Just about every two years StataCorp releases their latest version of their statistical software package, Stata. However, the newest release is sure to attract many more users to the tools that Stata offers to its users for several reasons.
The developers of StataCorp have been working hard to increase the intelligence of the program itself. For example, the – set memory – command will effectively be obsolete in the new release. The “Automatic Memory Management” component of Stata 12 is designed to detect the amount of RAM needed to analyze the data and will adjust the amount of RAM used to match this need. This is particularly helpful if you need to merge several datasets together, since the software will adjust memory allocation for you. They have also worked to redesign the user interface of the program. One of the best features of this is the new layout of the viewer (where you read help files, find user written packages, etc.). With a new table of contents drop down menu, you can get to the parts of the help file that you need and navigate through lengthy help files much easier than before. For example, try running – help estout – (estout can be installed from SSC if you do not have it already) and imagine how convenient it would be to have a drop down menu to navigate that help file.
There have also been a significant amount of modifications and additions made to the econometric and time-series models in Stata 12. ARFIMA, Multivariate GARCH, spectral density, and several other estimation features have been added to the new release. In addition to developing support for these time-series models, the developers have also created what they call “Time-Series Filters.” These filters allow you to have more flexibility, with greater ease, to specify how time functions within your model. For example, these new filters are able to split time-series data into cyclical and trend components. This can make estimating time-series models much more stable and precise.
A combination of new and improved visual routines adds to the capabilities to Stata 12. Now it is possible to produce contour plots that are often used to show results of geospatial analysis. Other new graphical components include enhanced ROC curve analysis and modeling, contrast plots, and margins plots.
However, the most impressive aspect of the new release, Stata 12, is the addition of Structural Equation Modeling (SEM) capabilities. If you’re like me, you use Stata all the time and then have to abandon your work there to use another statistical package (like LISREL, EQS, M-Plus, Amos, etc.) to fit SEM to your data. While it would have great to have the command line interface to run these models, Stata 12 now has a Graphical User Interface (GUI) to allow you to interactively create a graphical model that Stata can interpret and test for you.
Stata 12 is a great platform for allowing you to fit an amazing array of statistical models to your data. And even though the development efforts of StataCorp are making this software platform an amazing asset, it is also necessary to remind you that end-users of Stata’s products are also hard at work developing and refining statistical procedures that may be of interest to you. For example, Maurizio Pisati (a professor of sociology and social research in Italy) wrote a routine called – spmap – that can be used to show results of your analysis on a map. In my own research, I examined the effect of the availability of community-based resources on student achievement. Using the routine created by Dr. Pisati, I am able to display these results in a way that is ease to interpret for a wide variety of audiences.
About the Author
William Buchanan received his PhD at TUI University where his dissertation research focused on the causal effects of poverty on musical achievement in the US. He makes use of geospatial analysis (i.e., using ArcGIS software to measure resource allocation, concentration, and geographical differences in resource availability), econometrics (i.e., instrumental variable methods, regression discontinuity designs, differences-in-differences estimators, fixed-effects models, and others), and structural equation models (i.e., path analysis of observed data, latent variable modeling, etc…) to conduct quasi-experimental data analysis with a national data set from the US Department of Education. His use of multi-disciplinary approaches to research allowed him to integrate theory and statistical methodology from economics, education, and psychology. He has also provided statistical consulting and research services to a wide variety of public school districts, businesses, and institutions of higher education.
William Buchanan is also an aspiring Stata programmer and has helped his client to automate their data analysis by providing custom written Stata programs (i.e., .do and .ado). He also uses SPSS 16, LISREL 8.8, HLM 6.0, R, G*Power 3.2, ArcGIS, and other software packages to perform data analysis. Additionally, he uses StatTransfer 10 and is able to accept data files in almost any format and can provide data files in the format of your choice.
Specialties: music education, developmental psychology, neuropsychology, cognitive psychology, education, education policy, program evaluation, arts integration in education, econometric methods in educational research, contextual factors in education/developmental psychology, LaTeX typesetting software.
“He was patient and very thorough. I would highly recommend him to others in need of statistics help services. Received: statistical consulting, data analysis, interpretation of results, presentation of results.” Terrence Carter, Chief Learning Officer, AUSL
6. Describe your project: (e.g., book, business document, dissertation)
7. Describe the level of writing or editing required: (e.g., copyediting, proofreading, content editing, fact-checking, ghostwriting, formatting)
8. Current word count of document:
9. Your deadline date:
10. Required manual of style, if any: (e.g., Chicago Manual, APA, MLA, AP, AAA, CBE/CSE)
11. If you are a student, please provide the URL of your university's style manual:
12. Number of charts, tables, and pictures: ____________ Do you need them edited and/or formatted?
13. Do you have a budget for the project? (Please be specific.)
14. Number of footnotes and entries in reference list:
15. Do you want to contact a particular writer/editor?
16. How did you learn about our service?
Attach a sample chapter/section or other important documents related to your project. Please zip large files (max 1MB)
YOUR NAME MUST BE IN YOUR SAMPLE DOCUMENT OR IT CAN BE THE FILE NAME (e.g., johnsmith.doc).
A sample of the material is required to receive a quote for services. If you did not send a sample and/or description of the project with your first submission, please resend.
Once your e-mail is received, the network coordinator will forward it (plus any attached files) to the consultants you selected. If no selection is made, your submission will be forwarded to several consultants who might be a good match. Final choice of consultant is yours.
If you do not receive a response within 3 hours (M-F, 7a-7p Eastern), please RESEND your submission. You may also use the chat button or leave a message: 469-789-3030.
Allow a longer response time if you sent your submission during the weekend or after U.S. business hours.
All of the consultants listed on this site are freelance. They are located throughout the U.S. The coordinator cannot answer cost/timeframe questions for each consultant. You must go through the submission process to receive direct responses from the consultants listed on this site.
