We needed to give our work more guidance from our users. So while the team was already working on a visual design concept for the new SPSS Statistics, I was planning user interviews with participants across all personas. One of our discoveries after the interviews was that first-time users like Max needed to spend a considerable amount of time reading documentation and textbooks and viewing tutorials before getting started. Moreover, we have learned that as the attention of users gets worse, working with multiple open windows leads to visual overload.
MAR. 2017 | 5 MIN. READ
AUTHOR: Robin Auer, Design Research, IBM Cloud and Cognitive Software
We wanted to understand the user needs and pains in the current user interface. The design work focused on new users because we saw the biggest potential in the market when we can provide them with easy-to-digest user flow. Therefore their requirements were the reference point for the functionality of the new user interface. At the same time, we kept an eye on commercial experts (Current users) because they form the second big user base of SPSS Statistics. We knew that changes needed to be done thoughtfully and step-by-step. Big changes in the user interface like existing behaviors could confuse them and scare them away. It was essential for us not to lose our existing users. Better meeting the needs of our novice users—with our expert users always in mind—was crucial for the success of SPSS Statistics.
Based on this we selected participants for a row of user interviews. In the end, we talked to 13 different users in 23 remote interviews. Eight of these interviews were done with students (Max), five with power users (Angela), and ten with teachers (Robert). You can find more information about each persona in the project description from SPSS Statistics.
Here some facts about the participants: None of the Students, Power Users or Teachers used the Mac OS version from SPSS. So our User Tests were always prepared for windows operating systems. Nine participants told us that they have more than ten years of experience with SPSS. Moreover, two had less than two years of experience. Nine of our Sponsor Users also used SAS for statistical tasks. Six of them used R Statistics, and two used STATA or MPlus.
Before we talked with our users in 60-minute interviews, we sent them a survey with some basic questions about their environment and their existing user behavior in SPSS. This gave us an understanding of how they use the tool in their day-to-day work and which functionality was most important to them. The interview questions could then be very different in some cases. Sometimes the UX Designer needed insights to specific areas of SPSS to continue their work. Then I changed the question guide to getting answers from our users to these work specific items. Some of the later participants were even already confronted with new design concepts.
Through the initial interviews with our Sponsor Users, we wanted to get deeper insights into their work environment and user behavior in the current SPSS. The focus of one part of our interviews was on the usage of the three main windows Output View, Data Set (with Variable View) and Syntax Window.
Inspired from the tab concept of browser tools, we created the idea of tabs for the new SPSS. The tabs are the most important improvement in SPSS, and they are the foundation for the new Look & Feel. Five users told us in the interview that they like the tab concept. Also, two of them mentioned that this reminds them of the usability of browser tabs. Here is some of the user feedback:
If users want to change one or multiple variables in SPSS, they have to switch from the Data View to the Variable view. Very often they do this multiple times during their daily work.
The big idea is to make users life easier so they can edit variables in a side panel without changing the view from data view to variable view and being able to see both at the same time.
A student who uses SPSS Statistics
The users like the idea to have the variable view and the data view in the same window and the ability to edit both. However, one of the main needs is to copy variable properties from one variable to another and edit multiple variables at the same time. The users liked a clear overview of the current variable view table. If we want to get rid of the variable view, we have found a good combination out of both.
How does the user handle the huge amount of different dialog windows in SPSS? We asked them to lead us through a data analysis process and observed how they work with the various dialogs.
A lecturer who teaches SPSS Statistics
A good User Experience cannot work with 600 different dialog windows. Even the users didn’t mention that there is a pain point, it was clear for us that we have to find a better way for our users. We created three different concepts and confronted the users with them.
Concept A: Analysis in dialogs
The first concept is based on the old known way of dialogs which lead the user through the analysis process. The main difference was that the steps stay in a single dialog window and not in several windows.
Concept B: Analysis in a side panel
The second concept is based on the same idea as the side panels. While editing the settings for an analysis step, the user still can see the data set in the main window.
Concept C: Analysis in tabs
The idea of the third concept is, everything – a new analysis or transformation – will be opened in a new tab window. The users can run the analysis from there, and the output can appear in an existing or new output window.
To get a definitive answer about which analyze concept is the best for the users, we needed more user insights. This first evaluation with 13 different users was not enough to give a definite answer. The next step needed to be an A/B Test with more participants. Because I was leaving for Austin for the IBM Design Bootcamp, there was no time to run this follow up study.