This overview describes
research software that is available at UT and how OIT's Research Computing
Support (RCS) consultants can help you use it. To keep current about
new software versions and our services, see our monthly newsletter.
Type of Software
Full Support(Recommended, supported, tested, taught)
(Provided, but support for only installation, import/export)
JMP Genomics, SAS/Genetics
Data Acquisition & Web Surveys
SPSS Data Collection Web Interviews, LabVIEW, NVivo
Graphics & Visualization
ATLAS.ti, MATLAB, NVivo, R, SPSS Graphics, SAS/Graph
JMP, Origin, SAS/IML Studio, SigmaPlot, Stata
ATLAS.ti, MATLAB, NVivo, R, QDA Miner
Mapping & Geographic Information Systems
ESRI ArcGIS, JMP, MATLAB , R, SAS/GIS, SPSS
Scientific, Engineering, Mathematics
FORTRAN, Maple, MATLAB, PuTTY, Xming
Autodesk Suite, Mathematica, Origin, Scientific Notebook
EQS, HLM, JMP, LISREL, Mplus, SamplePower, Stata, SUDAAN
There are four ways you can use research software at
UT: licensed on your computer, in the OIT Computer Labs running on a PC or
Macintosh computer, from our Windows Terminal Server, Apps@UT (apps.utk.edu), or our high performance
Newton Linux Cluster.
If you own a computer and the software you need is
available at a price you can afford, you may prefer running software on
it. For a comprehensive listing of software available for departmental purchase
through OIT, see http://oit.utk.edu/software. You
will also see software there that you can download for free or
purchase on CD or DVD for a minimal fee from the UT Bookstore Technology Center
(see https://utbookstore.org/technology/). These
include Amos, ATLAS.ti, EndNote, JMP, LabVIEW, Maple, MATLAB, QDA Miner,
SamplePower, SPSS, SPSS Text Analysis for Surveys, and WordStat.
The OIT computer labs have an extensive selection of
software available on PCs and Macs and are open to students, faculty, and staff
for free. For a list of software, hardware, and locations, see http://oit.utk.edu/labs. Faculty
can also reserve labs for class use from that web page.
Our Apps@UT system,
lets you run software as if it was installed on your own computer or
on a computer in the labs. Actually, the software runs on a powerful server,
and it only displays an image of it on your computer. You can open and save
files on your local disk drive and print to your local printer. You can learn
the simple steps to use this system by reading How to Use
This server is free for students, faculty, and staff on the UT Knoxville
Our Newton Linux Cluster is our most powerful system
with over a thousand processors. However, it does require learning a way of
working that is generally considered harder than using a desktop computer. For
more information, see https://newton.utk.edu/.
you have a research problem to solve, you will of course want to find the
software best suited to the problem itself and to your computing resources. The
software description below are brief but do not hesitate to contact us to
discuss your analysis needs in more detail.
A very important consideration in choosing software is
the type of computer you prefer. Due to the wide popularity of Microsoft
Windows, all of the 50+ packages discussed below are available for it. However,
of the software listed below, only about 35% are available on Macintosh and
Linux computers. Macintosh users have only SPSS, JMP, Stata, ImageJ,
MATLAB, Maple and R to choose from. Linux users have SAS, SPSS,
Stata, ImageJ, MATLAB, Maple, and R to choose from. Any computer you use
can run the software we make available on our Windows Terminal Server and our
UNIX systems. On the Mac, both VMware Fusion and Parallels usually run Windows
software well. Booting a Mac or Linux system directly into Windows usually
guarantees full compatibility.
When using a complex software package, sooner or later
you will have questions about how to get the most out of it. You should ensure
that the package you choose has support available at the level of depth, cost,
and timeliness to meet your needs. You may get support from colleagues,
Internet discussion lists, vendor web pages or support lines, or from our
Finally, cost is an important factor. Much of the
software described below is centrally funded at UT. That means it is
free as long as you are using it for internal teaching and research. SAS and
Stata are not centrally funded so you will have to pay to use them. And, for
administrative use, one copy of SAS is several hundred dollars. Using any
commercially available software for the benefit of outside organizations, even
non-profits or any government agency, requires a commercial license. Those
often cost between $500 and $15,000. Such fees make open source software such
as R or ImageJ much more attractive.
We support software that is popular within the UT
research community, that solves problems well, that runs on as many operating
systems as possible, and, hopefully, whose market share is growing rather than
shrinking. The field of research software is rapidly changing and so we
frequently evaluate new software and monitor the popularity of existing
software. If you would like us to consider licensing and/or supporting a new
package, please contact the OIT HelpDesk.
We offer two levels of support: Full and Minimal. For both levels, we make the software available to the UT
community by arranging site licenses or server installations. We also keep the
software running and up to date. For the 50+ packages we support, this requires
a substantial amount of work.
that we recommend the software, and we support both the software and its
related research methods. We test new releases thoroughly, teach its use
through workshops or one-on-one tutorials, and usually have more than one
knowledgeable consultant available to assist you with it. OIT is a
customer-driven organization, so fully supported software is typically in use
by at least 500 people in a broad range of departments.
Support means we can only assist you with installing it on your
computer or starting it on our computers; we can point you towards tutorials
and documentation, and we may be able to help with importing or exporting data.
Software for which we offer only Minimal
Support is still of high quality; however, our resources do not
currently allow us to provide Full
Support for them. Minimally
supported software is usually in use by fewer than 500 people on campus, or
its use is focused in a single department or college.
We offer most UT students, faculty, and staff of UT's
Knoxville campus up to 10 free hours of assistance per semester for their
research computing needs. Assistance is available via appointment and walk-in
service. Appointments during busy times of the semester may take over a week to
get, so please plan ahead. For more details see https://oit.utk.edu/research. We also
offer training workshops each semester. See https://oit.utk.edu/training or
call the HelpDesk at 974-9900.
Bioinformatics software is used to study the
information stored in DNA. Although we only offer Minimal Support for this area, we
are studying this area closely to determine if our support level should
increase in the future.
As the product web site states,
"Whether you’re working with large data sets from next-gen sequencing
studies or microarrays, JMP Genomics provides the tools you need to analyze
rare and common variants, detect differential expression patterns, discover
reliable biomarker profiles, and incorporate pathway information into your
analysis workflows. In addition, our latest release extends the software’s
capabilities for creating and manipulating genetic linkage maps, and then
utilizing these maps in downstream QTL mapping for important agronomic
JMP Genomics is available as a free download for your
Windows computer and also runs on Apps@UT. We
offer Minimal Support for
* Study and obtain summary statistics on
genetic markers more easily.
* Examine the relationship between two or
more markers by combining genotype data with well-recognized methods.
* Find associations between markers and
traits such as disease using case-control or family data more quickly. The
flexibility and scalability of the application means any number of alleles can
* Correct any inconsistencies caused by
multiple testing effects on results, thus improving the quality of results.
* Interpret results more efficiently since
disequilibrium and association data are displayed simultaneously in an
For more information, see http://support.sas.com/documentation/onlinedoc/genetics/.
Acquisition & Web Surveys
Data acquisition software and hardware is used to
collect data quickly while minimizing errors.
National Instruments is a graphical programming language that uses icons
instead of lines of text to control laboratory instruments. It can help you
with the creation of tests and measurements, data acquisition,
instrument control, data logging, measurement analysis, and report generation
applications. Our license includes use only on UT-owned machines.
We offer Full Support for LabVIEW. We
also offer a basic training workshop on it each semester. See http://oit.utk.edu/training for
details. A LabVIEW CD-Based Training (CBT) CD is available from http://oit.utk.edu/software and UT Bookstore Technology Center. LabVIEW manuals
are available in Hodges Library Reserve.
data acquisition capabilities for computer mediated communication,
including some social media sites and YouTube, and we offer our Full Support for NVivo. For
details, see NVivo below in the Qualitative Analysis section.
Qualtrics is a
web survey design tool that will be available at the end of Spring Semester,
2013. At that point, we will begin Fully Supporting it. This software makes it easy for anyone to
create a web-based survey, and it has an extensive feature set. For
announcements, follow the OIT News (News and Events link on left of this
Collection Web Interviews from IBM is a web survey design tool
we Fully Support. This software
is easy to use, and the server that we use to collect the data is backed up
nightly. You can get an account to use this server by making a half-hour
appointment to get trained in the basics. You can also take one of our training
workshops that we offer each semester. You can sign up for them at https://oit.utk.edu/training/. For more
details, see https://oit.utk.edu/research/websurveys/. We will
be encouraging people to use the newer Qualtrics solution (above) starting
Summer Semester, 2013.
Data mining software automatically searches for useful
patterns in your data. The three main types of models used in mining are
statistical models, decision trees, and neural networks. For ease of use, most
data mining programs are controlled by drawing a diagram that represents the
flow of data through various analyses.
Miner is a
comprehensive data mining package from the SAS Institute that features all the
popular methods for both numeric and text data. We offer Full Support for its use. You
control it by drawing a diagram which represents the flow of data and results.
To learn it, choose Getting Started with Enterprise Miner Software from the
Help menu, or stop by for a demonstration by an RCS consultant. The
manuals are online at http://support.sas.com/documentation/onlinedoc/miner/.
You can read more about the product's features at http://www.sas.com/products/miner/.
Togaware is a free data mining user interface to R. It is easy to learn by
reading the documentation at http://rattle.togaware.com/.
called Clementine) is a comprehensive data mining package from SPSS, an IBM
company. It is started as a separate program, but it can read SPSS datasets, as
well as import data from many other sources. Due to its low usage, we currently
offer Minimal Support for
To learn the basics of Modeler, start the package and
choose Tutorial from the Help menu. The manuals are included in the
installation files. Modeler is the only product from the
SPSS company that is limited to installation on computers owned by
the University. It is available on Apps@UT. You can
read more about it at http://www-01.ibm.com/software/analytics/spss/products/modeler/.
Data graphics visually displays information about
measured quantities. It helps you discover relationships and conveys
information to others. Data Visualization is the subset of data graphics that
deals with interactive, dynamically changing data displays that are used to
explore data. Scientific Visualization is a type of data graphics that is used
to visually analyze data from real objects such as a tornado, a brain, or an
airplane. Statistical models often do not exist to analyze the latter type of
data. The displays often use color to display a fourth dimension, such as
velocity, tissue type or stress. Animation (or freeze-frame) is often
used to show changes through time.
data graphics and visualization capabilities, and we offer our Full Support for ATLAS.ti. For
details, see ATLAS.ti below in the Qualitative Analysis section.
data graphics and visualization capabilities, and we offer our Full Support for them. For details, see
R below in the Scientific, Engineering, and Mathematics section.
data graphics and visualization capabilities, and we offer our Full Support for NVivo. For
details, see NVivo below in the Qualitative Analysis section.
Origin from OriginLab is
a very powerful data graphics package. Its ability to transform data is
supported by the entire set of routines developed by the Numerical Analysis
Group (NAG). The package is available on Apps@UT where two
(2) people can run it at the same time. A series of tutorials is available on
the Help menu. Online multimedia tutorials are available for viewing or
downloading at www.originlab.com. We
offer Minimal Support for
data graphics and visualization capabilities and we offer our Full Support for them. For details, see
R below in the Statistics section.
a data visualization package. It can display a rotating 3D scatter plot of, for
example, blood pressure, cholesterol level, and salt intake. If you then
display a bar chart of gender, clicking on the bar for males will cause the
males in the scatter plot to become highlighted, so you can quickly see if
males differ from females. All plots are interactive and linked in a similar
manner. Stat Studio comes as a part of the SAS System. To learn Stat
Studio, read its manual at http://support.sas.com/documentation/onlinedoc/. We offer
Minimal Support for SAS/IML
SAS/Graph is a
set of graphic procedures that are built into the SAS system. This product was
greatly improved in SAS 9.2 and now offers superb quality graphics that are
easy to program. The manuals are online at http://support.sas.com/documentation/onlinedoc/ . We
offer Full Support for this
SigmaPlot from Systat Software
is particularly good at scientific graphics. We offer Minimal Support for it.
a data graphics package for which we offer Full
Support. It is extremely flexible, and SPSS Inc. claims it can create every
type of data graph. Its graphical user interface can create a reasonable range
of graphs, but its full power resides in its Graphics Production Language.
Since it is built into the popular SPSS package, it offers a set of popular
graphs integrated into most analyses. Even if you are not interested in doing
statistics, this is a good choice for data graphics. To learn how to use it,
take one of our training workshops or make an appointment for a quick
one-on-one tutorial. The tutorial listed under the SPSS Help menu includes a
brief overview of SPSS Graphics. The SPSS Base User's Guide describes its
use in more detail. The SPSS manuals are a part of the SPSS Suite.
Stata SE from StataCorp is
quite good at 2D data graphics. For details, see Stata under Statistics below.
We offer Minimal Support for
You can analyze images both quantitatively and qualitatively.
To analyze video, you can capture key frames and then apply image analysis to
Quantitative image analysis extracts measurements from
images, such as photographs, microscope images, and satellite imagery. The
software can count things such as cells, insects, or any type of spots. You can
use it to measure the length of items in any units you specify and angles
in degrees. You can also calculate area in square units or percent
coverage. The most popular package for this type of analysis is ImageJ,
which is described below.
Qualitative image analysis helps you classify images or
parts of images. For example, a psychologist might study photographs of people
interacting and manually select parts of the image that contain various types
of body language. Art historians might classify various types of symbolism used
by different artists and then cross tabulate their findings to compare the
artists or track their changes through time. Qualitative analysis
packages that assist in image analysis, including ATLAS.ti and NVivo, are
described in the Qualitative Analysis section.
Image J is a
free, open source image analysis program for Windows or Macintosh. It is easy
to learn, and it analyzes images quickly. We offer Full Support for it. You can get started with it by taking our
hands-on workshop towards the beginning of each semester. The manual that comes
with it is also a good way to learn it. It is installed on the Apps@UT server and is available for free download
at http://rsb.info.nih.gov/ij/ .
& Geographic Information Systems
Mapping software ranges from tools that display a
static statistical theme, such as population by country (thematic maps), to
full-blown Geographic Information Systems (GIS) that allow you to input,
update, query, analyze, and display geographic information. If you need full
GIS capabilities, we recommend taking a course in the Geography Department to
learn the ArcGIS products from
the ESRI Company.
an integrated suite of packages for Geographic Information Systems and Spatial
Analysis. It is available in most of OIT's computer labs. We currently
offer Minimal Support of
the ESRI product line.
free and open source software that does data mapping and advanced geospatial
analysis. We currently offer Minimal
Support mapping for R. For more about R, see it below in the Statistics
SAS/GIS is a
powerful GIS package, but its usage on campus is so low that we offer Minimal Support for its use.
Documentation is available for it at http://support.sas.com/documentation/onlinedoc/ .
thematic maps from within the SAS system. We offer Minimal Support for mapping in SAS/Graph.
Qualitative analysis programs help you manage and
analyze text or multimedia data through the manually assignment of codes or annotations.
They also may contain tools to build theories about relationships among
concepts. In certain software packages, numerical code counts or indicators can
be exported to statistical packages, if desired. This type of software is also
known as Computer Assisted Qualitative Data Analysis Software, or CAQDAS.
A helpful review of this software is located at: http://www.restore.ac.uk/lboro/research/software/caqdas_comparison.php.
ATLAS.ti Scientific Software Development GmbH, is a qualitative analysis
package that allows the researcher to organize, analyze, and visualize data in
both text and multimedia formats.
ATLAS.ti also allows users to transcribe directly into the program and
has an accompanying iPad app for multimedia data collection. ATLAS.ti is available for download by
UT students, faculty, and staff at the OIT software distribution page at https://web.dii.utk.edu/softwaredistribution/. It is also available on the Apps@UT server where 30 people can run it
simultaneously. We offer Full
Support for this product and hold several workshops each semester on
QDA Miner is a
program from Provalis Research that supports qualitative analysis of
text and images. It also links to its sister text mining program,
WordStat. See the Text Mining section for a description of that. The two
provide a unique combination of manual and automated features. For example, you
might manually select various blocks of text and then pass only those on for
QDA Miner is available for download by UT
students, faculty, and staff at the OIT software distribution page at https://web.dii.utk.edu/softwaredistribution/. It is also available on the Apps@UT server where 50 people can run it
simultaneously. We offer Full
Support for this product. The best way to get started is to watch the
tutorials or read the manual located at http://provalisresearch.com. The help
files are also quite good. OIT teaches a hands-on workshop on QDA Miner each
spring and fall semester.
QSR International, is a qualitative analysis package that supports both
qualitative and mixed methods research. It
lets you collect, organize, and analyze content from textual, multimedia, and
computer mediated communication data. It
also offers ways to visualize data, including cluster analysis. It is
available on the Apps@UT server
where 20 people can run it simultaneously. We offer Full Support for this product and hold several workshops each
semester on NVivo.
Engineering and Mathematics
This category includes a variety of tools of interest
to scientists and engineers. See also the section on Graphics &
Education Master Suite consists of engineering, design, and
modeling software developed by Autodesk, Inc. and is available for free to
faculty and staff at UTK and UTSI for installation on University-owned computers.
The Suite includes more than two dozen individual applications, including
AutoCAD for 2D and 3D design and drafting. The OIT Autodesk/AutoCAD
installation document includes a list of all the applications. You can browse
the Autodesk online manuals at http://usa.autodesk.com/support/documentation. In
addition, the Autodesk Education Community at http://www.autodesk.com/edcommunity offers
students and faculty free downloads of Autodesk 3D design software, including
AutoCAD, Maya, Revit and Inventor. This free software cannot be installed
or used on institutional computers located in classrooms and laboratories, but it
may be installed and used on the personal computer of the Education Community
Microsoft is very flexible, and we offer Full
Support for its use on research problems. The only thing we do not
recommend using Excel for is statistical analysis. There is, however, an Excel
add-in for R that we do recommend for statistics. Excel is also useful as a
data entry tool for research data as long as you know the proper way to
structure your spreadsheet.
Excel is available for your PC or Macintosh and on
those machines in our computer labs. To learn how to use it, sign up for
training at https://workshop.utk.edu/.
one of the oldest programming languages, and it is still widely used in
research. As its name implies, it is best at mathematical FORmula TRANslation. We
offer Full Support for this language,
which is available for your computer and on our Linux and UNIX servers (not in
our computer labs).
Maple, from Maplesoft, is
a symbolic math package, because it can work with formulas rather than just
numbers. It lets you enter a problem in the form of a mathematical formula and
returns answers that can be a number or another formula. It can perform tasks, such
as factoring equations or taking derivatives or integrals. For more details, see
http://www.maplesoft.com/. It is
available for your computer running Windows, Macintosh, or Linux, and it is
installed on the Windows computers in our labs. We offer Full Support for this package.
There are several ways to learn Maple. We offer a Maple
Basics training workshop that we teach each semester. You can sign up https://workshop.utk.edu/. For a
general overview of Maple, within a Maple session, select Help>>New
Users>> and then select either Quick Tour or Full Tour. Online Maple
tutorials can be downloaded from the Maple Application Center at www.mapleapps.com/tutorial.shtml.
Wolfram Research, is a symbolic math package, because it can work with formulas
rather than just numbers. It lets you enter a problem in the form of a
mathematical formula and returns answers that can be a number or another
formula. It can perform tasks such as factoring equations or taking derivatives
or integrals. For more details, see http://www.wolfram.com.
Mathematica is available on the Apps@UT where 7
people can use it at once. We offer Minimal
Support for it since we have a site license for Maple, a very similar
product. To learn Mathematica, see the Tutorial that is available when you
start it under the Help menu.
The MathWorks, is a MATrix LABoratory package that
excels at solving problems in matrix algebra. It is the most widely used
package in Engineering and is popular in many other scientific fields. For more
details see http://www.mathworks.com or OIT's MATLAB information page. You
can install MATLAB on your Windows, Mac, or Linux computer at the OIT software
distribution page at https://web.dii.utk.edu/softwaredistribution/. To learn
MATLAB, you can use the tutorial that is available under Help > MATLAB Help
> Getting Started, or you can sign up for our MATLAB Basics workshop, which
is taught every semester. You can sign up at https://workshop.utk.edu/. We offer
Full Support for MATLAB.
PuTTY is a
free and open source program Telnet, SSH, and xterm client for Windows and
UNIX. For the software documentation to help you learn it, see the project web
Notebook from MacKichan Software
is a word processing package that specializes in handling mathematical and
statistical formulas. It can save its output in the TEX format that is required
by many journals that make heavy use of formulas. It is available on Apps@UT where ten people can run it at the
same time. We offer Minimal Support for it. For a quick tour of the Scientific
Notebook features, within a Scientific Notebook session, select Help >
Contents > Take a Tour. For a more comprehensive tutorial, select Help
> Contents > Learn the Basics.
Xming allows your
PC to display X-Windows output from other computers, especially those running
Linux or UNIX. This allows you to control our Linux or UNIX server computers
using a graphical user interface similar to Windows or Macintosh. Most of our
UNIX software provides graphics output only in an X-Window environment. We
offer Full Support
for Xming. For documentation on how to configure Xming, see https://oit.utk.edu/research/documentation/Documents/HowtoInstallXmingForWindows.pdf. It is available for your computer
from http://www.straightrunning.com/XmingNotes/ and
in our computer labs.
Statistics packages are perhaps the most widely used
type of research software. Over half of all theses and dissertations at UT
require the use of statistics. Stat packages can summarize data into numbers,
such as percents or means, and optionally test whether groups differ on such
measures. It can also measure the strengths of relationships among measurements
and make predictions from them, as well. The packages for which we currently
offer Full Support in this area are
the SAS System and SPSS. Those products are covered on the following pages.
Statistics packages for which we offer Minimal Support are JMP, R, S-PLUS, and
A package that we do not recommend for statistical
analysis is Microsoft Excel. It is good for data entry, creating new variables,
manipulating data, and creating graphics. However, we do not recommend its use
for statistical analysis. The flexibility that it offers makes it easy to do
analyses that make little sense, its range of analysis methods is limited, and
its statistical algorithms have problems, especially in the presence of missing
values. Its ability to sort columns independent of one another is a
particularly dangerous feature that essentially destroys a dataset. There is
however an Excel add-in that allows it to do statistics with R. To see how best
to use Excel for data entry, see https://oit.utk.edu/research/documentation/Documents/HowToUseExcelForDataEntry.pdf.
of Mment Structures) from SPSS analyzes Structural Equations
Models (SEM) by either drawing a path diagram or via a command language. These
models are also known as LISREL models, confirmatory factor
analysis, analysis of covariance structures, path analysis, or causal
modeling. The web page is www.smallwaters.com,
and SEM information is at www.gsu.edu/~mkteer/semnet.html.
We offer Full
Support for Amos. To learn Amos, we recommend working through the examples
of the excellent Amos User’s Guide, which is on reserve at Hodges Library. A
Programmer's Reference is available in the \Program Files\AMOS 5\Documentation
folder. Amos is available for your PC and in our computer labs. It is part of
the SPSS Suite of software. To help Macintosh and Linux users access it, it is
also available on Apps@UT where
5 people can run it at once.
Multivariate Software analyzes Structural Equations Models (SEM). These models
are also known as LISREL models, confirmatory factor
analysis, analysis of covariance structures, path analysis, or causal
modeling. It handles categorical variables better than Amos currently does. The
web page is http://www.mvsoft.com ,
and SEM information is at www.gsu.edu/~mkteer/semnet.html. To
learn EQS, see the User's Guide available on the Help menu. We offer Minimal Support of EQS. It is
available on Apps@UT, where 2
people can run it at once.
Scientific Software International is a package that does Multilevel Models that
are also often called Hierarchical Linear Models (HLM). An example of a
multilevel model is a regression equation showing how studying increases
grades. That relationship may differ at different levels, such as teachers
within schools within districts. Although we offer Full Support for the solution of multilevel models using SPSS and
SAS, we currently offer only Minimal
Support for the HLM program. You can learn HLM from the
Help files, which contain the entire manual. However, to understand the manual,
you also need the book by the program's
developers, Raudenbush and Bryk, Hierarchical Liner Models,
Applications and Data Analysis Methods. HLM is available on Apps@UT where 2 people at a time can run
SAS Institute is a statistics and graphics package that is easy to learn. We
offer Minimal Support for JMP, because
it is not widely used at UT for theses and dissertations, but it is a
very good package. It has a range of statistical procedures that is adequate
for most researchers, and its design of experiments ability is top notch.
Running an analysis is as easy as selecting variables and options in dialog
boxes. However, SAS and SPSS offer much more extensive selections of
procedures. JMP's main strength is its interactive visualization
capabilities which make data exploration quick and easy. However, since the
displays are optimized for exploration, they are not publication quality. JMP
also offers the best Design of Experiments capability.
The best way to learn JMP is to read the Introductory
Guide which is available in the JMP folder and under the Start menu in Windows.
The product web page is www.jmp.com. JMP is
available for your own computer and in our computer labs on both Windows and
Macs. Although a LINUX version is available, we do not yet have it due to low
Scientific Software International analyzes Structural Equations Models (SEM).
Variations of these models include confirmatory factor analysis, analysis of
covariance structures, path analysis, or causal modeling. LISREL is
viewed by many as the most powerful and most difficult to learn of the SEM
packages we have available. It is available on Apps@UT where 2 people can run it at once. We
offer Minimal Support for its use.
The manual is available in the folder \Program Files\lisrel854\IBOOK. The web
page is http://www.ssicentral.com/lisrel/mainlis.htm, and SEM
information is at www.gsu.edu/~mkteer/semnet.html.
Mplus from Muthén & Muthén analyzes
Structural Equations Models (SEM) that can include both continuous and
categorical latent constructs. As their web site states, it can also analyze
"both cross-sectional and longitudinal data, single-level and multilevel
data, and data that come from different populations with either observed or
unobserved heterogeneity. Analyses can be carried out for observed variables
that are continuous, censored, binary, ordered categorical (ordinal), unordered
categorical (nominal), counts, or combinations of these variable types."
The web page is http://www.statmodel.com/features.shtml,
and SEM information is at www.gsu.edu/~mkteer/semnet.html. To
learn Mplus, we recommend the program's help files. We offer Minimal Support for this
package. Mplus is only available on Apps@UT where 1 person can run it at a
free, open source software for data analysis and graphics for which we offer Full Support. There are free versions
available to run on many different types of computers at http://www.r-project.org/. R is
available for your Windows, Mac, or Linux computer and on our Apps@UT server and our high performance Newton Linux Cluster.
an SPSS-like graphical user interface for R available at http://socserv.mcmaster.ca/jfox/Misc/Rcmdr/. See also
Rattle. If you already know SAS or SPSS, you can learn R by reading the free
version of R for SAS and SPSS Users at http://r4stats.com.
you plan your study by figuring out the smallest number of subjects (or
experimental units) you would need to find a significant result. To learn this
package, read the SamplePower manual on reserve at Hodges Library. SamplePower
is available for your PC and is on the PCs in our computer labs. It is part of
the SPSS Suite of software. We offer Minimal
Support for its use.
SAS from SAS
Institute is a powerful statistics package, and we offer Full Support for its use. It contains multiple menu systems,
multiple programming languages, and pre-written solutions to most commonly seen
research problems. The package comes with over 70 manuals! Luckily, most
people do not need more than a few of these manuals for their research needs.
SAS used to stand for Statistical Analysis System, but it grew far beyond
statistics into other areas, such as applications development and data
warehousing. So, the SAS Institute now says that SAS doesn't stand for
anything. SAS is available for your own computer, in the Humanities computing
lab, on Apps@UT, and on
our high performance Newton
You can control SAS many different ways. The most
widely used method is the SAS programming language. You learn commands and type
them in to control your analysis. Enterprise Guide is the main menu-based
interface to SAS that is easy to use. SAS Stat Studio is another menu system that
has extensive data visualization capabilities. SAS is available for Windows,
Linux and most UNIX computers.
The best way to learn SAS is to take one of our
training workshops that we offer each semester. You can sign up for them
at https://workshop.utk.edu/. You can
also make an appointment with one of our consultants to get an overview of SAS
and get started using your data. There is a tutorial available on the Help menu
in Enterprise Guide. There is also computer based training called SAS/TUTOR,
but it focuses on learning the SAS language rather than its analytic
procedures. It is available at http://web.utk.edu/sas/basic.html and at http://web.utk.edu/sas/advanced.html.
The SAS manuals are also quite complex and not a good place
to begin learning SAS. They are, however, an excellent reference, and they are
available at http://support.sas.com/documentation/onlinedoc/.
SPSS, an IBM company, is by far the most widely used statistics package at UT,
and we offer Full Support for it. It
does a wide range of statistics and graphics, and it is easy to learn and use.
It is also easy to transfer data from SPSS into SAS, so that any esoteric
analysis can then be run in SAS. SPSS also has a programming language that you
can use to automate repetitive tasks, to leave an audit trail of the steps you
took, and to get the package to do something that it was not originally
designed to do. It is available for your computer, in our computer labs (both
PCs and Macs), and on Apps@UT.
The best way to learn SPSS is to take one of our
training workshops that we offer each semester. You can sign up for them at https://workshop.utk.edu/. You can
stop by our walk-in support area on the fifth floor of Greve Hall for an
overview by one of our consultants. There is also an introductory tutorial
available in SPSS under Help. The other help files are also very easy to learn
Stata SE from StataCorp is
a general purpose statistics package that is designed to be easily extendable.
There are many hundreds of extensions available for it in Internet
repositories, all free of charge. Stata is particularly strong in time series,
panel data, and the analysis of complex, non-random samples. To learn
Stata, see the program's help files or see the manuals on reserve at Hodges
Library. Stata is available on Apps@UT where
14 people can run it at the same time. You can read more about it or order it
at http://www.stata.com/. For
pricing, it is important to know that UT has "GradPlan" educational
pricing. Be sure to select that option when placing your order. We offer Minimal Support for Stata. Don
Bruce in CBER is the GradPlan representative on
campus. Feel free to contact him with your questions at firstname.lastname@example.org or
974-6088. To get started with Stata you can watch the tutorials at http://youtube.com/user/StataCorp.
SUDAAN, from RTI International,
is SUrvey DAta ANalysis software that can analyze data from
complex samples, such as cluster or stratified designs. For example, making a
random selection of high schools from which you randomly
select students, is no longer a simple random sample. As a result,
the standard procedures from SAS and SPSS do not give the proper results. For
this type of data, we offer Full Support
for the use of SAS' specialize procedures (e.g. PROC SURVEYREG), but
provide Minimal Support
SUDAAN is available on Apps@UT where 3 people can run it at the same
time. To learn SUDAAN, read the User’s Manual which is on reserve in
Hodges Library and is in the server folder N:\Program
Files\SUDAAN\Release900\Manual. You can download example programs from http://www.rti.org/sudaan/. Click on
About SUDAAN and then Examples. Be sure to change the location of the
data file for any example program before running it.
Various methods of text analysis include qualitative
analysis, content analysis, latent semantic analysis, linguistic analysis. The
latter three methods are also referred to as text mining.
Qualitative analysis programs help you manually study
text and are discussed in the section above on Qualitative Analysis.
Content Analysis software uses a
"dictionary" of words and phrases to automatically classify
documents. The dictionaries may be standardized collections of categories used
by many researchers or may be custom dictionaries created for a single research
project. Standardized dictionaries classify words into linguistic categories,
such as nouns and verbs, or psychological categories, such as personality
types. There is even a scale which attempts to determine when someone is
telling the truth or lying. Researchers use custom dictionaries to study areas,
such as open-ended survey questions, interviews, and speeches. Tools, such as
cluster analysis and multidimensional scaling, can assist in creating custom
dictionaries. WordStat is the software we support that does content
Latent Semantic Analysis (LSA) automatically extracts
concepts contained in text documents. In simplified statistical terms, it
factor analyzes the text to extract concepts and then clusters the documents
into similar categories based on the factor scores. This might be used to
analyze interviews, open-ended questions on surveys, collections of email, or
to study the literary style of authors. It is most useful when analyzing
hundreds or thousands of documents, but it can be applied to smaller numbers of
short documents if they describe a small number of concepts. SAS Text Miner
Linguistic analysis applies an understanding of the
rules of language to extracting concepts and/or summarizing documents. A major
part of this is extracting terms or phrases and comparing them to large
databases of pre-categorized terms or phrases. SPSS Text Analysis for Surveys
uses the linguistic approach.
Text Mining is a loosely defined term that applies to any
of the automated methods above to large volumes of text.
Due to the complexity of importing data into text
analysis packages, our Minimal
Support level does not include importing data. However, SAS Text Miner
has by far the most extensive ability to import and export documents, making it
possible to get text from many sources into some of the other packages listed
Provalis Research, uses the content analysis approach to help you use or create
dictionaries of categories that you can then quickly apply to large numbers of
documents. Since you start WordStat from within QDA Miner,
you can apply the content analytic approach to all your documents or just the
sections that you manually coded.
Simstat is a
basic statistics package that can apply quantitative analyses to the output
from either WordStat or WDA Miner.
The standardized dictionaries installed
with WordStat are: WordNet 2.0, the Regressive Imagery
Dictionary (RID), Roget's Thesaurus, and the Forest Value Dictionary. For
details about the dictionaries, go to http://provalisresearch.com/products/content-analysis-software/ and click on Dictionaries.
These packages are available for download for free at http://oit.utk.edu/software. They
are also available on Apps@UTwhere 50
people can run them simultaneously. That server includes the dictionary for
Linguistic Inquiry and Word Count (LIWC). We offer Full Support for these products. The best way to get started
with these packages is to watch the tutorials or read the manual located
at http://provalisresearch.com. The help
files are also quite good. OIT teaches a hands-on workshop on QDA Miner and
Wordstat each spring and fall semester.
Analysis for Surveys is a linguistic package designed
to help you categorize open-ended survey items. Its ability to differentiate
positive from negative responses is quite helpful. It is limited to around
5,000 responses of a page or less.
This package is easy to learn using the manual that is
included with it. It is available for use on your Windows computer at no
charge. We offer Minimal Support for
this package. You can learn more about it at http://www-03.ibm.com/software/products/us/en/spss-text-analytics-surveys/.
the latent semantic analysis approach to help you automatically discover
underlying themes, cluster the documents into homogeneous groups, and/or
classify new documents into categories. It can read text from many different
formats such as Excel, PowerPoint, RTF, Text, Word, WordPerfect, PDF, and even
entire web sites. To learn it, see Introduction to Text Mining in the Enterprise
Miner help files. To read more about it, see http://www.sas.com/products/textminer/. It is
available for your Windows computer via a site license and in our computer
labs. It is the only part of SAS that runs exclusively on Microsoft Windows. We
offer Minimal Support of
Knoxville, Tennessee 37996 | 865-974-1000
The flagship campus of the University of Tennessee System