SPSS - Great tool for broad range of usershttps://www.trustradius.com/predictive-analyticsIBM SPSSUnspecified8.31451012013-11-07T20:03:04.341Z
Updated November 07, 2013
SPSS - Great tool for broad range of users
Score 8 out of 101
- IBM SPSS Statistics
- Great tool for a broad range of users - One of my favorite aspects of SPSS is the low barrier for beginners. Unlike other statistical packages that rely heavily on syntax to run, this one is comparatively user friendly. Training interns on this is extremely easy as SPSS has point and click in addition to the syntax editor for running. With this version, many of the preliminary tests (ex. Levene's test) on the data are automatically executed when running analyses. Deep statistical training is not needed to on board a new user. This is so straightforward that even clients without any statistical background and be walked through simple tasks, and they can easily understand the output.
- Continuously improving user interface - This is an extension on the previous point. IBM SPSS is by far, much easier to use than other statistical packages. Because of the intuitive design, it takes no explanation to do simple tasks like creating formulas or editing variable metadata. When using SPSS as a teaching tool, you can jump directly onto the software without having to explain how to get to each function.
- Ability to run additional analyses not built or sold together - it might take some iterations for SPSS to include specific types of analyses, however, it is highly likely that current SPSS tools can be used in such a way to run those analyses. One example is HLM. By using linear mixed models in SPSS, it is possible to run HLM models with just a few additional steps in the set up. Another example of of analyses that can be performed in SPSS is conjoint (currently sold separately by IBM).
- The ability to manipulate (transpose or easily rearrange columns) makes this data set up very time consuming. More often than not, it is easier to manipulate data in Excel and then reopen the data in SPSS. That method is also a source of problems as Excel has column limits and occasionally different file versions of Excel do not port well into SPSS.
- The data visualization capabilities of SPSS are quite limiting. I often have to use a different tool (Excel or Tableau) to create graphics for clients.
- The output method is easy to read when running a few analyses. However, after running ten or so tests, the output tends to run together with no easy labeling of which output box goes with which test. Blocking off output boxes for each test or labeling each output box would make large sets of tests easier to read.
- With larger data sets, SPSS has a tendency to crash or freeze.
- The positive impact of SPSS is that it is very easy to train users. Efficiency is increased when easy tasks like creating crosstabs can be delegated to junior analysts.
- The negative impact of SPSS is primarily the usability of the output and lack of visually pleasing graphs. In order to create report decks, all the output needs to be ported to Excel, reformatted, and then turned into charts. On average, this task takes eight hours after the data is already cleaned for a 30-40 item instrument.
Despite the output and graphics shortcomings, the ease of use and intuitive design leaves SPSS as one of my top choices for statistical packages. Additionally, the easier use of the variable view (as compared to other packages) cuts down on the the need for multiple data files (variable code sheet + data file(s) + instrument file ) when running a project. It is a win-win when the client feels like he/she can understand how to read the output easily.
SPSS is like a multi-tool. It has a broad range of functions that can easily be used and understood by many people regardless of background. However, I don't recommend it for anyone doing large data sets (> 20,000 rows) or someone needing to use it for specialized analyses not included. Setting up the other analyses often includes manipulating the data file or many steps in the analysis process. SPSS also is not likely to be the correct tool for anyone working "big data." The attributes that make it easy to use would likely be too limiting for the broad approaches seen in big data. Also, I don't think SPSS would be able to handle that type of data size when it already has issues handling relatively small data sets.
I have not contacted IBM SPSS for support myself. However, our IT staff has for trying to get SPSS Text Analytics Module to work. The issue was never resolved, but I'm not sure if it was on the IT's end or on SPSS's end.
No - Instruction and support for SPSS is better from individual blog posts and websites not owned by IBM. Stats and education groups seem to have far more useful tips and tricks for SPSS.