North West University Mafikeng Data Carpentry workshop

This post originally appeared on the Data Carpentry website

Data Carpentry workshop, 1-3 November, 2017

Background/Introduction

Conducting the first Data Carpentry workshop by members of NWU Mafikeng campus and teaching my (Caroline Ajilogba) first workshop was a great task with support from NWU and our mentor, Anelda van der Walt. The support from the co-instructors Martin Dreyer (NWU Potchefstroom) and Amy Hodge (Stanford University), and from helper Bennett Kankuzi (NWU Mafikeng) was great. Other helpers from the R study group at Mafikeng’s NWU campus, like Olubukola Aremu and Ayansina Ayangbenro, were available to enhance the workshop.

This workshop was planned for 2.5 days, with spreadsheets, OpenRefine, and the Intro to R on the first day; the remainder of the R lesson on the second day; and SQL on the last half day.

The day before the workshop, Caroline and Bennett met to make sure nothing was left behind only to find out on the first day of the workshop that the venue had not been included on the workshop web site. Thanks to Martin and Amy who helped to salvage the situation and quickly updated the website, while Caroline sent e-mail to as many participants as she could who had sent messages to her requesting the venue.

Another interesting issue was the attendance register which was not available immediately but was also sorted before the workshop started on the first day. I must really say thank you to the organizers especially in making the connecting plugs, the stickies and also the badges available. Thank you so much.

The caterers were on time and participants were learning in a good atmosphere as some of them commented that it was a plus to come for a workshop and be taught and also given tea breaks and lunches.

Ist November 2017

The first day was great, as participants trooped in with enthusiasm and were being helped to settle down and have their software installed. Though the installation had its hitches here and there, I think that was great as we had to find solutions to how to handle the hitches and that for me was ‘learning’!!!

We waited nearly an hour for everyone to arrive the first day so that we could do data downloads and OpenRefine installation together. This took quite some time, as there were many issues with the install, including quite a few who had trouble with the required Java installation. Because of this and a later fire alarm and power outage, we were very short on time for teaching on the first day.

The instructors did a great job with the spreadsheet (Martin) and OpenRefine (Amy) lessons. Students later indicated they felt it was rushed, but when asked repeatedly if they had questions or wanted things repeated, they would say all was fine. We were supposed to cover the first part of R on the first day, but did not have time for that because of the delays.

Because of these issues, it seemed the point of the spreadsheet and OpenRefine lessons and their importance to the overall workflow and subsequent use of R seemed to get lost on the learners. The modules make sense in this order, as it is the same as someone’s own workflow might be, but seems sometimes difficult to connect for the learners. More emphasis should be placed on making these connections. For example, from the worksheet where the raw data is can be better organized, to using OpenRefine where more cleaning up is done, to loading it into R where data is analyzed and other databases are used.

Comments at the end of this day generally indicated that people were enjoying the workshop but that much of it was just too fast.

2nd November 2017

This morning was started with installation of R and RStudio. More of the students came with this software already installed, and many came early to do the installations, so that we were able to start right on time at 9am. Caroline started on a good note with R, but since we could not start off R on the first day, she thought she could be a bit fast after laying the foundation. When it was time to use data in R, she saw that for some it was fun, while others were trying to catch up. Since we planned to cover SQL on the third day, Caroline thought she had to go fast in order to cover everything in that one day. We also had technical difficulties on this day with the room technology, as the school’s desktop she was using kept shutting down and wiping everything so that she would have to reload RStudio, reinstall packages, create data frames, etc.

Again, comments on the second day indicated enthusiasm for R, but that it was going too fast. We never had anyone say that we were going too slowly. At this point, we still had the ggplot lesson to do and had decided that we would continue with R on the last half day and skip the SQL lesson altogether. Caroline did an excellent job responding to comments from day 2 that she slow down, and day 3 seemed to finally hit the right pace for these students.

3rd November 2017

The third day was better than I expected as participants were still ready to work, though some had told me they were travelling early Friday morning and were not available. The workshop concluded with enthusiasm from participants about being part of the study group.

We did not have any engagement from people on the etherpad at this workshop, and even some reluctance to use the sticky notes. Often only one or two people would put up a sticky but when helpers walked around, many more people actually needed and accepted help. Several students were doing well and consistently helping the learners sitting next to them.