Fix Markdown

Author: richelbilderbeek

evaluations/20260424_day_4/README.md

@@ -110,7 +110,7 @@
   Pedro has some amazing material and is a great lecturer
   but i am not intrested in parallel programming.
   Bjorn was mostly ok but maybe a bit fast (packages).
-  Rachel should have more advanced material,
+  Richel should have more advanced material,
   we are smarter than he thinks.
   Brigitte i didn't get through the sessions as batch is not intresting
   but i need it until you install ondemand everywhere.

reflections/202604_whole_course/README.md

@@ -4,7 +4,7 @@
 - Dates: 2026-04-20, 2026-04-22, 2026-04-23, 2026-04-24
 - Author: Richel
 
-## Self-rated confidence
+## Self-rated confidence again proves useless
 
 The average confidences per course, colored by course day:
 
@@ -19,6 +19,72 @@ Let's compare to the previous course:
 
 ![Confidences per question, colored by course day](../../evaluations/202511_course/confidences_per_question.png)
 
+From that, I see no obvious difference.
+To me, this again shows that the self-assessed confidences are unrelated
+to course quality (e.g. the review paper `[Yates et al., 2022]`,
+where no relation is claimed between self-assessment
+and the actual skill being taught,
+hence being used to assess teaching).
+
+## Other feedback
+
+I have already
+[reflected on feedback from day 3](../../reflections/20260423/README.md).
+
+Here I add the feedback from other days that is relevant to me:
+
+### Day 2
+
+- I didn't like the way in which they have taught during day 2.
+- big step up from day 1
+- Proper instructions and more time for breakout rooms
+  where we can talk to an instructor for help....
+- A big difference from day 1.
+  Sorry, but it needs to be improved a lot
+- Day 2 was much more disorganised and less pedagogical than day 1.
+
+Days 1 and 2 are different. Some of these mention a preference for day 1.
+
+### Day 4
+
+- Overall, [the pace] is OK, but it may be too fast for beginners,
+  even though it is meant to be an introduction.
+- Overall, while I think that the course has some good parts and some teachers
+  were very good and pedagogical, the course still goes too fast
+  for the most part, assumes that attendees know more prerequisites
+  than they actually need in order to attend,
+  and several of the teachers need to improve their teaching skills.
+- Teachers were all good, although Richèl stood out (positively).
+- I think the course has potential, but it could benefit from more
+  interaction with the students.
+- Finally, I would suggest removing the section on how to connect
+  to the HPC clusters.
+  Since this is a course about Python and Python in HPC clusters,
+  connecting to the cluster should be a prerequisite;
+  alternatively, it could be sent as a pre-course exercise
+  before the program starts.
+  The course should remain focused on Python
+  and its application within HPC clusters.
+- Richel should have more advanced material,
+  we are smarter than he thinks.
+
+Here seems to be some topics:
+
+- Q1: Who is the course aimed at: beginners of more experienced users?
+- Q2: Which prerequisites does the course have and do we **not** teach
+  what is assumed in those prerequisites?
+- Q3: How much Python should we teach (i.e. beyond HPC Python)?
+
+My answers:
+
+- A1: I follow the pace of the learners. This means I am beginner-friendly.
+- A2: I am quite strict in what I do teach and I do assume the prerequisites
+  being fulfilled
+- A3: As minimal as possible: the goal is to get things to work in an HPC
+  environment
+
+## Listen to feedback
+
 My question:
 
 Should we listen to feedback from the learners when they recommend
@@ -30,3 +96,25 @@ No. When we know that an advice goes against the literature, we should
 follow the literature instead. Would we be in doubt, then checking the
 literature is a great next step.
 
+## Actually listen to feedback
+
+Here is some feedback from [the May 2023 evaluations](../evaluations/20230523/evaluation_20230523.csv):
+
+- Please be prepared and make it more hands-on.
+- Test the modules work the day before instead of a week before!
+  More exercises for us to try
+- More exercises all over.
+- more time for code-along and on-your-own to try it
+- More GPU and ML examples. More about how to install packages in
+  virtual environment as I had problems
+- longer time to do the exercises
+
+This means already since May 2023 the learners ask for more time for
+exercises. Are they being listened to?
+
+## References
+
+- `[Yates et al., 2022]` Yates, Natasha, Suzanne Gough, and Victoria Brazil.
+  "Self-assessment: with all its limitations, why are we still measuring
+  and teaching it? Lessons from a scoping review."
+  Medical Teacher 44.11 (2022): 1296-1302.