SPSS Statistics is a software package used for statistical analysis. It is now officially named "IBM SPSS Statistics". Companion products in the same family are used for survey authoring and deployment (IBM SPSS Data Collection), data mining (IBM SPSS Modeler), text analytics, and collaboration and deployment (batch and automated scoring services).
$99
per month per user
TensorFlow
Score 7.7 out of 10
N/A
TensorFlow is an open-source machine learning software library for numerical computation using data flow graphs. It was originally developed by Google.
I described earlier that the only scenarios where I use SPSS are those where we have legacy projects that were developed in the late 90s or early 2000s using SPSS, and for some reason, the project (data set, scope, etc.) hasn't changed in 24+ years. This counts for 1-2 out of around 80 projects that I run. Whenever possible, I actively have my team move away from SPSS, even when that process is painful.
TensorFlow is great for most deep learning purposes. This is especially true in two domains: 1. Computer vision: image classification, object detection and image generation via generative adversarial networks 2. Natural language processing: text classification and generation. The good community support often means that a lot of off-the-shelf models can be used to prove a concept or test an idea quickly. That, and Google's promotion of Colab means that ideas can be shared quite freely. Training, visualizing and debugging models is very easy in TensorFlow, compared to other platforms (especially the good old Caffe days). In terms of productionizing, it's a bit of a mixed bag. In our case, most of our feature building is performed via Apache Spark. This means having to convert Parquet (columnar optimized) files to a TensorFlow friendly format i.e., protobufs. The lack of good JVM bindings mean that our projects end up being a mix of Python and Scala. This makes it hard to reuse some of the tooling and support we wrote in Scala. This is where MXNet shines better (though its Scala API could do with more work).
SPSS has been around for quite a while and has amassed a large suite of functionality. One of its longest-running features is the ability to automate SPSS via scripting, AKA "syntax." There is a very large community of practice on the internet who can help newbies to quickly scale up their automation abilities with SPSS. And SPSS allows users to save syntax scripting directly from GUI wizards and configuration windows, which can be a real life-saver if one is not an experienced coder.
Many statistics package users are doing scientific research with an eye to publish reproducible results. SPSS allows you to save datasets and syntax scripting in a common format, facilitating attempts by peer reviewers and other researchers to quickly and easily attempt to reproduce your results. It's very portable!
SPSS has both legacy and modern visualization suites baked into the base software, giving users an easily mountable learning curve when it comes to outputting charts and graphs. It's very easy to start with a canned look and feel of an exported chart, and then you can tweak a saved copy to change just about everything, from colors, legends, and axis scaling, to orientation, labels, and grid lines. And when you've got a chart or graph set up the way you like, you can export it as an image file, or create a template syntax to apply to new visualizations going forward.
SPSS makes it easy for even beginner-level users to create statistical coding fields to support multidimensional analysis, ensuring that you never need to destructively modify your dataset.
In closing, SPSS's long and successful tenure ensures that just about any question a new user may have about it can be answered with a modicum of Google-fu. There are even several fully-fledged tutorial websites out there for newbie perusal.
collaboration - SPSS lacks collaboration features which makes it near impossible to collaborate with my team on analysis. We have to send files back and forth, which is tedious.
integration - I wish SPSS had integration capabilities with some of the other tools that I use (e.g., Airtable, Figma, etc.)
user interface - this could definitely be modernized. In my experience, the UI is clunky and feels dated, which can negatively impact my experience using the tool.
Theano is perhaps a bit faster and eats up less memory than TensorFlow on a given GPU, perhaps due to element-wise ops. Tensorflow wins for multi-GPU and “compilation” time.
Both money and time are essential for success in terms of return on investment for any kind of research based project work. Using a Likert-scale questionnaire is very easy for data entry and analysis using IBM SPSS. With the help of IBM SPSS, I found very fast and reliable data entry and data analysis for my research. Output from SPSS is very easy to interpret for data analysis and findings
Probably because I have been using it for so long that I have used all of the modules, or at least almost all of the modules, and the way SPSS works is second nature to me, like fish to swimming.
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
Community support for TensorFlow is great. There's a huge community that truly loves the platform and there are many examples of development in TensorFlow. Often, when a new good technique is published, there will be a TensorFlow implementation not long after. This makes it quick to ally the latest techniques from academia straight to production-grade systems. Tooling around TensorFlow is also good. TensorBoard has been such a useful tool, I can't imagine how hard it would be to debug a deep neural network gone wrong without TensorBoard.
Have a plan for managing the yearly upgrade cycle. Most users work in the desktop version, so there needs to be a mechanism for either pushing out new versions of the software or a key manager to deal with updated licensing keys. If you have a lot of users this needs to be planned for in advance.
I have used R when I didn't have access to SPSS. It takes me longer because I'm terrible at syntax but it is powerful and it can be enjoyable to only have to wrestle with syntax and not a difficult UI.
Keras is built on top of TensorFlow, but it is much simpler to use and more Python style friendly, so if you don't want to focus on too many details or control and not focus on some advanced features, Keras is one of the best options, but as far as if you want to dig into more, for sure TensorFlow is the right choice
I found SPSS easier to use than SAS as it's more intuitive to me.
The learning curve to use SPSS is less compared to SAS.
I used SAS, to a much lesser extent than SPSS. However, it seems that SAS may be more suitable for users who understand programming. With SPSS, users can perform many statistical tests without the need to know programming.
