Performance Measurement and Metrics

LibQUAL+™ spring 2001 comments: a qualitative analysis using Atlas.ti

Julie Anna Guidry,
Article information:
To cite this document:
Julie Anna Guidry, (2002) "LibQUAL+™ spring 2001 comments: a qualitative analysis using Atlas.ti", Performance
Measurement and Metrics, Vol. 3 Issue: 2, pp.100-107, https://doi.org/10.1108/14678040210429008
Permanent link to this document:
https://doi.org/10.1108/14678040210429008
Downloaded on: 15 August 2017, At: 07:14 (PT)
References: this document contains references to 5 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 586 times since 2006*
Users who downloaded this article also downloaded:
(2002),"LibQUAL+™ and the small academic library", Performance Measurement and Metrics, Vol. 3 Iss 2 pp. 85-95 <a
href="https://doi.org/10.1108/14678040210429982">https://doi.org/10.1108/14678040210429982</a>
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)

(2002),"Extracting meaningful measures of user satisfaction from LibQUAL+™ for the University Libraries
at Virginia Tech", Performance Measurement and Metrics, Vol. 3 Iss 2 pp. 48-58 <a href="https://
doi.org/10.1108/14678040210440937">https://doi.org/10.1108/14678040210440937</a>

Access to this document was granted through an Emerald subscription provided by emerald-srm:264987 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.


TM
LibQUAL+ spring 2001
comments: a
qualitative analysis
using Atlas.ti
Julie Anna Guidry
The author
Julie Anna Guidry is a PhD student in the Marketing
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
Department, Texas&M University, College Station, Texas,
USA. E-mail: jguidry@cgsb.tamu.edu
Keywords
Qualitative techniques, Computer software,
Service quality, Libraries
Abstract
Respondents' comments to the LibQUAL+TM spring 2001
survey were examined to refine the instrument and
reduce non-sampling error. Using qualitative data analysis
software, Atlas.ti, respondents' unsolicited e-mail
messages were analyzed. Results showed that the major
problem with the survey was its length, which was due to
a combination of factors. This information helped the
survey designers in reducing the number of library service
quality items from 56 to 25 and in addressing technical
problems from the Web-based survey. An in-depth
discussion of the steps followed in conducting the Atlas.ti
analysis will also be discussed.
Electronic access
The research register for this journal is available at
http://www.emeraldinsight.com/researchregisters
The current issue and full text archive of this journal is
available at
http://www.emeraldinsight.com/1467-8047.htm
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . pp. 100±107
# MCB UP Limited . ISSN 1467-8047
DOI 10.1108/14678040210441008
100
Introduction
One of the major philosophies behind the
design of the LibQUAL+TM survey
instrument is the use of an iterative mixed
methods approach. A combination of
qualitative research in the form of interviews
with faculty, staff, and students and
quantitative research in the form of surveys
administered to that same population has
been employed through two iterations, with
the first round of qualitative and quantitative
research conducted in 2000, and the second
round conducted in 2001. The purpose of the
qualitative research presented here is to
complement the quantitative findings of the
spring 2001 LibQUAL+TM survey to refine
the instrument (cf. Cook and Thompson,
2001; Thompson et al., 2002). While
quantitative research can provide much of the
needed information for fine-tuning
LibQUAL+TM, qualitative research
associated with the survey provides insights as
well. Specifically, this article presents the
findings of an analysis of respondents'
unsolicited comments to the 2001 survey. An
additional purpose of this paper is to
introduce readers to the qualitative data
analysis program, Atlas.ti, which was used to
conduct the analysis.
Spring 2001 LibQUAL+TM survey
The second iteration of the LibQUAL+TM
survey took place in spring 2001. The
instrument consisted of four demographic
items, 56 library service quality items, and five
optional items. For each of the 56 library
service quality items, survey participants were
asked to provide their minimum, desired, and
perceived service levels, on a nine-point scale.
Thus, the library service quality section
demanded 168 (3  56) responses from
participants. The survey required respondents
to navigate through several WWW pages that
incorporated Java script design in which
participants clicked on buttons to indicate
their responses. Those included in the sample
were e-mailed three invitations to participate
in the survey, which could be accessed
through a hypertext link included in the
e-mail. Respondents were informed in the
e-mail message that the survey would require
ten minutes to complete, and that those who
completed the survey would be included in a
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
drawing for a PalmPilot. A total of 43 libraries
participated in the survey, and the final
sample size totaled 20,416.
Although the survey did not include open-
ended questions, many participants felt
compelled to provide feedback about the
survey or their institution's library by
e-mailing the LibQUAL+TM survey
administrator at their institution. These
emails were collected and analyzed using
Atlas.ti, a qualitative data analysis program.
The remainder of the paper discusses how the
analysis was completed using Atlas.ti, the
results of this analysis, and the refinements of
the survey based on this analysis.
Atlas.ti analysis
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
Atlas.ti, a qualitative data analysis program,
was used to discover patterns in respondent
comments. While Atlas.ti allows the
researcher to manage large amounts of
qualitative data, it does not perform the
analysis for the researcher, i.e. it is not an
artificial intelligence program that sifts
through data to discover themes. Rather,
Atlas.ti provides a tool for the researcher to
organize and document themes within his/her
data. In order to provide a better
understanding of the analysis and to
familiarize readers with the program, the
three steps used to conduct the analysis in
Atlas.ti will be discussed. This illustration is
not meant to show the entire array of features
available in Atlas.ti but rather to provide a
general understanding of the program.
The first step in the analysis was to prepare
and import the primary data. This step is
somewhat analogous to data preparation in
quantitative analysis, in which the data file
must be cleaned and edited before it is
imported into a quantitative data analysis
program such as SPSS or SAS. For the
LibQUAL+TM project, the primary data, the
text of the respondents' e-mail, was collected
from various libraries. Duplicate e-mails were
removed. Additional primary data cleaning
included removal of nonsensical characters
and symbols associated with the e-mail
transmission. One distinct advantage of
analyzing e-mail messages in Atlas.ti is that
they are already in text format and do not
have to be transcribed, as does a tape-
recorded interview or focus group. Once the
primary data for each institution was cleaned,
101
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
each institution's collection of e-mail
responses was saved as a text only file. These
files were uploaded into a hermeneutic unit in
Atlas.ti. A hermeneutic unit is an Atlas.ti file
in which all information, links or paths to this
information, related to a specific project, are
stored. Thus, all information regarding the
spring 2001 LibQUAL+TM survey comment
analysis was stored in one hermeneutic unit.
E-mail comments for 37 of the 43 libraries
were collected and analyzed.
Once the data were imported into Atlas.ti,
the second step, coding the data, was
initiated. This step was conducted using the
document view in Atlas.ti, in which one
window displays the primary document, one
of the 37 text files of the actual e-mails
imported for each school, and a second
window displays the codes that are assigned to
each quotation, the text within the primary
document that the researcher considers to be
meaningful. Assigning codes involves first
highlighting a particular quotation within the
primary document and then creating a code
for that quotation, i.e. open coding, or
selecting a previously created code, i.e. code
by list.
There are several considerable advantages
of using Atlas.ti as opposed to a non-
computerized method for coding the data.
First, once the data have been coded, Atlas.ti
allows quick access to the quotations for a
particular code. A window can be opened for
each code that displays the first few words of
every quotation for that code with its location,
noting the primary document of the quotation
and its line number within the primary
document. By clicking on the quotation, the
quotation's primary document is retrieved
and shown within the primary document
window with the particular quotation
highlighted. Second, Atlas.ti provides a search
feature that lets the researcher find patterns or
strings, specified by the researcher, within the
primary document. Third, the program allows
the researcher to assign more than one code
to a quotation. The double coding strategy is
often useful in the beginning of the analysis in
which the codes have not been thoroughly
developed; the researcher may not be able to
decide at that point exactly how the data
should be coded. Finally, Atlas.ti also allows
the researcher to assign more than one
quotation to the same piece of text. For
instance, the first two-thirds of a sentence
could belong to one quotation, while the last
 LibQUAL+TM spring 2001 comments Performance Measurement and Metrics
Julie Anna Guidry Volume 3 . Number 2 . 2002 . 100±107

two-thirds could belong to a second
quotation. Thus, in this case, the middle third
of the sentence would be associated with two
different quotations and, most likely, two
different codes. If this were attempted
manually, the researcher would likely be
required to duplicate the sentence so that it
could be stored in two locations, one for each
code.
In all, 36 codes were created, as shown in
Table I. The length of the quotations, for the
most part, was approximately one line or
sentence long. However, some quotations
were as short as one word while others were
several paragraphs long. The longer
quotations were most often those comments
associated with a particular institution's
library rather than comments about the
survey itself. Because the purpose of this
study was to refine the survey instrument,
comments about a particular library most
often, as opposed to the survey, were not
analyzed in detail. Thus, this information
was most often left in long strings.
Additionally, some long quotations captured
specific error messages produced while the
respondent was taking the survey. Many
respondents cut and pasted these error
messages into their e-mails, some of which
were up to 35 lines long.

Table I List and frequency of codes generated from LibQUAL+TM 2001 survey comments
Code Frequency
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)

Already completed/stop reminding me 44

Anti-spamming 33
Cannot look ahead/backward 3
Cannot start/access survey 54
Comments about library 119
Complete page one but said they did not 8
Completed page four but said they did not 37
Confused about how to finish 6
Confusing format 14
Cookie problem (browser set but still does not work) 7
Does not address major issues or problems 47
Does/has not used library 81
Error/froze after demographic information 18
Font too small 26
Forced choice/may not have an opinion 60
General comment about survey 60
General confusion/ambiguity 50
Graphics hide buttons 5
No distinction between librarians and students 9
No distinction between libraries 28
No distinction between service and holdings 4
No space for additional comments 66
No time 28
Other/specific technical problem 81
Positive comments 13
Problem with specific question/aspect 27
Problem with demographic items 27
Questions too intrusive 5
Repetitive/long 241
Right side of screen not visible 3
Scaling issues 16
Server not responding/too slow/timed out/froze 38
Technical requirements too demanding 26
Too specific/tedious/detailed 20
Unable or unwilling to complete survey 22
Understanding of minimum/desired/perceived 18
Total 1,344

102
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
The third step in the analysis involved
synthesizing the codes to form broader
categories. This process is then used to
develop analytic frameworks for the data
collected under a constructivist grounded
framework (Charmaz, 2000). Although
Atlas.ti provides several methods for
organizing data, this analysis was achieved
using the following method, which was
thought to be the most flexible. Within
Atlas.ti, the network view provides a graphical
representation of the data for model building
purposes. This view allows the researcher to
import the codes that were previously created,
that are represented by boxes called nodes.
Nodes may represent entities other than
codes, such as memo, quotations, or primary
documents; however, in this analysis, nodes
represented only codes. These nodes can then
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
be related to one another through the use of
lines, or links. Invoking grounded theory,
each link may show a different type of
relation, such as ``is associated with'', ``is part
of'', ``is cause of'', etc., that is selected by the
researcher.
` ... Atlas.ti allows any node within the
network view to be linked with any
other node, allowing for a rich
representation of the data... '
The general structure of the synthesized data
in this analysis forms a hierarchy, in which
codes are grouped into categories, categories
are grouped into higher-level categories, and
so forth. It is important to understand that
within the network view, Atlas.ti does not
distinguish between these different levels.
Rather, these are all considered nodes in
Atlas.ti. The researcher, however, can form a
hierarchical structure. To do this, the
researcher would first create a new node in
the network view. The original codes
developed during the coding process can then
be linked to this new node, or, in the
researcher's terms, category. For example, the
codes/nodes of ``right side of screen not
visible'' and ``graphics hide buttons'' were
linked to the new node named ``visual
problems''. In order to further synthesize the
data, higher-level categories were created in
the same fashion, which subsumed codes and/
or categories. For example, the higher-level
category labeled ``technical problems''
included both ``visual problems'', a category,
103
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
and ``other/specific technical problems'',
a code.
While a hierarchical representation of the
data is normally desired, it may be restrictive
if complex relationships exist in the data.
Fortunately, Atlas.ti allows any node within
the network view to be linked with any other
node, allowing for a rich representation of
the data.
Results
Figure 1 shows the overall dimensions of all e-
mail messages using Atlas.ti. Additionally,
tallies of each code are provided in Table I. In
all, 1,344 comments were coded. It must be
emphasized that these comments were
unsolicited, in that there was no question in
the survey that asked respondents to provide
their opinions. Thus, it is likely that these
respondents were highly motivated to provide
feedback and that less motivated people who
did not provide feedback felt similarly.
As can be seen in Figure 1, the comments
were broken into two major categories: those
related to the survey itself and those related to
a particular library. E-mail comments related
to a specific library totaled 119, as shown in
Table I. Again, because the purpose of this
analysis was to improve the LibQUAL+TM
survey instrument, comments about a library
were not further analyzed.
Those comments related to the survey itself
were also broken into two categories: positive
comments and problems. Positive comments
totaled 13 and reflected varying aspects of the
survey. Problems with the survey, on the
other hand, accounted for a very large
number of comments about the survey itself:
1,212 from 1,225. While this number may
seem alarmingly high, it must be kept in
perspective: 20,416 people participated in the
survey and some people provided more than
one comment. Thus, no more than 6 percent
(1,212/20,416) of respondents provided
negative feedback.
The problems with the survey reported by
respondents are likely a reflection of non-
sampling error, i.e. error in a survey method
design other than sampling error. Sampling
error refers to the random error involved in
drawing a sample and is the statistically
estimated distance between the sample results
and the estimated true population results
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
Figure 1 Dimensions of e-mail comments received during spring 2001 LibQUAL+TM survey
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
(Hair et al., 2000). Unlike sampling error,
however, non-sampling error:
. is systematic;
. is controllable;
. cannot be statistically estimated; and
. is iterative, meaning one type of error
often generates another type.
The purpose of the Atlas.ti comment analysis
was to determine the sources of non-sampling
error, correct the errors and, thus, improve
the next iteration of the LibQUAL+TM survey
instrument. Although this analysis used a
grounded theory approach, in that the
literature on non-sampling error was not
consulted immediately before the analysis was
initiated, allowing patterns in the data to
emerge without the influence of previous
theory, the results found here are consistent
with non-sampling concepts. Thus, the
problems with the LibQUAL+TM survey
identified through respondent feedback will
be discussed employing a non-sampling error
conceptual framework.
The problems with the survey were grouped
into five categories: technical problems;
survey content ± demographic items; survey
content ± service quality items; administrative
and formatting issues; and respondent
problems.
Technical problems
The first category of problems, ``technical
problems'', included three sub-categories.
Blocked from finishing survey referred to
104
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
several general technical problems that
occurred during the survey. These were likely
due to Internet traffic, a slow connection, or
possible server problems. Another set of
technical problems that prevented
respondents from completing the survey
concerned survey software dynamics. For
instance, 37 comments noted that the survey
would not move past the fourth page,
indicating the respondent had not completed
the page even though they had. Additionally,
another bug, in which an error occurred after
respondents completed the demographic
items, was reported 18 times. An
incompatibility problem between the
respondents' computer system and the survey
software was another technical problem that
respondents faced. Visual problems referred
to technical problems that may have caused
error, although the respondents were still able
to complete the survey, e.g. small size, was
mentioned 26 times. All of the above-
mentioned technical problems can be referred
to as survey instrument design error, which
``represents a `family' of design or format
errors that produces a questionnaire that does
not accurately collect the appropriate raw
data'' (Hair et al. 2000, p. 276). These
findings helped the LibQUAL+TM team
remove the technical bugs and also motivated
them to design a survey with fewer Java script
applications, which would allow more
expedient downloading of the survey, as well
as address incompatibility problems.
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
Survey content ± demographic items
The second category of problems was labeled
``survey content ± demographic items'', with
27 respondents reporting problems with these
items. Within in this category, respondents
noted that either their ``academic discipline'',
e.g. law, or their ``position'', e.g.
administrator, was not one of the possible
responses. This problem is considered scaling
measurement error, which ``occurs when
inaccuracies are designed into the various
scale measures used to collect the primary raw
data'' (Hair et al., 2000, p. 275). A small
number of respondents pointed out that the
age and gender items were inappropriate or
unnecessary. These items may have caused
response error, which occurs when
respondents unconsciously or deliberately
misrepresent their answers (Hair et al., 2000).
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
These problems were noted and plans were
made to address them in the 2002 survey
iteration.
Survey content ± service quality items
``Survey content ± service quality items'' was
the third category of problems. Several factors
associated with the service quality items
influenced the length of the survey. By far, the
most common problem respondents reported
was that the questions were repetitive, which
accounted for 241 comments, more than any
other. Another comment was that the survey
was too tedious or detailed (20 respondents).
That comment may have been driven by the
fact that nine scale points (rather than the
more common five or seven) were used. Some
respondents (16) felt that nine points were
too many, and these people felt that it was too
hard to discriminate between the scale points
when responding to an item. Together, the
repetitiveness of the questions and the
tediousness of the survey contributed to the
amount of time needed to complete the
survey. An overly long questionnaire is
considered a survey instrument design error
and can increase forms of nonresponse error,
which occurs when a number of people are
not included in the final sample (Hair et al.,
2000). Indeed, some reported that they did
not finish the survey or expressed doubts that
others would due to the survey's length. The
survey's length may have also caused response
error. Some respondents showed hostility
(Hair et al., 2000), noting that survey was
``torture'' or ``waaaaaaaay too long'' and that
the questions were ``mind-numbing''.
105
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
Furthermore, the survey's length may have
influenced ``yea- and nay-saying'' (Hair et al.,
2000). For instance, several respondents
noted that they began filling out the survey in
good faith but, once they realized the survey's
length, they began mindlessly completing the
survey simply to enter the PalmPilot drawing.
These problems were addressed when
developing the spring 2002 survey
instrument. Based on the results of this
qualitative analysis, along with the results of
the quantitative analysis, the number of
library service quality items was reduced from
56 to 25. The use of the nine-scale point,
however, has remained the same in order to
properly differentiate between respondents'
minimum, perceived, and desired levels of
service, as recommended by Parasuraman,
Berry and Zeithaml in gap analysis (cf.
Parasuraman, 2000).
A number of respondents reported that
some questions were ambiguous. For
instance, some commented that they could
not accurately indicate the level of service
quality because they did not know to which
branch library the survey was referring.
Additionally, some respondents stated they
could not readily discern the differences
among the three service levels, minimum,
desired, and perceived, upon which the gap
model is based. Both of these problems may
have contributed to scaling measurement
error, in that the ambiguity of the items may
have confused the respondents. Some of these
problems were addressed by the removal of
the less reliable items, as mentioned above.
` ... The repetitiveness of the questions and
the tediousness of the survey
contributed to the amount of time
needed to complete the survey. An
overly long questionnaire is considered
a survey instrument design error and
can increase forms of nonresponse
error... '
The third problem associated with the library
service quality items was that some
respondents felt as though the survey did not
address important issues. Several respondents
(47) indicated this explicitly in their
comments, often stating that there were not
enough items that addressed the library's
holdings. Others (66 respondents) indirectly
expressed this by complaining that the survey
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
did not provide space for respondents'
feedback, implying that the survey did not
address their particular problems with their
library. Both of these problems would suggest
construct development error, which occurs
when the researcher does not adequately
identify important constructs related to the
research problem. Furthermore, some
respondents noted that the absence of an area
to provide feedback also made them feel as
though the survey was too impersonal. To
rectify these concerns, a free text comments
section was added to the 2002 version of the
survey.
The forced choice format was the final
problem associated with the library service
quality items. Each item allowed the person to
either score all three levels of service quality or
choose the ``N/A'' response. A total of 60
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
respondents felt that neither of these options
allowed them to accurately reflect their
attitude. For some items, these respondents
could not indicate a perceived level of service
since that had not experienced that service,
yet they did wish to express attitudes about
their minimum and desired levels of service.
The problem with forced choice is likely a
reflection of sample design error, or
``systematic inaccuracies by using a faulty
sampling design to identify and reach the
selected `right' respondents . . .'' (Hair et al.,
2000, p. 276). Thus, those persons who felt a
need to express their desired and minimum
service levels but did not have attitudes of
perceived levels of service were often those
new to the library who had not yet had the
opportunity to form perceptions of some
services.
Administrative and formatting issues
The fourth category of problems pertained
to ``administrative and formatting issues''.
Many respondents felt that the processes
associated with administering the survey
were too intruding, with 44 persons annoyed
that they were continually reminded to
complete the survey although they had
already done so. Another interesting
problem in this category was that several
people (26) felt the survey had too many
technical requirements, in that respondents
were requested to change their computer's
configurations, e.g. accepting cookies and
modify their monitor setting. Many of these
comments indicate nonresponse error was
possible. Another issue related to this fourth
106
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
category concerned the directions provided,
with some claiming that the instructions
were not truthful (completing the survey
took longer than specified) or were not clear
regarding how to begin or finish the survey.
The actual format of the survey was also a
recurring comment. Some respondents
specifically indicated that the number of
buttons was ``dizzying'', while others noted
that the font was too small, which was
actually related to a technical problem.
Respondent problems
Finally, the last set of problems was labeled
``respondent problems'', meaning the
respondents were either unwilling or unable
to complete the survey. For instance, 33
respondents were upset about being sent
unsolicited e-mail. Also, several respondents
simply did not have time to complete the
survey, which may have been related to the
survey's length. Both of these comments
influenced nonresponse error. In addition to
these comments, 81 respondents stated that
they either no longer used the library (were
retired or moved) or had recently begun to
use the library and, thus, had not yet formed
an opinion about particular aspects of it.
These comments may reflect sample design
error and were related to the forced choice
format problem mentioned previously. These
problems reflect issues associated with the
Web-based survey design, in which
respondents must be contacted via e-mail.
Unfortunately, these problems are not easily
remedied. However, the advantages of a
Web-based survey, such as the convenience
afforded to the respondent and the absence of
data entry error, were believed to outweigh
these problems.
Conclusion
Although the quantitative analysis of the 2001
LibQUAL+TM results helped to refine the
survey, that analysis could not reflect all of the
non-sampling errors associated with the
survey. The major impetus for this qualitative
analysis of unsolicited respondent comments
was to uncover these non-sampling errors
and, thus, to further refine the LibQUAL+TM
survey instrument. The most common
problem respondents reported was the
survey's length, which was due to a
combination of factors, such as the number of
 LibQUAL+TM spring 2001 comments
Julie Anna Guidry
library service quality items, the tedious
nature of the survey, and the time it took for
respondents to download the survey. These
problems were addressed in the redesign of
the current 2002 LibQUAL+TM instrument
in which the number of items was reduced
from 56 to 25, and the survey programming
was rewritten to be less dependent on Java
script, making the downloading time more
reasonable. Additionally, Atlas.ti, the
qualitative data analysis program used to
conduct this analysis, was introduced to the
researchers. The benefits of employing
qualitative analysis in research using this new
Atlas.ti software will encourage researchers to
include qualitative analysis in their future
research designs.
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)
107
Performance Measurement and Metrics
Volume 3 . Number 2 . 2002 . 100±107
References
Charmaz, K. (2000), ``Grounded theory: objectivist and
constructivist methods'', in Denzin, N. and Lincoln,
Y. (Eds), The Handbook of Qualitative Research,
Sage, New York, NY.
Cook, C. and Thompson, B. (2001), ``Psychometric
properties of scores from the Web-based
LibQUAL+TM study of perceptions of library service
quality'', Library Trends, Vol. 49, pp. 585-604.
Hair, J.F. Jr, Bush, R.P. and Ortinau, D.J. (2000), Marketing
Research: A Practical Approach for the New
Millennium, Irwin McGraw-Hill, Boston, MA.
Parasuraman, A. (2000), ``The SERVQUAL model: its
evolution and current status'', paper presented at
the New Measures Symposium of the Association of
Research Libraries, Washington, DC.
Thompson, B., Cook, C. and Thompson, R.L. (2002),
``Reliability and structure of LibQUAL+TM scores'',
Portal: Libraries and the Academy, Vol. 2, pp. 3-12.
 This article has been cited by:

1. Porcia Vaughn, Cherie Turner. 2016. Decoding via Coding: Analyzing Qualitative Text Data Through Thematic Coding
and Survey Methodologies. Journal of Library Administration 56:1, 41-51. [CrossRef]
2. Michael Luther Kennesaw State University, Kennesaw, Georgia, USA . 2015. Comment sifting: pragmatic qualitative analysis
of LibQUAL + Comments. Library Hi Tech News 32:9, 8-13. [Abstract] [Full Text] [PDF]
3. Wonsik Shim, Eun-Chul Lee. 2013. Service Quality Assessment of University Libraries in Korea using LibQUAL+ : A Case
Study. Journal of the Korean Society for information Management 30:2, 245-268. [CrossRef]
4. Jennifer Rosenfeld and Raida GattenB. Jane ScalesWashington State University Libraries, Pullman, Washington, USA. 2013.
Qualitative analysis of student assignments: a practical look at ATLAS.ti. Reference Services Review 41:1, 134-147. [Abstract]
[Full Text] [PDF]
5. Michael J. Roszkowski, John S. Baky, David B. Jones. 2005. So which score on the LibQual+™ tells me if library users are
satisfied?. Library & Information Science Research 27:4, 424-439. [CrossRef]
Downloaded by Georgetown University At 07:14 15 August 2017 (PT)