Reflection

• Teacher: Richel
• Date: 2024-05-24
• Course: Intermediate Bianca course
• Lesson plan

This was the schedule of that day:

When	Who	What
9:00	R	Introduction
9:10	R	Transferring files to and from Bianca
10:00	.	Break
10:15	R	Transferring files p2
10:35	D	Slurm jobs p1
11:00	.	Break
11:15	D	Slurm jobs p2
12:00	.	Lunch
13:00	B	Software and packages installation
13:45	.	Break
14:00	BR	IDEs on Bianca
14:45	.	Break
15:00	L	NAISS-SENS summary
15:30	R	Summary
15:35	R	Evaluation
15:45	All	Optional Q/A and extra material
16:00	.	END

• Who: B: Björn, D: Diana, L: Lars, R: Richèl

First session: 9:00-10:35

Before the session, the three teachers discussed on a rule on when we teach: until 8:58 there was no learner at all! One minute later, the first learner arrived. From that moment on, every minute, a learner would trickle in, until we'd be with six learners.

The start was messy, learners being late requires improvisation and I did that, at the cost of following my lesson plan. I forgot to do a Prior, as instead I felt a stronger need to set the learners to work quickly. Only after talking about the shared document, I did a quick Prior on file transfer.

I had all learners turn on their camera, including those that were late. I shared the notes I prepared with every new learner coming in. In around 3 minutes, the learners there were set to work on using the shared document, while I welcomed the latecomers and set these to work too.

As I could see the number of answers, I saw that only 1 learner filled this in. Something was wrong. I asked the learners and let one share his screen and demonstrate. Now everyone saw what to do and the answers can in fast. Below are the results, marked by 'Start of shared document' and 'End of shared document' headers:

Start of shared document

File transfer

Goal of this exercise is:

• to practice using HackMD
• getting to know your prior knowledge

Below are three tallies, such as this example:

What is X?

- [ ] A
- [ ] B

Pick one character for you, e.g. * or R or ? or whatever (duplicates are fine!). Between the [ ] you agree on, write that character. In the end this may look like:

What is X?

- [s#k%vq*#6] A
- [q.vR] B

Now we know 9 people think A is the answer, where 4 people think it is B.

1. My favorite way to transfer files to/from Bianca now is using

• [xjsr] FileZilla
• [ ] fuse-sshfs and mounting the wharf
• [ ] lftp
• [Z] rsync
• [ ] scp
• [sl] sftp
• [ ] sshfs and mounting the wharf
• [ ] Rclone
• [ ] WinSCP
• [ ] other

2. About rsync

• [ r] I have never followed a lecture on rsync, I don't know what it is
• [sL] I have never heard a lecture on rsync, but I've never gotten it to work
• [j] I have never heard a lecture on rsync, I have gotten it to work, but don't use it in practice
• [Z] I have never heard a lecture on rsync, I have gotten it to work, and use it
• [x] I have heard a lecture on rsync, but I've never gotten it to work
• [] I have heard a lecture on rsync, I have gotten it to work, but don't use it in practice
• [ ] I have heard a lecture on rsync, I have gotten it to work, and use it

3. About FileZilla

• [ ] I have never followed a lecture on FileZilla, I don't know what it is
• [ ] I have never heard a lecture on FileZilla, but I've never gotten it to work
• [] I have never heard a lecture on FileZilla, I have gotten it to work, but don't use it in practice
• [slj] I have never heard a lecture on FileZilla, I have gotten it to work, and use it
• [r] I have heard a lecture on FileZilla, but I've never gotten it to work
• [ ] I have heard a lecture on FileZilla, I have gotten it to work, but don't use it in practice
• [] I have heard a lecture on FileZilla, I have gotten it to work, and use it

End of shared document

So, 4 learners use FileZilla, 1 rsync and 1 sftp. There has never been a lecture on rsync where (the one) learners got rsync to work. As most learners used FileZilla as their main file transfer program, I ended this poll earlier.

During the poll, I announced I would probably ask the person Z using mainly rsync to to show him/herself and consider to help teaching. Z identified himself and stated he was rusty in it, and we agreed he'd be treated like a regular learner.

After the Prior, I did the Present, mostly focusing on the 'Why', the terms that they can read themselves and how to do the exercises.

The learners were set to work in groups at around 9:20, later then I planned. Group sizes were 2x 3 learners, just what I prepared for. I was happy being able to have two groups.

At around 9:22, I went through the two breakout rooms. I was inconsistent and only shared all info in both rooms:

• (both) Ignore me in the future + reason why
• (only in room 1) Start reading first
• (only in room 2) Try all to do it individually

I should have had a checklist of things to say here ...

I checked the breakout rooms regularily and was commonly (as I encouraged them!) ignored: they were busy reading. I was happy with that.

2025-05-27

OK, I make a rougher evaluation now:

Newest team member

Our newest team member showed up in the Zoom. I set him to work. I told him some ideas behind my teachings. There were other colleagues in the Zoom room too, but they had to teach after my session, so I let them prepare their sessions instead

rsync teaching success rate

At the end of my session on rsync, I asked the learners to share if they transferred a file using rsync.

Results are:

• 5x yes
• 3x no
• 1x learner was in the wrong course: she has never accessed Bianca at all
• 1x learner used Windows, which was not documented in the UPPMAX documentation
• 1x learner just came into the Zoom room, as he wanted to follow the next session

So, I'd give my success rate a 5 out of 6, which is a 83%. Not a 100%, but reasonably good already.

IDE teaching success rate

At the end of the session on IDEs, I asked the learners to share if they have been able to start RStudiom

Results are:

• 4x yes
• 0x no

That is a 100% success rate!

Number of learners through time

Taking a look at the amount of learners in time looks like this:

• One learner was on time out of the 6 that would arrive in the first 6 minutes
• During the break, 4 left. As I can explain 2 of these (see below), I'd say that a third of the learners left without an explanation.
• One person just started using Bianca and was in the wrong course
• One person came for the Slurm part only (although he'd come back at 13:18 for a short while)

Comparing this with another online (but less interactive) course, does not give a good comparison, but already some insights.

Insights are:

• Learners are generally late

However, here I plot the cumulative numbers of learners, being present, having been present and having left:

Comparing this with another online (but less interactive course), does give some insights:

Insights are:

• In both courses, after the 12:00-13:00 break, only half of the learners are present
• In this course, however, until 12:00, no-one had left yet, compared to a third in the other online (but less interactive) course
• In this course, after the first break, no-one had left yet, compared to a quarter in the other online (but less interactive) course
• In this course, until 12:00, no-one had left yet, compared to a third in the other online (but less interactive) course

Evaluation

We had 4 learners filling in an evaluation. These were all the learners that stayed around till the end, hence a bias for the persistent learners. I announced the link to the evaluation at the start of the day, in the hope of catching the complete set of learners.

We decided to do an evaluation by looking at the raw evaluation results. I suggested to look at it later. The colleagues, however, wanted to take a look together and immediately.

We, including me, drew the wrong conclusions about which sessions scheduled enough time for exercises. The Google Form image, that I just took a screenshot of, shows a trend I can understand:

Out of the 4 learners there was 1 that wrote 'more time' at 'In the course, what should we improve?'.

Also, the confidences that learners had were hard to interpret:

So, I used some R scripting to give a proper plot like this:

All my sessions have above 'I have good confidence I can do this', so I think that is good enough. However, my aim is to have all to have 'I absolutely can do this!'. And there is one person that has low confidence in using rsync. This is because he is a Windows users, and he could not figure out how to install rsync. That should be fixed for next course iteration.

One evaluation result stated at 'Other':

Excellent, thank you

This should be celebrated more I think.

Taking a look at meeting_saved_chat.txt ...

R = me. Names starting with L are learners, names starting with C are colleagues.

one can see I sent by far the most chat messages to everyone. Weakness of this plot is that I can only see the messages to everyone and my private ones. Maybe others use more private communication.

Lessons for the future

• The timespan for the file transfer lesson was OK
• [DONE] Add installing rsync for Windows to UPPMAX doc
• [DONE] Use sftp in optional extra exercise instead of FileZilla
• Using a Google Form for evaluations works fine. However, analyzing a raw Google Form with colleagues has proven to be useless to me again
• Using the confidence learners have on sessions in the evaluation seems to be useful
• Asking the learners if sessions scheduled enough time for exercises seems to be useful
• We maintained having all learners in the morning session. This is better than I see in some other less-interactive courses
• We lost a third of the learners in the 12:00-13:00 break without explanation. Is this a common pattern?