IBM SPSS Statistics Review

Offers good duration, correlation, and nonparametric statistics


What is our primary use case?

I am using SPSS Statistics in many, many, many fields - more than you can imagine. The last one I was working on was with women at the Human Rights Council. I'm also using SPSS Amos for structural equation modeling. I'm using it to analyze data for business, data for engineering, and many other fields. Besides being a professional, I also sometimes advise or solve problems for others who are working with it in business or doing their Ph.D. or master's degree in it. Again, it is useful in many, many fields.

What is most valuable?

In terms of the features I've found most valuable, I'd say the duration, the correlation, and of course the nonparametric statistics. I use it for reliability and survival analysis, time series, regression models in different solutions, and different types of solutions. I'm using most of the content of SPSS.

What needs improvement?

I think the duration and time-series need improvement.

Some things need more explanation. For instance, if you are using the automated part in SPSS, you have automatic linear modeling. It gives you the meaning of the result, explaining if it is significant. But most of the package will give you the fixed value, or the p-value, without an explanation as to whether it significant or not. Some beginners might need not just the results, but also an explanation for them.

This already exists in two parts of the SPSS. One for the automatic linear modeling, and the second for nonparametric. The automation, the automatic part, gives you not only the results and the output but also what this output means. For me as a professional, I can explain the results even if it's not included in the output. But if they were to include some explanation for the results, such as is it significantly based on the P-value, that would make SPSS great.

Additionally, there is something else that really needs improvement - the part related to multivariate analysis. The chi-squared for instance. Some packages can give you the chi-squared directly, but in the multivariate part of SPSS, it's not as clear as it should be. So working with the multivariate could be improved or expanded. This last semester I was teaching multivariate to postgraduate students and it was not as clear as it should be. In conclusion, regarding improvements for the package, I can say that multivariate needs improving.

Lastly, some results need to be presented in a more diverse way. For instance, if you are working in biostatistics then sensitivity, specificity, and relative risk need to also be included. Some people don't use SPSS because they need some medical calculations and they think that SPSS cannot calculate it. But if you are working with the older, survival analysis, you can add this part to the medical field. Actually, it is really easy to include it and calculate it for it. As I told you, you can add a medical field part for the study design, such as screening tests, sensitivity, specificity and so on.

For how long have I used the solution?

I have been using IBM SPSS Statistics for more than 20 years.

I have written five books on using SPSS Statistics and one book for using the Modeler. Two of them were in English for bio-statistics, and the others are in other fields, but in Arabic.

What do I think about the stability of the solution?

I like the stability and would rate it at more than 95%. If you're working with the package for a long period of time, you understand most of the texts that you need to know.

I told you I wrote five books about using SPSS, so of course, I'm totally impressed with this package more than any other one.

When I was in the United States, I started to use SAS. But for the Middle East, SAS is not as easy as SPSS.

What do I think about the scalability of the solution?

I believe more than 50% of those in the university are working on it, so it's a long scale of users. For instance, tomorrow I am leading training for using the Modeler in a department in the university. All the time, a lot of people ask for a daily cost for using SPSS because it is the most common solution here in Saudi Arabia.

How are customer service and technical support?

I have never contacted support. I hope I can contact them with a call to improve my notes about the outputs or something.

How was the initial setup?

The initial setup is straightforward and doesn't take more than 20 minutes on average.

What about the implementation team?

I do the deployment myself. We can sometimes download from IBM's site. It doesn't need any help.

What's my experience with pricing, setup cost, and licensing?

We have a package for the university students and teachers. It's a package for graduate students, and the students can use it. They can download it and use it.

Which other solutions did I evaluate?

We evaluated other options. In general, in the Middle East, we use many different software packages, like SAS, Minitab, and SPSS. But I believe that SPSS might be the number one in the Middle East. I'm talking about Saudi Arabia or Egypt. 

What other advice do I have?

Of course I would recommend SPSS. I advise everyone to use it. As I mentioned, I wrote books about using SPSS. They are used for training and I knew most of the lecturers were using the package.

Time series is a branch in and of itself, and it's very important for forecasting the future. I was suffering until I could simplify this using the SPSS for forecasting or for building time series modeling.

On a scale of one to ten, I would give IBM SPSS Statistics at least a nine.

Which deployment model are you using for this solution?

On-premises
**Disclosure: I am a real user, and this review is based on my own experience and opinions.
More IBM SPSS Statistics reviews from users
...who work at a University
...who compared it with SAS Analytics
Add a Comment
Guest