Analyzing survey responses from virtual hands-on workshop
Apr 20, 2020 9:33 AM
| Last Modified: Apr 22, 2020 6:09 AM
We hosted the first JMP Virtual Hands-on Workshop in Europe recently. After the workshop, we surveyed the participants to find what they thought of it and how we could improve their experience. We thought we’d share what we learned with everyone, to show you how we do our analysis in JMP.
One response was nearly universal: Post a recording of the workshop so people can watch it whenever they have the time. Consider your wish granted. The link to the recording can be found here, along with all data sets needed to follow along.
Virtual hands-on workshop participants answer survey questions to help us better understand their needs and preferences.
Length and Time of Day
We wanted to know two things about time: what time we should hold future workshops and how long they should be. As can be seen from the figures below, there is a significant interest in having shorter workshops, while the time of day doesn’t seem to make a difference. With that in mind, we plan to offer two sessions for each upcoming workshop, one in the morning and one in the afternoon.
Most people prefer shorter workshops (a, left), but no significant preference for time of day (b, right).
Next we asked for the topics that users would like to see covered in future workshops. Rather than give respondents a finite set of topics, we allowed them to answer with their own words, which we put into a word cloud using Text Explorer, which you can see below. Three of the top four words are “data,” “analysis” and “studies.” While these are inherently good words, they are rather nebulous. The fourth word – DOE – gave us more to work with. About 40 percent of respondents made some reference to DOE or designed experiments, such as Definitive Screening Designs.
Most used words in “What topics would you like us to cover in future workshops” free-form response.
The remaining 60% requested a range of topics, which we divided into three clusters:
Statistical modeling and analytical techniques (e.g., neural nets, principal components analysis).
Automating analysis through scripting and dashboards.
Unique challenges specific to their companies’ situations.
With those requests in mind, we will schedule workshops on DOE, modeling and analytical tools, and scripting and dashboards. For those with specific challenges, we will be reaching out to them to see how we can support them on a one-to-one level.
The last question was also free form: Any other comments? We used Latent Semantic Analysis in Text Explorer to group the responses for us, which yielded meaningful results into three clusters (see below):
Problems playing the posted video (red)
Complaints about the sound during the event; we had some muting problems (blue)
Clustering terms from “Additional Comments or Suggested” free-form response.
In response to the technical issues, we will involve more JMP team members to help with the workshops: one person to deliver the workshop, another to answer questions via the chat, and a third solely responsible for the managing the presentation medium and ensuring attendees are muting upon entry.
Thanks so much to everyone who attended the workshop and then took the survey. We’re happy to report that in the two weeks following our workshop, JMP has rolled out JMP Discovered, an online series on a range of topics. New workshops will be scheduled as part of JMP Discovered so keep an eye out for additional information. In the meantime, stay safe!