-
Notifications
You must be signed in to change notification settings - Fork 9
/
Copy pathteaching-notes.txt
80 lines (38 loc) · 2.73 KB
/
teaching-notes.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
2018-10-25 10am-1pm (3 hours, 15 minute break at 11:30)
Skipped limma section. Rushed in teeth section. Skipped multiple testing. Pace was otherwise good.
Formulas were a significant source of confusion:
- dropping the intercept and the factor getting an extra column was very confusing
- ~0 and ~1 models in slideshow were a bit confusing.
Move some discussion of formula syntax earlier.
Maybe go up to four hours to allow time for limma. Still had energy at the end of the workshop.
Here are some introductory talking points:
View this workshop as "vocational training".
Theory and practice will re-enforce each other.
Exploration remains important, but not covered in detail today.
Trying to provide a "complete" toolbox, up to a certain point.
Trying to provide generic rather than specialized methods, skipping useful shortcuts.
Statistics and R have a lot of history.
I'm skipping over a lot -- people view the same underlying mathematics in different ways.
Particularly ANOVA is a separate but often equivalent line of development.
Presentation is generic, but somewhat shaded by needs in bioinformatics.
Not covering GLMs or mixed models, however you will have a good foundation to learn about them.
Many of the same functions apply.
Now introduce yourself to your neighbours and introduce yourself on the etherpad.
2019-02-13 11am-4pm (5 hours total: 4 hours teaching, 1 hour break for lunch)
Pretty much covered the material, but rushed at the end.
- Didn't do the demonstration that limma exactly matches glht -- probably superfluous.
Final slides summarizing things learned could be trimmed back.
It would be nice to get to a challenge earlier, first challenge reached at 12:30pm.
2019-08-28 10:30am-4pm (5.5 hours total: 4.5 hours teaching, 1 hour lunch)
Had multiple new people teaching. Very rushed at end.
The ~0+group is very hard, you need to follow three tricks.
- 0+ means no intercept
- but group encoding adapts to this
- and solution process also adapts
... maybe a different challenge or add a challenge earlier
log(1/40) is poorly thought out
2019-11-21 10:30am-4pm + 30mins overtime (5.5+.5 hours total with 15min morning tea, 30min lunch, 15min afternoon tea)
Had multiple people teaching, still rather rushed at end. Went 30min overtime talking very quickly about RNA-Seq -- this section really is too much at the end of a day and needs to be a separate workshop. Even without the RNA-Seq section, adding 30 minutes to bring this up to a 6 hour workshop would be ideal.
Revised step-by-step challenge about ~0+group seemed to go well.
Need to keep up a slightly faster pace in the introductory vector/matrix section.
The morning and afternoon tea breaks are very much needed.