Doing Quantitative Grounded Theory: A theory of trapped travel consumption
Mark S. Rosenbaum, Ph.D.
All is data. Grounded theorists employ this sentence in their
quest to create original theoretical frameworks. Yet researchers
typically interpret the word gdatah to mean qualitative data or,
more specifically, interview data collected from respondents. This
is not to say that qualitative data is deficient; however, grounded
theorists may be missing vast opportunities to create pioneering
theories from quantitative data. Indeed, Glaser and Strauss
(1967) argued that researchers would use qualitative and/or
quantitative data to fashion original frameworks and related
hypotheses, and Glaserfs (2008) recently published book, titled
Doing Quantitative Grounded Theory, is an attempt to help
researchers understand how to use quantitative data for
grounded theory (GT).
Quantitative Grounded Theory
Glaser introduces quantitative grounded theory (QGT) by
providing readers with a historical background of the
methodology, which has ties to Glaserfs sociological training by
Paul Lazarsfeld at Columbia University. Although some readers
may question the purpose of this introductory section, they
should understand that QGT is willing to forgo some empirical
rigor to generate frameworks that can be empirically tested at a
future time. This stance contradicts the empirical rigor that
Lazarsfeld was requesting and which represents the standard in
the United States today. As a result, as social scientists, it is not
surprising that we continue to learn increasingly more about
increasingly less. Today, we clamor for complex structural
equation models that illustrate putative causal relationships
between and among exogenous and endogenous variables, which
are either observed or latent. These research endeavors are far
from inexpensive, as researchers must spend tremendous
amounts of money gathering large data sets that are sizable
enough to replicate an identifiable variance.covariance matrix. I
am confident that contemporary social scientists who publish in
leading journals can relate to the pressures involved in collecting
data and to waiting for a gGod-likeh approval through an RMSEA
or a CFI indicator.
Glaser is not denigrating empirical rigor; however, he makes
readers question whether they can learn as much about the world
by relying on relatively simple chi-square tests. He questions
whether social scientists are wasting data, which many may
perceive as meaningless based on an interpretation of model fit
indexes. Then, QGT sets out to fashion creative models based on
extant data sets and to do so in a way that assumes the data is
nonparametric, but rich enough to capture a social phenomenon.
Glaser spends the next major portion of the book providing
novices with a thorough methodological instruction of QGT.
One of the greatest challenges to accomplishing a QGT study
may be obtaining a data set. The data set must contain variables
that evaluate a socially relevant and interesting condition. In
essence, a core category must arise from quantitative data; thus,
irrelevant empirical data can never lead to a core category.
Although Glaser suggests that researchers can talk to fellow
researchers to obtain raw data sets, I encourage researchers to
explore Internet sites such as the Centers for Disease Control and
Prevention, Roper Center for Public Research, Pew Research
Center, and the U.S. Census for data sets that capture real social
phenomenon. As a way to discuss QGT, I provide an actual
example from one of my data sets.
The data set I selected for QGT was based on an actual
project that I conducted for a 400-square-foot retail store aboard a
ferry that transports people between Oahu and Maui. The trip
lasts for three hours. The ferryfs management sought input
regarding product assortment. On a broader level, many
retailers, such as airlines, bus tours, and railroads, serve
consumers who are gtrappedh for a few hours during
transportation. For this project, a group of MBA students and I
developed a list of products that might be of interest to ferry
customers. In addition, we removed the influence of tourists by
probing responses among residents. The reason for doing so is
because the ferry targets residents who are seeking a more
affordable, albeit longer, form of transportation to Maui; in
contrast, tourists tend to fly interisland. We collected 377
questionnaires from the lower part of the Ala Moana Shopping
Level, which caters to Honolulufs middle-class residents. Of the
respondents, 58% were female, and 42% were male. The average
ages of the respondents were between 25 and 35 years.
Respondents were asked to analyze a list of 70 products and
to indicate (0 = No, 1 = Yes) which of these they would be
interesting in purchasing if they were taking a ferry to Maui.
Respondents were also asked to indicate how much money they
thought they would spend on the ferry. Finally, we obtained data
about respondentsf income level, residency status (as mentioned
previously, only residents could participate), and ethnicity.
Core Index
The first step in QGT is to create a core index. Similar to a
core category, a core index represents a main concern among
respondents. In the ferry study, the core index is based on the
types of products customers planned to purchase to satisfy their
needs during the three-hour excursion to Maui. The core index
represents the combination of two or more questions or items in a
questionnaire. In this study, ten different apparel items, such as
menfs and womenfs T-shirts and childrenfs T-shirts, were added
together to derive an apparel index; three items (dog bowl, collar,
dog T-shirt) were added together for a total pet index, and so
forth. During the summation process, the averages between 0
and .49 were coded as 0, and those between .50 and 1.00 were
coded as 1. In terms of core index validation, Table 1 shows that
at least one respondent displayed an interest in purchasing
products from the indexes. Thus, the core index possesses face
validity.
Table 1 (cannot be shown)
Consistency Indexes and Elaboration
After a researcher finds a core index, the next task is to
develop a theory around the core index by presenting clusters of
items that are associated with the core index. In this study, I
chose to analyze gender as a consistency index. After completing
a series of exploratory cross-tab analyses between gender and
each product from the core index, I elaborated on the study by
considering planned spending on the ship as a third test variable.
To accomplish this task, I divided the respondents into two
mutually exclusive groups: one based on respondents who
planned to spend $1.$20 and one based on respondents who
planned to spend more than $20 (see Table 1).
Creating Formal QGT
Whereas Glaser uses cross-tab analysis to highlight
percentage differences among variables, I encourage researchers
to use cross-tab analysis procedures from statistical packages,
such as Excel, SPSS, SAS, and Minitab, to obtain p-values
associated with a Pearson chi-square, which yield measures of
statistical significance.
A Theory of Trapped Travel Consumption
Modern transportation often means that travelers are
trapped in one place for an extended period, usually any time
from one to three hours. Table 1 represents a first attempt to put
forward a theory of trapped travel consumption. More
specifically, the framework shows the various types of products
that respondents are likely to purchase on a three-hour ferry
service, based on gender and planned expenditures on the ferry.
Although it is intuitive that ferry customers would purchase cold
beverages and motion sickness medication, it is less clear which
other products consumers would purchase during the three-hour
ride. However, the findings indicate that every respondent
planned to spend at least $1 on products during the ferry service.
Thus, the ferry service can essentially increase its profit potential
by fine-tuning its product mix.
Magazine Time
Nearly three-quarters of respondents, in both spending
groups, planned to purchase a local newspaper on the ferry. In
addition, approximately 40% of the respondents, in both spending
groups, planned to purchase a national news publication, such as
USA Today or Newsweek. Over half the women in the highexpenditure
group planned to purchase a female-oriented
magazine, such as Cosmopolitan, and 37% of the women in the
low-expenditure group planned to do so. In contrast, 17% of the
men in the high-expenditure group planned to purchase a maleoriented
magazine, and 14% of the men in the low-expenditure
group planned to do so. Of interest, nearly 20% of the men in the
low-expenditure group indicated that they planned to purchase a
leading U.S. paper, such as the New York Times or the Los
Angeles Times, and 11% of the men in both the low- and the highexpenditure
groups planned to purchase a leading business
magazine, such as Forbes or the Economist. These results show
that reading is integral to travel consumption, and thus the ferry
service should consider outsourcing this department to a national
chain such as Barnes & Noble. This tactic would enable
consumers to use their loyalty card to save money and enable the
ferry service to benefit from Barnes & Noblefs merchandising
expertise.
Catch Some Rays
The data also reveal that the overwhelming majority of
respondents indicated likelihood to purchase sunscreen on the
ferry. This finding shows that the purchase of sunscreen is not
planned but rather arises from the circumstances. Furthermore,
a desire to purchase sunscreen on the ferry emerged in both the
low- and the high-expenditure groups. Thus, the ferry service
could capitalize on such a last-minute purchase by selling
sunscreen at value prices.
Eat, Drink, and Be Merry
It is intuitive that ferry customers would purchase cold
beverages during the service. Yet the data also reveal that more
than one-half of the women in the low-expenditure group and
nearly two-thirds in the high-expenditure group planned to
purchase snacks on the ship, including candy, chips, gum, and so
forth. Only 44% of the men in the low-expenditure group planned
to purchase snacks, and this figure increased to approximately
59% among men in the high-expenditure group. Thus, the ferry
service might consider focusing on snacks that are appealing to
women, such as those that are healthier for the family in general.
Reading and Writing
Another major finding is that approximately one-third of the
respondents, across gender and planned expenditures, planned to
purchase books, stationary, and greeting cards, perhaps in an
effort to catch up on a novel or to send friends a letter. Indeed,
without e-mail access, many consumers might recall how
enjoyable it was to receive a post card from a friend or a greeting
card from a family member.
Get the Sun Out of My Eyes
Given that most of the respondents forgot sunscreen at
home, it is not surprising that more than one-third of the men in
the low-expenditure group and one-half in the high-expenditure
group planned to purchase a baseball cap on the ferry.
Need Tunes
Approximately one-third of the men in the high-expenditure
group planned to purchase CDs, DVDs, or game cartridges on the
ferry, and more than 20% of the women in that group planned to
do so. This percentage reduces to half among those in the lowexpenditure
group.
A Floating Pharmacy
Another surprising finding is that between 20% and 25% of
all respondents indicated that they would purchase cold medicine
or eye solutions on the ferry, respectively.
Conclusion
Overall, the data reveal that consumers are in an active
purchasing state during trapped travel occurrences, which
provides an organization the opportunity to maximize its profit
potential by developing an appropriate product mix. Taken as a
whole, the data reveal that consumers view a three-hour trip as a
means to catch up on enjoyable pastimes that are lost in todayfs
fast-paced, high-technological world. Although planned
purchasing rather than actual purchasing was explored in this
example, the emergent theory typifies consumers seeking a
respite from technology by purchasing newspapers, books,
magazines, greeting cards, and even CDs and DVDs. In addition,
a trapped travel occurrence is a time when consumers tend to
take a break from their health regimen and indulge in sweet and
salty snacks. Finally, trapped travel occurrences spur the
purchase of essential items, such as sunscreen and baseball caps.
Future researchers can easily expand on the theory of
trapped travel consumption, especially those in hospitality and
marketing. Pioneering opportunities abound for research to
explore family size and household income as additional variables.
Furthermore, the rising cost of fuel in the United States will most
likely influence peoplefs consumption during travel excursion.
The impetus of this article was to show readers how to follow
Glaserfs (2008) QGT methodology by illustrating a theory of
trapped travel consumption from actual quantitative data. In
many ways, QGT is somewhat easier than traditional GT because
the researcher has the flexibility to learn from each respondent
and to constantly change questions. A questionnaire is a finite
representation of reality, and whether QGT can develop a theory
that possesses conceptual grab remains unknown. Thus, a QGT
researcher faces the challenges of working with a completed data
set to see latent theoretical patterns in a series of numbers, which
can be conceptually interesting.
I anticipate that researchers who opt to use QGT will be
questioned as to their desire to use cross-tab analyses to tease out
an emergent framework from a data set rather than employing a
more rigorous structural equation modeling or analysis of
variance program. By easing methodological constraints, such as
multivariate normality or non-correlated error terms, researchers
acquire the freedom to explore a data set and to generate an
original framework that yields hypotheses that can be empirically
tested at a future time. Glaserfs contribution in doing QGT is
that he encourages researchers to counterbalance empirical rigor
with theoretical sensitivity, as doing so leads to hypotheses that
can be empirically verified. Unfortunately, too many social
scientists clamor for rigor by forgoing theoretical sensitivity.
Author:
Mark S. Rosenbaum, Ph.D.
Northern Illinois University
College of Business Administration
Department of Marketing
DeKalb, IL 60115
mrosenbaum@niu.edu
References
Glaser, B. G. (2008), Doing Quantitative Grounded Theory. Mill
Valley, CA: Sociology Press.
Glaser, B. G., & Strauss, A. L. (1967), The Discovery of Grounded
Theory: Strategies for Qualitative Research. New York:
Aldine de Gruyter.
