubois, B. (2000) ‘Understanding the Consumer’,
the internet. Thus, any generalisations to Prentice Hall, Upper Saddle River, NJ, p. 229.
the general population should be made 14 Van Trijp, H., Hoyer, W. and Inman, J. (1996)
‘Why switch? Product category-level explanations
with caution. Internet usage is, however, for true variety-seeking behavior’, Journal of
increasing in all segments of the Marketing Research, Vol. 33, No. 3, pp. 281–292.
population. Given that this study was not 15 Kotler and Kelle (2006), see ref. 3 above,
pp. 191–192.
intended to establish the proportion of 16 Ibid.
the population that use the internet to 17 Dubois (2000), see ref. 13 above, p. 231.
purchase online, but was intended to 18 Alba, J. and Hutchinson, J. (1987) ‘Dimensions of
consumer expertise’, Journal of Consumer Research,
investigate the buyer decision process of Vol. 13, No. 4, pp. 411–454.
those that do, the Net Generation’s 19 Beatty, S. and Smith, S. (1987) ‘External search
highly literate and heavy online users are effort: an investigation across several product
certainly more qualified and better able categories’, Journal of Consumer Research, Vol. 14,
No. 1, pp. 83–95.
to provide the requisite information. 20 Biswas, D. (2004) ‘Economics of information in the
Additionally, a third reference point Web economy: Towards a new theory?’, Journal of
Business Research, Vol. 57, No. 7, pp. 724–733. 42 Auh, S. and Johnson, M. (2005) ‘Compatibility
21 Urbany, J., Dickson, P. and Wilkie, W. (1989) effects in evaluations of satisfaction and loyalty’,
‘Buyer uncertainty and information search’, Journal of Journal of Economic Psychology, Vol. 26, No. 1,
Consumer Research, Vol. 16, No. 2, pp. 208–215. pp. 35–37.
22 Kotler and Kelle (2006), see ref. 3 above, pp. 43 Ball, D., Coelho, P. and Machás, A. (2004) ‘The
192–193. role of communication and trust in explaining
23 Peter, J. and Olson, J. (1996) ‘Consumer Behavior customer loyalty’, European Journal of Marketing, Vol.
and Marketing Strategy’ (4th edn), 38, No. 9/10, pp. 1272–1293.
Irwin/McGraw-Hill, Boston, Massachusetts, p. 200. 44 Evanschitzky, H., Iyer, G., Hesse, J. and Ahlert, D.
24 Klenosky, D. and Rethans, A. (1988) ‘The formation (2004) ‘E-satisfaction: a re-examination’. Journal of
of consumer choice sets: a longitudinal investigation Retailing, Vol. 80, No. 3, pp. 239–247.
at the product class level’, Advances in Consumer 45 Szymanski, D. M. and Hise, R. T. (2000)
Research, Vol. 15, No. 1, pp. 13–18. ‘E-satisfaction: an initial examination’, Journal of
25 Narayana, C. and Markin, R. (1975) ‘Consumer Retailing, Vol. 76, No. 3, pp. 309–322.
behavior and product performance: an alternative 46 Cho, Y., Im, I., Hiltz, R. and Fjermestad, J.
conceptualization’, Journal of Marketing, Vol. 39, (2002) ‘The effects of post-purchase evaluation
No. 4, pp. 1–6. factors on online vs offline customer complaining
26 Häubl, G. and Trifts, V. (2000) ‘Consumer decision behavior: implications for customer loyalty’,
making in online shopping environments: the effects Advances in Consumer Research, 29, No. 1, pp.
of interactive decision aids’, Marketing Science, Vol. 318–326.
19, No. 1, pp. 4–21. 47 Hellier, P., Geurse, G., Carr, R. and Rickard, J.
27 Rowley, J. (2000) ‘Product search in e-shopping: a (2003) ‘Customer repurchase intention: a general
review and research propositions’, Journal of structural equation model’, European Journal of
Consumer Marketing, Vol. 17, No. 1, pp. 20–35. Marketing, Vol. 37, No. 11/12, pp. 1762–1800.
28 Huber, J. and Klein, N. (1991) ‘Adapting cutoffs to 48 Palanisamy, R. (2004) ‘Impact of gender differences
the choice environment: the effects of attribute on online consumer characteristics on web-based
correlation and reliability’, Journal of Consumer banner advertising effectiveness’, Journal of Services
Research, Vol. 18, No. 3, pp. 346–357. Research, Vol. 4, No. 2, pp. 45–74.
29 Bhatnagar, A. and Ghose, S. (2004) ‘A latent class 49 Rodgers, S. and Harris, M. (2003) ‘Gender and
segmentation analysis of e-shoppers’, Journal of e-commerce: an exploratory study’, Journal of
Business Research, Vol. 57, No. 7, pp. 758–767. Advertising Research, Vol. 43, No. 3, pp. 322–329.
30 Hauser, J., Urban, G. and Weinberg, B. (1993) 50 Dholakia, R. and Chiang, K. (2003) ‘Shoppers in
‘How consumers allocate their time when searching cyberspace: are they from Venus or Mars and does it
for information’, Journal of Marketing Research, Vol. matter?’, Journal of Consumer Psychology, Vol. 13,
30, No. 4, pp. 452–466. No. 1/2, pp. 171–176.
31 Lee, A. and Labroo, A. (2004) ‘The effect of 51 Gabriano, E. and Strahilevitz, M. (2004) ‘Gender
conceptual and perceptual fluency on brand differences in the perceived risk of buying online
evaluation’, Journal of Marketing Research, Vol. 41, and the effects of receiving a site recommendation’,
No. 2, pp. 151–165. Journal of Business Research, Vol. 57, No. 7,
32 Ibid. pp. 768–775.
33 Kotler and Armstrong (2005), see ref. 8 above, 52 Ibid.
pp. 36–37. 53 Sähköisen Kaupan Palvelukeskus (2004) ‘Suomi:
34 Ibid, p. 169. internet-kuluttajakaupan arvo 2,1 miljardia euroa’,
35 Dubois (2000), see ref. 13 above, p. 239. available at: http://www.e-finland.org/
36 Baumeister, R. (2002) ‘Yielding to temptation: cgi-bin/center/xhtml?/pageID=53//list=3/
self-control failure, impulsive purchasing and from=4/size=1/sort=down+25/find=+/show,
consumer behavior’, Journal of Consumer Research, accessed 24th May, 2005.
Vol. 28, No. 4, pp. 670–676. 54 The Census Bureau of the Department of
37 Ibid. Commerce (2005) US Census Bureau News,
38 Sojka, J and Giese, J. (2003) ‘Using individual http://www.census.gov/mrts/www/data/pdf/
differences to detect customer shopping behavior’, 05Q1.pdf, accessed 24th May, 2005.
The International Review of Retail, Distribution and 55 Oliver (1999), see ref. 41 above.
Consumer Research, Vol. 13, No. 4, pp. 337–353. 56 Shang, R., Chen, Y. and Shen, L. (2005) ‘Extrinsic
39 Kotler and Kelle (2006), see ref. 8 above, versus intrinsic motivations for consumers to shop
pp. 198–199. on-line’, Information and Management, Vol. 42, No. 3,
40 Mittal, V. and Kamakura, W. (2001) ‘Satisfaction, pp. 402–413.
repurchase intent and repurchase behavior: 57 Auh and Johnson (2005), see ref. 42 above.
investigating the moderating effect of customer 58 Ball, Coelho and Machás (2004), see ref. 43 above.
characteristics’, Journal of Marketing Research, Vol. 38, 59 Kotler and Kelle (2006), see ref. 3 above, pp.
No. 1, pp. 131–142. 192–193.
41 Oliver, R. (1999) ‘Whence consumer loyalty?’, 60 Welch, B. L. (1951) ‘On the comparison of several
Journal of Marketing, Vol. 63, No. 4, pp. 33–44. means values: an alternative approach’, Biometrika,
Vol. 38, No. 3/4, pp. 330–336. 108–140.

61 Brown, M. B. and Forsythe, A. B. (1974) ‘The 64 Draper, N. and Smith H. (1981) ‘Applied
ANOVA and multiple comparisons for data with Regression Analysis’ (2nd edn), John Wiley and
heterogeneous variances’, Biometrics, Vol. 30, No. 4, Sons, Inc., Hoboken, NJ, pp. 241–250.
pp. 719–724. 65 Kotler and Armstrong (2005), see ref. 8 above, p.
62 Greenwood, P. and Nikulin, M. (1996) ‘A Guide to 204.
Chi-Squared Testing’, John Wiley and Sons, Inc., 66 nua Internet Surveys, 3 May, 2001,
Hoboken, NJ, pp. 86–90. http://www.nua.ie, accessed 28th May, 2001.
63 Christensen, R. (1996) ‘Analysis of Variance, Design 67 Welch (1951), see ref. 60 above.
and Regression’, Chapman & Hall, London, pp. 68 Brown and Forsythe (1974), see ref. 61 above.

Nghiên cứu khoa học - Sưu tầm

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nghiên cứu khoa học - Sưu tầm

Uploaded by

Copyright:

Available Formats

Longitudinal comparison of

Finnish and US online shopping

INTRODUCTION devices almost throughout their entire

Need Information Evaluation Purchase Postpurchase

Recognition Search Decision Behavior

Figure 1: Five stage buying decision process model

(hunger, thirst) or external (passing have suggested that higher purchase

the awareness set to consist of three Evaluation of alternatives

maximise the equation 兺svs(ts) ⫹ v0(t0), technology has its advantages. By

are some genuinely irresistible impulses, and (postpurchase) satisfaction. Typically,

H1: Gender influences the online

To examine this general hypothesis more following general hypothesis is

threshold of purchasing online should be online information search and the

Table 1: Hypotheses H1: Gender

Variable Chi-square value p-value Contingency coefficient

Table 2: Summary of results for H1 hypotheses

Table 3: Hypotheses H2: Time of sampling

Variable Chi-square value p-value Contingency coefficient

support to four of the first five in a positive direction, it was concluded

Table 4: Summary of results for H2 hypotheses

Table 5: Hypotheses H3: Purchase volume

followed by p-values from regression the hypotheses, with the exception of

Table 6: Summary of results for H3 hypotheses

Table 7: Hypotheses H4: Change in purchase volume

Table 8: Summary of results for H4 hypotheses

findings of hypothesis H4. E-marketers would have been beneficial in obtaining

Vol. 38, No. 3/4, pp. 330–336. 108–140.

You might also like