| 27 February 2017 |
If you are an experienced SPSS user, the leap to using Displayr should not be a big one: Displayr is designed as an improvement over SPSS (I have more than a decade of experience using SPSS). This post describes the key benefits of using Displayr, showing why it is a much better alternative to SPSS.
Displayr is “live”
In SPSS, you regularly have to redo things. If you have to change the value labels for a variable, you need re-run any analyses that used that variable. To filter the data, you need to rerun your analyses. Weighting the data means you have to rerun the analysis. Etc.
In Displayr, all of this re-running happens automatically. Displayr remembers how all of the analyses in a document were created, and re-runs them automatically if you make any changes that necessitate the analyses being re-run.
Editing rather than pasting syntax
In SPSS, you can either work from the menus or work using syntax. If working from syntax in SPSS, a useful shortcut is to use a menu and then press the Paste button, which pre-writes the syntax for you. In Displayr, you can do something quite similar. Once you have used the menus to do something (e.g. run a regression or a machine learning algorithm), you can click on the output and choose Properties > R CODE to see the underlying R Code. You can edit this code live, and it will automatically update the analyses, a better alternative to SPSS. Check out Introduction to Displayr 5: Machine learning and multivariate statistics for an example.
In Displayr, the smart way to work is to avoid writing syntax
In the world of SPSS users there is a clear divide between users, and it relates to syntax. Newbies or a cowboys might do everything from the menus. If you are serious, you will spend a lot of time creating syntax files, as you know that these have two great advantages: they allow you to redo analyses with revised data, and they provide a record of everything that was done.
Experienced SPSS users might even think that this approach to using syntax reflects how one should work. But it doesn’t. Rather, it is a consequence of SPSS having been created 50 years ago. As I am sure you have noticed, the world of computing has changed a lot in the last half century.
Displayr automatically remembers everything you do. It does so in a really smart way. It does not keep a log of everything you have done. Instead, it works out all the steps needed to convert the original raw data file into your final analyses. This means that you can replace the data and, when you do so, Displayr will automatically redo all your work. If there are problems (e.g., due to missing variables), Displayr will identify all the outputs in your documents that you need to fix.
With Displayr, you can even get it to automatically update the entire report with live data feeds. For more information, see Introduction to Displayr 6: Reporting – automated and reproducible.
Our experience is that SPSS users take quite a while to really appreciate this fundamental change in workflow. The more experienced the user, the longer it takes to unlearn, so you should start now because it is a much better alternative!
You cannot break your data, and you can always figure out how things were computed
One of the reasons that SPSS users like syntax is that in SPSS you can easily “break” your data. For example, if you (or a colleague) recodes some data and saves the data file, then, unless you have diligently kept records, you run the risk of losing, or having to recreate, some of the data. In Displayr, by contrast, the underlying data never changes. You can always undo any data modifications, and you can always figure out how data has been modified. Again, making Displayr a better alternative to SPSS.
Merging by dragging and dropping
In SPSS, if you want to merge categories on a table, you typically do so by either recoding the underlying data, creating new variables, or by writing some code in custom tables.
In Displayr, if you want to merge categories on a table, you do so by dragging and dropping (see Introduction to Displayr 3: Creating tables, charts, and other visualizations). When you do this, Displayr makes the change to the underlying data, simultaneously updating all other analyses that use it. If you do not want this to happen, you can have multiple versions of the same data.
Variable type, measurement, multiple response sets, and variable sets
In SPSS, there are three different ways of encoding information about the structure of any variables. There is the variable type, measurement scale, and, the more exotic multiple response set. In Displayr, the same information is stored as the variable set structure. There are 14 different variable set structures at the moment. The table below shows how these relate to the SPSS concepts.
Displayr supports lots of different structures that contain multiple variables. This is particularly useful for survey research. See our wiki for definitions of the various structures.
|SPSS Type/Measurement/Set||Displayr Variable Set Structure|
|NA||Text - Multi|
|Multiple Response Set||Binary - Multi
Binary - Multi (Compact)
|NA||Nominal - Multi|
|NA||Numeric - Multi|
|NA||Binary - Grid|
|NA||Numeric - Grid|
Automatic selection of analyses based on variable set structure
In SPSS, the user has the job of determining the most appropriate type of analysis given their data. For example:
- To compute an average: Analyze > Descriptive Statistics > Frequencies.
- To compute a frequency table: Analyze > Descriptive Statistics > Frequencies.
- To compute a multiple response frequency table: Analyze > Multiple Response > Frequencies.
In Displayr, you instead just drag the data onto the page, and it automatically works out the appropriate tables based on the structure of the variable set. See Introduction to Displayr 3: Creating tables, charts, and other visualizations and Understanding Variable Sets in Displayr: A Tutorial. The time saved on this alone makes Displayr a better alternative to SPSS.
Multiple data sets
In SPSS, while you can analyze multiple data sets in a single session, you are essentially doing each in parallel. In Displayr, you can create a single document, which combines analyses from the different data sets. And, you can create analyses that use data from multiple data sets, setting relationships between the files (do do this, click on a data set in the Data tree, and press Edit relationship).
For a more general overview of Displayr’s benefits (and for more reasons that it is a great alternative to SPSS) see Introducing Displayr: the data science and reporting app for everyone.
Author: Tim Bock
Tim Bock is the founder of Displayr. Tim is a data scientist, who has consulted, published academic papers, and won awards, for problems/techniques as diverse as neural networks, mixture models, data fusion, market segmentation, IPO pricing, small sample research, and data visualization. He has conducted data science projects for numerous companies, including Pfizer, Coca Cola, ACNielsen, KFC, Weight Watchers, Unilever, and Nestle. He is also the founder of Q www.qresearchsoftware.com, a data science product designed for survey research, which is used by all the world’s seven largest market research consultancies. He studied econometrics, maths, and marketing, and has a University Medal and PhD from the University of New South Wales (Australia’s leading research university), where he was an adjunct member of staff for 15 years.