Changes to 2021

[mattansb]

Each lesson should have a story driving the data operations...

1. Intro to R

(2 lessons)

Lesson 1

• [x] Operations
• [x] types
• [x] lists
• [x] sub-setting, pulling, replacing
• (Don't forget to assign)

Lesson 2

• refresher
• [x] for ()
• [x] if () and ifelse()
• [x] functions
• ~[ ] apply (or purrr::map()) (optional)~

2. Data wrangling

(2-3 lessons)

Explain how we normally work in R

Lesson 3

• [x] data frames
• [x] importing
• [x] dplyr and co
• [x] joining and making data long / wide
• [x] exporting (csv, rds, even sav)

Lesson 4

• [x] describing data
• [x] summarizing data
  • [x] reliability (psych)
• [x] Dealing with outliers (optional)
• [x] Dealing with missing data (optional)

3. Plotting

(1 lesson)

Lesson 5

• [x] ggplot2 <3

4. NHST

(1 lesson)

Lesson 6

• [x] htest-functions + correlations (psych)
• [x] Bayesian alts
• [x] effectsizes
• [x] power (optional)
  • simulation...

5. Modelling

(5 lesson) All models will have accompanying plots with ggeffects.

Lesson 7

• [x] linear regression
• [x] parameters
• [x] standardized coefficients ("beta")
• [x] predict() + newdata=
• [x] plot

Lesson 8

• [x] dummy
• [x] plot
• [x] emmeans and contrasts
• [x] comparing models
  • [x] indices of fit
  • [x] Bayesian

Lesson 9

• [x] moderation
  • What is this y ~ a + b*c / y ~ (a + b)*c / y ~ a*b*c
  • the parameters and also who cares about them?
  • centering
• [x] simple slopes (by cat / cont) with emmeans
• [x] plot with ggeffects (+ with interactions?)
• [x] curvilinear / poly

Lesson 10

• [x] (rm-)ANOVA
  • [x] long data
• [x] interaction
  • [x] Link to more....

Lesson 11

• [x] glm()s
  • [x] binomial
  • [x] Poisson - https://doi.org/10.1080/00223890802634175
  • there are more...
• [x] looking at response vs link scale

6. Assumptions

(1 lesson)

For fixed effect linear models...

Lesson 12

• [x] assumption checks
  • [x] Replace effects with ggeffects?
• [x] permutation test + basic bootstrap

7. Other topics (optional)

• [x] mediation (optional) (move to SEM?)
• [ ] Factor analyses? (already in SEM and partially in ML)

[mattansb]

https://CRAN.R-project.org/package=faux