E-surveying
for tourism research: Legitimate tool or a
researcher's fantasy?
Journal of Travel Research;
Boulder; Feb 2001; Stephen W Litvin; Goh Hwai Kar;
Abstract:
Tourism
researchers continually seek to improve their primary data collection methods.
With the spread of Internet and e-mail
technologies, various researchers have begun to explore the potential and
efficacy of electronic data collection. This article reports on a study that
compared respondent data from two data sets based on the same survey instrument.
the first, collected via a traditional mall-intercept-type data collection
exercise and the second, an "e-sample" collected from random e-mail
addresses. Analysis of the responses found similar travel psychographic
characteristics but significant differences in demographics and travel patterns.
The article concludes that while there seems to be legitimate potential for tourism
research "e-surveying, " current logistical problems and issues of
sample bias remain serious stumbling blocks precluding widespread use of the
method
[Headnote]
Tourism researchers
continually seek to improve their primary data collection methods. With the
spread of Internet and e-mail technologies, various researchers have begun to
explore the potential and efficacy of electronic data collection. This article
reports on a study that compared respondent data from two data sets based on the
same survey instrument. the first, collected via a traditional
mall-intercept-type data collection exercise and the second, an
"e-sample" collected from random e-mail addresses. Analysis of the
responses found similar travel psychographic characteristics but significant
differences in demographics and travel patterns. The article concludes that
while there seems to be legitimate potential for tourism research
"e-surveying, " current logistical problems and issues of sample bias
remain serious stumbling blocks precluding widespread use of the method
That the Internet and
e-mail are here to stay may be the greatest understatement of the new century.
The technologies have become pervasive, while the Internet has emerged as the
fastest-growing technology adoption in history. Among American adults, Internet
penetration, only 9% in 1995, is expected to exceed 60% by the end of year 2000
(Beniger 1998). Internationally, there are currently more than 1 million users
in China, 10 million in Japan, and 500,000 in India (Computer Almanac 1999),
while the United Kingdom records more than 10,000 new Internet subscribers daily
(Reed 1999).
For travel researchers,
the extension of electronic telecommunication technologies across the general
population has the potential to create new and interesting research tools.
Tourism researchers, requiring primary data, always weigh alternative approaches
to data collection. Traditional methods include mail surveys, telephone
canvasses, and mall-intercept approaches, each with its own distinct advantages
and disadvantages as a means of extracting data from a small group of
respondents that can then be generalized to the greater population of interest.
These methods are discussed in most research methods or tourism marketing texts.
Good examples include Sekaran (1992) and Rubin and Babbie (1993) for general
research methods and Kotler, Bowen, and Makens (1999) and Seaton and Bennett
(1996) for tourism marketing.
Now peeking over the
horizon, but not discussed in the above texts, is the use of electronic
communications for primary data collection. To coin a phrase (undoubtedly
already claimed), "e-surveying" is an obvious extension of traditional
data collection methods, offering researchers the potential to reach mass
numbers of respondents in a potentially efficient and cost-effective manner. A
review of the literature found a limited but growing body of research exploring
the potential use of the Internet for sampling purposes. Bradley (1999) looked
at the still embryonic field and prepared a technical overview of available
electronic sampling methods, in which he discussed various approaches to Web
page questionnaires versus three types of e-mail forms-- simple, attached, and
URL embedded-and segmented the different types of users into 13 categories
dependent on their different computer skill levels and computer capabilities.
Readers contemplating the use of e-surveying would be wise to invest some time
familiarizing themselves with the technical basics, from this or other
introductory articles.
A fair amount of the
early research on the use of electronic surveying has looked at its pros and
cons specifically versus a mail survey approach (e.g., Schaefer and Dillman
1998; Oppermann 1995; Mehta and Sivadas 1995; Medlin, Roy, and Theong 1999).
Among the plusses discussed are the elimination of the tedium of folding and
stuffing of the instruments, the immediate delivery of the instrument and the
opportunity of an equally quick response, the fact that responses can be
received around the clock-a major plus for overseas respondents, the current
mystique still associated with electronic communications, the fact that the
method is environmentally friendly, and the opportunity to introduce a more
complex instrument design (Weible and Wallace 1998). Some of the disadvantages
include the bias of limiting respondents to Internet subscribers, who may not
constitute a representative sample frame (Tse 1998, p. 354), and the inability
to enclose reply incentives. In addition, Caroll (1994) has suggested that the
medium's lack of format and layout flexibility results in less attractive forms
destined to yield unsatisfactory response rates. It has also been argued that
e-mail respondents find unsolicited mail to be more of an intrusion of privacy
than they do such mail received via post (Mehta and Sivadas 1995).
A review of the
literature, however, did not find any work comparing e-mail surveying with
mall-intercept data collection, so common for tourism research. Such a
comparison is the focus of this research.
Why is mall-intercept
so popular in tourism research? For academics the main attraction is cost.
Mall-intercept tends to be less expensive than other traditional methods and can
often be accomplished with student helpers and research assistants. It is also
quicker than mailed survey techniques, as the data can often be collected during
a reasonably short period of time without the need to wait for postal responses
and second requests. Could e-surveying provide these same benefits?
This article compares a
mall-intercept survey sample with an "e-sample." (Although it would
have been useful to also compare the data with telephone or mail surveys,
collection of such data was beyond the scope of the research.) In doing the
comparison of the two exercises, there were two main goals. The first was to
explore the relative efficacy of the two methods. The second was to enable
comparison of the resultant samples, specifically from a tourism research
perspective. Comparability was measured in four distinct domains: demographics,
selected travel-related psychographic characteristics, travel usage, and use of
the Internet and e-commerce, including online travel purchases. If the two
samples proved to be relatively similar (the primary research proposition), it
seems appropriate to begin to consider e-surveying as a replacement or
supplement to mall-intercept sampling techniques, thus providing tourism
researchers with a new tool in their research toolbox.
METHOD
Samples
Two independent data
collection exercises were conducted in Singapore, the site for this research
during the fourth quarter of 1999. Identical (except for one question related to
e-mail addresses) two-page 30-item survey instruments were used for each.
The first survey was
administered using a modified mall-intercept approach. Data were collected
during one weekend from a busy mass-rapid transit station and attached midscale
shopping mall in the central shopping district of Singapore (thus the qualifier
modified, as not all data were collected in a shopping mall). To gather the
data, research assistants were instructed to approach each fifth adult and when
encountering refusals, to simply ask the next available person. No respondent
younger than 16 years was asked to complete the form, as Cooper et al. (1993)
indicate that this is the age when one first takes some responsibility for
travel decision making. The research assistants were provided 300 forms. During
the weekend, they exhausted their preallotted supply, yielding 279 substantially
complete and usable responses, a number sufficiently large to support the
statistical methods employed in this research. This sample has been termed
mall-intercept.
Mall-intercept
surveying by design always introduces an element of convenience sampling into an
otherwise random sampling effort as respondents are restricted to the specific
locales selected for data collection, and, as such, the resultant sample is not
expected to be an exact representation of the population as a whole. In this
research, the mall respondents are, in fact, skewed toward the middle and upper
class, representative of those Singaporeans more prone to be found at an in-town
shopping area on the weekend. Rather than being a limitation of the research,
however, this can be viewed as a strength, as it has provided a sample, it
seems, more reflective of those likely to partake in travel than would a less
skewed sample population.
The 279-element sample
consisted of 146 men and 133 women, ranging in age from 16 to 60, with a mean of
28.9 years (SD = 8.9). Approximately 40% of the sample reported having attained
either an "A-level" certification or a polytechnic diploma (both
somewhat the equivalent of a U.S. junior college-level education), and a further
34% had earned university or higher qualifications. Approximately 84% had annual
household incomes greater than U.S.$15,000, and more than 21% reported income
levels in excess of U.S.$57,000. Most participants were Chinese (81.3%), with
Malays (7.2%), Indians (6.8%), and others (4.7%), an ethnic diversification that
represents a reasonable reflection of Singapore's multicultural population
(Central Intelligence Agency 1999). Eighty-five percent of respondents were
Singaporeans. (See Table 1 for full demographic disclosure.)
The second sample was
collected via e-mail, with the goal of obtaining responses from approximately
the same number of respondents as from the mall-intercept collection exercise.
Two thousand questionnaires, sent via e-mail to a randomly selected set of
e-mail addresses, yielded 237 usable responses. To find these 2,000 addresses, a
random number table was used to select names from the Singapore telephone
directory. For each name selected, an e-mail address was sought via one of the
several free-access commercial e-mail directories available on Internet, such as
Yahoo, Switchboard, and so on. It was not expected that exact matches would
necessarily be found. When no exact name match was found, as was often the case,
or when a selected common name yielded multiple matches, an e-mail address
related to a name, as close alphabetically as possible, with either a Singapore
domain or terrestrial address, was selected.
It was determined, for
sake of simplicity, that a document format versus a Web-based format would be
used (see Bradley 1999). Doing so allowed the research to be accomplished
without outside, in this case campus computer center, assistance. The first 500
names on the list were sent a standard e-mail letter serving as cover letter and
participation request, with the survey instrument an attached document. The
cover letter requested the receiver to open the attachment file, to click on the
appropriate response boxes, and to return the form to the sender as an attached
reply. The survey form appeared very professional and when pretested with 20
students, worked effectively and efficiently. However, when sent to the first
500 names, quite a few recipients replied, via e-mail, that either they would
not open the attachment for fear of infection by computer virus (certainly a
valid concern) or that they tried but found their system unable to open the
attached document. As such, the approach was modified and the remaining requests
were sent with the 30 questions incorporated into the body of the e-mail
document itself, with respondents asked to type responses onto the form itself
and send a reply message to the authors. Although not as professional in
appearance, this approach met far less resistance. Seven days following each
mailing, second requests were sent to nonrespondents.
The selection method
used to gather the addresses for the e-sample can only be assumed to have
yielded a representative cross section of the population of e-mail address
holders in the republic. There is, however, no means to identify the specific
characteristics of these 2,000 potential respondents. What is learned from this
research is limited to the characteristics of the 237 actual respondents, who,
in fact, may or may not be representative of the gross sample of potential
respondents. This, however, is neither a limitation nor a problem, as the intent
was specifically to learn about those respondents who would choose to
participate in an e-mail-based sample. (Characteristics of e-survey respondents
are also detailed in Table 1.) Of the 2,000 forms e-mailed, 766, or 38.3%,
yielded variations of the dreaded electronic response "Mail System Error,
Returned Mail," indicating a wrong address, closed e-mail account, or a
full mail box. For the remaining 1,234 addresses, 237 responses were received,
52% from first requests, the balance from second responses, representing a quite
credible 11.9% total response and 19.2% effective response rate. The average
time-lag between first requests and responses received was 1.4 days. For second
responses, the mean response time was 2.8 days, from the date of second mailing.
Questionnaires
Singapore is a
multilingual state with English as the common unifying language. As such, all
data collection was conducted in English only, avoiding translation difficulties
or linguistic bias (Smith and Bond 1993). The two questionnaires used were
identical, except for one question that asked if the respondent had an e-mail
address. In addition to classification data, the questionnaires asked a series
of questions to learn travel frequency, use of the Internet, and several travel
psychographic characteristics of the respondents. Travel frequency was
determined by asking subjectively, "How often do you travel for vacation
travel purposes?" and "How often do you visit or speak with a travel
agent about vacation plans?" with 5-point response formats of 1
=frequently, 2 = often, 3 = sometimes, 4 = rarely, and 5 = never provided. The
participants were then asked to report the number of vacation trips taken in the
past 12 months.
Three dependent
psychographic variables-vacation travel innovativeness, subjective vacation
travel knowledge, and vacation travel opinion leadership-were explored using
tests previously applied to tourism research by Goldsmith and Litvin (1998,
1999). Vacation travel innovativeness was measured using the Domain Specific
Innovativeness scale, or DSI. This six-item self-report Likert-type scale asks
survey respondents to reveal their tendency to purchase new products or services
soon after they appear on the market. One item read: "In general, I am
among the last in my circle of friends to visit a new vacation spot."
Previous studies have demonstrated the reliability and validity of the DSI for
both products (Goldsmith and Hofacker 1991) and services (Flynn and Goldsmith
1993). Perceived vacation travel knowledge was measured by a three-item scale
that reflects the respondent's subjective knowledge of the given product field.
One item read: "Among my circle of friends, I'm one of the 'experts' on
tours or where to travel." A 5-point agree-- disagree response format was
used. Vacation travel opinion leadership was measured by the seven-item,
5-point, King and Summers (1970) Opinion Leadership Scale as modified by
Childers (1986). One item read: "Compared with your circle of friends, how
likely are you to be asked about places to visit on vacations?" The travel
innovativeness, vacation travel knowledge, and vacation travel opinion
leadership scales were each independently summed so that higher scores indicated
greater levels of the constructs.
Finally, the
respondents were asked several questions related to their use of the Internet.
The first two were dichotomous questions that read, "Do you have an e-mail
address?" (for the e-sample, this question read, "Do you have multiple
e-mail addresses?") and "Have you ever used the Internet to research
travel options?" Two additional questions determined frequency of Internet
use for online travel and for other product purchases.
RESULTS
Demographically, when
the two samples were tested for significant differences, three variables were
found to vary between the groups. The e-sample was found to be higher educated,
from a household earning a higher income, and more likely to be single than the
sample collected using the mall-intercept method. (See Table 2 for results and a
description of the statistical methods used for comparison of the two samples.)
For travel frequency,
the data revealed that e-survey respondents took 50% more vacation trips in the
past 12 months (2.4 vs. 1.6), but, interestingly, when responding to the
question regarding subjective travel frequency, each sample mean reflected the
descriptor sometimes, the midrange response on a scale ranging from frequently
to never. The e-sample, however, indicated a higher usage of travel agencies
than did the mall-intercept sample.
The four questions
regarding the use of the Internet and e-mail all found significant differences
between the two populations, with the e-sample, not surprisingly, more likely to
have e-mail and more likely to both research and purchase travel-related and
nontravel products online.
Psychographically,
there were no differences between the two populations. The three variables
tested-travel innovativeness, subjective travel knowledge, and travel opinion
leadership-each yielded similar results between the samples. (Nondemographic
variables are detailed in Table 3, with statistical methods used for comparison
of the samples noted.)
OBSERVATIONS AND
DISCUSSION
Today, any general
tourism research based on primary data collected via a random e-mail survey
technique would be highly suspect. Down the road, however, e-surveying seems to
have potential as a new data collection method for tourism researchers. But
before research can be entrusted to the method, work such as the current
research needs to enlighten researchers to both the reliability issues and the
operational pitfalls of using the new technologies. This article, decidedly
exploratory in nature and not intended for generalization to other locales where
Internet usage profiles may be different from Singapore's, should be seen as a
starting point in the process.
What was learned from
the exercise? In response to the proposition that for travel research a
mall-intercept sample would be similar to an e-sample, one finds mixed results.
Psychographically, for those variables tested on a Singapore population, it was
found that electronically solicited respondents were not significantly different
from those sampled randomly using the mall-intercept method. This implies that
Singaporean researchers concerned with such dimensions as travel innovativeness
or travel opinion leadership may consider using e-survey techniques to conduct
their study, with reasonable confidence that their findings can be generalized
to their broader local population.
However,
demographically, an e-sample would not likely cover the full population range,
as researchers can expect to find respondents generally more educated, of higher
income, and less likely to be married than the population as a whole. In
addition, the electronic respondents were heavier travelers, reporting having
taken 50% more vacation trips in the past year. For much tourism research these
differences would not constitute a problem. In fact, travel market researchers
would likely find such a sample population more attractive than a true cross
section of the population.
Logistically,
e-surveying a broad population is still a difficult proposition. In the
literature, a fair number of examples of research based on electronic data
collection were noted. However, in each case, the population sampled was a
closed set. For example, Tse (1998) selected his sample from his university
directory, Weible and Wallace (1998) sampled Management Information Science
professors selected from a faculty list, Oppermann (1995) based his findings on
an electronic sample selected from the membership ranks of the Association of
American Geographers, and so on. These tests, however, unlike the current
research, do not attempt to inform the effectiveness of the technique for a more
generalized population.
For the current
research, the actual process of e-survey data collection was relatively
easy-requests were sent and responses received without having to venture beyond
the office, avoiding the costs and discomfort of face-to-face solicitation
required by the mall-intercept effort. Easy, that is, except for the gathering
of random addresses, which was both time-consuming and tedious. Not only was
gathering of addresses a difficult task, the list collected was found to be
disappointingly noncurrent, with, as mentioned previously, more than one-third
of the addresses undeliverable. A plus, however, was the instant feedback of
undeliverable mail, which, Oppermann (1995) pointed out, facilitates the
addition of names when necessary. In the future, it is likely that researchers
will have access to commercial e-mail lists similar to today's mailing lists,
both reducing the effort of obtaining and increasing the accuracy of e-mail
address lists. If respondents from these lists can be targeted to be
representative of a desired population, the logistics of e-mail surveying will
be greatly enhanced.
The 19% response rate,
for nonrejected addresses, seemed fairly encouraging, although Schudlt and
Totten's (1994) metaresearch indicated that users of electronic sampling
consistently received lower response rates than those using mail methods, a
finding mirrored by Tse et al. (1995); Medlin, Roy, and Theong (1999); and
Weible and Wallace (1998).
Cost advantages offered
by e-surveying over a mall-intercept effort can be substantial. D'Onofrio (1999)
has commented in his study that, with literally no paperwork, e-mail surveying
represents an inexpensive method. Mehta and Sivadas (1995) have also concluded
that the marginal costs of collecting data electronically are much lower than
costs of interviewing, telephoning, and sending questionnaires through the mail.
These savings may be even more substantial if the respondents are scattered
worldwide. For the current research, it is difficult to compare costs of
collecting data for the two samples. Data collection for the mall-intercept
method, including printing and compensation to research assistants, incurred a
cost of approximately U.S.$400. The only cost of the e-surveying was the time
invested in gathering addresses and sending the mail, which was significant, and
clearly in excess of the amount spent on the mall-intercept approach. However,
the lion's share of this time investment was the manual task of mining the 2,000
random e-mail addresses. In the future, what will be the cost of a good random
or targeted list of Singapore e-mail addresses? If it were, let's say, U.S.$.05
to .07 per address (approximate current average cost of purchased street address
lists available in the market), then the cost of the 1,234 accurate addresses
used for this research would have been about U.S.$75.00. With software that
would allow mailing at the push of a button, the cost of distribution and
collection would have been less than one-third that of sending research
assistants to the mall and clearly a smaller fraction of a mailed survey
instrument.
Tangential to the key
purpose of this research, but of interest, are the differences between the two
samples regarding their use of new technologies. It is not surprising that
e-mail respondents use the Internet more often to purchase items online. What is
surprising, though, for both samples, is the limited dispersion of online travel
purchasing, with only 11% of the e-mail sample and 6% of the mall-intercept
sample reporting having made online travel purchases. Respondents are, however,
using the Internet to research travel (77% of the e-sample and 56% of the
mall-intercept sample report having done so), but their failure to use the
channel for purchases leaves unsettled, at least in Singapore, the argument
regarding inroads that electronic commerce will make on the travel distribution
system. At the current time, it seems that the Internet has been a plus for
travel agencies. Consumers' apparent willingness to do their own research should
result in more informed customers-and informed customers tend to be good
customers. In this time of decreasing travel agency commissions, having clients
share in their own research is an advantage for which travel agencies should be
grateful, as long as agents remain the booking channel of choice.
CONCLUSION
As previously
acknowledged, mall-intercept surveys are known to produce an inherently biased
sample, always creating some concern when the research findings are generalized
to a broader population. Despite this, and through the use of such strategies as
strata sampling or segment weightage to compensate for nonresponse bias (Gelman
and Little 1998; Rogers 1991), the method has been well embraced by tourism
researchers, particularly when studying the attitudes and behaviors of potential
travelers. If e-surveying provides a sample that comes close to being similar to
that obtained by a mall-intercept, then the advantages of the new method,
particularly its quick turnaround and cost-efficiency, should lead researchers
to consider its adoption.
For the Singapore
market, the current research seems to indicate that there may be occasions when
e-surveying may be used with caution, perhaps as a hybrid supplement to
traditional data collection methods. For example, as suggested by Schaefer and
Dillman (1998), e-mail could be used as a first mode of contact, followed by
progressively more expensive methods directed to nonrespondents until an
acceptable response level is reached.
A concern, however, is
that with rapid change of the underlying technologies, it is possible that the
research community will find it difficult to reach a consensus regarding the use
of e-survey techniques. At one time, mail and telephone surveys were likewise in
their infancy. With experimentation and usage, protocols were established for
these methods, and researchers now use them with confidence. However, as
conventions for e-surveying are being developed, researchers may find that just
as they begin to understand the issues, diffusion, usage habits, and
technological progression may have left them behind.
In line with this
concern, it is interesting to consider, even as e-mail usage becomes ubiquitous,
how use of e-surveying may change as the gap between e-mail users and a more
generalized population narrows and the two populations in effect become one. In
Singapore, this should not take long, as already 42% of all households have
e-mail access, up from 9% in 1996 (Tee 2000). (For comparison's sake, Tee [2000]
reported 1999 Internet penetration of 40% in the United States, 22% in
Australia, and 13% in Japan.) However, this does not address the question of who
among this wider market will respond to future e-mail surveying requests. Will
respondents be reflective of the broader user base? At this time, many
respondents undoubtedly find electronic survey participation novel, which may
explain the decent response rate in the current research. Bombard these same
people with multiple requests, however, and it is likely they may begin to click
and delete future survey requests as quickly as one can say "click and
delete."
Given the challenges of
having e-surveying as an accepted tourism data collection method, much more
research, in many more locales, is required before the research community can
begin to feel comfortable with the technique as a replacement for mall-intercept
or any other sampling method. Other researchers are encouraged to repeat and
refine this work in a search for answers. It would also be very useful if
researchers incorporated, at least on a limited basis, e-surveying as a
supplement to their primary data collection method. The resultant findings may
not, in and of themselves, prove valuable to their immediate research task at
hand, but their reporting of similarities and differences discovered between
their traditionally gathered sample data (whether collected using
mall-intercept, telephone surveys, or any other method) and their e-sample will
add to our e-- surveying body of knowledge, perhaps speeding the day when
tourism researchers will be able to use this new data collection tool with
confidence.
[Reference]
REFERENCES
[Reference]
Beniger, J. R. (1998).
"Presidential Address: Survey and Market Research Confront Their Future on
the World Wide Web." Public Opinion Quarterly, 62 (3): 442-50.
Bradley, N. (1999). "Sampling for Internet Surveys. An Examination of
Respondent Selection for Internet Research." Journal of Market Research
Society, 41 (4): 387-95.
Caroll, S. (1994). "Questionnaire Design Affects Response Rates."
Marketing News, 28 (1): 14.
Central Intelligence Agency (1999). The World Factbook. Assessed January 2000 at
http://www.cia.gov/cia/publications/factbook/country.html
Childers, T. (1986). "Assessment of the Psychometric Properties of the
Opinion Leadership Scale." Journal of Marketing Research, 23 (May): 184-88.
[Reference]
Computer Almanac (1999).
"Numbers about Computers: The Internet." Assessed in December at
http://www.cs.cmu.edu/afs/cs.cmu.edu/user/ bam/www/numbers.html
Cooper, C., J. Fletcher, D. Gilbert, and S. Wanhill (1993). Tourism: Principles
and Practice. London: Pitman.
D'Onofrio, M. (1999), "Web, E-mail Surveys Cheap but Have
Disadvantages." Advertising Age's Business Marketing, 84 (6): 17.
Flynn, L., and R. Goldsmith (1993). "Identifying Innovators in Consumer
Service Markets." The Service Industries Journal, 13 (3): 97-109.
Gelman, A., and T. C. Little (1998). "Improving on Probability Weighting
for Household Size." Public Opinion Quarterly, 62 (3): 398-404.
Goldsmith, R., and C. Hofacker (1991). "Measuring Consumer
Innovativeness." Journal of the Academy of Marketing Science, 19 (3):
209-21.
[Reference]
Goldsmith, R., and S. Litvin
(1998). "Identifying the Vacation Travel Innovator." Journal of
Vacation Marketing, 4 (4): 327-36.
------(1999). "Heavy Users of Travel AAAgents: A Segmentation Analysis of
Vacation Travelers." Journal of Travel Research, 38 (2): 127-33.
King, C., and J. Summers (1970). "Overlap of Opinion Leadership across
Consumer Product Categories." Journal of Marketing Research, 7 (February):
43-50.
Kotler, P., J. Bowen, and J. Makens (1999). Marketing for Hospitality and
Tourism. 2d ed. Upper Saddle River, NJ: Prentice Hall.
[Reference]
Medlin, C., S. Roy, and H. C.
Theong (1999). "Internet versus Mail Surveys: A Comparison and
Report." Presented at a seminar at Ngee Ann Polytechnic, Singapore, August
7. Unpublished paper distributed at the seminar.
[Reference]
Mehta, R., and E. Sivadas
(1995). "Comparing Response Rates and Response Content in Mail versus
Electronic Surveys." Journal of the Market Research Society, 37 (4):
429-39.
Oppermann, M. (1995). "E-mail Surveys-Potential and Pitfalls."
Marketing Research, 7 (3): 29-33.
Reed, M. (1999). "E-Commerce: An Era of Confusion." Marketing, June
17, 27-28.
[Reference]
Rogers, J. (1991). "A
Non-Technical Perspective on Data Collection Methodologies for Travel Surveys: A
Discussion Paper." Journal of Travel Research, 29 (3): 43-47.
Rubin, A., and E. Babbie (1993). Research Methods for Social Work. 2d ed.
Pacific Grove, CA: Brooks/Cole.
Schaefer, D., and D. A. Dillman (1998). "Development of a Standard E-mail
Methodology: Results of an Experiment." Public Opinion Quarterly, 62 (3):
378-97.
Schudit, B. A., and J. W. Totten (1994). "Electronic Mail vs. Mail Survey
Response Rates." Marketing Research: A Magazine of Management and
Applications, 6 (1): 36-39.
Seaton, A. V., and M. M. Bennett (1996). Marketing Tourism Products; Concepts,
Issues, Cases. London: International Thomson Business Press.
Sekaran, U. (1992). Research Methods for Business, 2d ed. New York: John Wiley.
[Reference]
Smith, P., and M. Bond (1993).
Social Psychology across Cultures, Analysis and Perspectives. New York:
Harvester Wheatsheaf.
Tee, E. (2000). "S'pore Beats US, Japan in Internet Use at Home." The
Straits Times [Singapore], 21 January, 74 (Home Section).
Tse, A.C.B. (1998). "Comparing the Response Rate, Response Speed and
Response Quality of Two Methods of Sending Questionnaire: E-mail vs. Mail."
Journal of Marketing Research Society, 40 (4): 353-61.
Tse, A.C.B., C. T. Ka, H. Y. Chow, B. T. Choy, W. Y. Ko, P. Y. Kwan, and C. H.
Wing (1995). "Comparing Two Methods of Sending Out Questionnaire: E-mail
versus Mail." Journal of Market Research Society, 37 (4): 441-47.
Weible, R., and J. Wallace (1998). "Cyber Research: The Impact of the
Internet on Data Collection." Marketing Research, 10 (3): 19-26.
[Author note]
Stephen W Litvin is an
associate professor and Goh Hwai Kar is a tourism management graduate research
student, both in the Nanyang Business School of Nanyang Technological University
in Singapore.
