future_changes

lectures overall
* add 60 second break in the middle 
* email students -- share FYP for class?
* more in-class activities (breakout groups like simple slope drawings)
* make a flow chart or organization for material? See example shared by S Lieber

course website:
* include in required materials: LaTex
* Add link to this: https://statmodeling.stat.columbia.edu/wp-content/uploads/2020/07/raos_tips.pdf

labs overall
* show only broom and kable for creating table summaries throughout
* in lab 9, show them papaja in depth (add links to other options, like stargazer, apaTables, and sjPlot to just create tables in Word)
* incorporate in-text R code to summarize results throughout 
* most useful to teach from R, not from html output
* should we replace papaja with data cleaning (for loops, functions, regular expressions, etc)

final-project.html
* add descriptive statistics back in?

1-correlation.html
- use t or z for statistical test?
- useful to read (skim) chapter before (send email week before)

2-correlation.html

- move scatterplots to before correlation matrices
- add slides about MAR/MCAR/MNAR 

1-lab.html

  * maybe add missingness (but included in class already) -- minihack?
  * change Dawtry example to something else
  
3-regression.html
  
  * go back and add i's 
  * double check z for standardized regression equation
  * remove sxy etc from slide 19
  * positive feedback on discussion of eugenics

4-regression.html
  
  * check z's for standardized regression equation
  * add the squares
  * add this! https://joshualoftus.com/post/2020-11-23-least-squares-as-springs/least-squares-as-springs/ 
  
2-lab.html

  * more interactive labs are good! This one is already interactive, so good template for future labs.

5-regression.html

  * move gganimate from lab 3 to here 

6-glm.html

  * possible terms to use for model comparison 
    * compact model (simpler)
    * augmented model (more complex, wanted to test)
    * proportional reduction in R^2
  * why is GLM taught before ANOVA?
  
3-lab.html

  * labs feeling repetitive
    * could this be a pandemic-specific thing?
    * options
      * move papaja lab here?
      * more applied lab? 
        * here's a question -- how would you test it?
    

7-partial.html

  * arrow from parent ses to peer relations
  * when to use each: only care about X and Y and controlling for variable; comparions between diff x's and incremental validity; network analysis
  
8-m_regression.html
  
  * remove ID variable from dataset
  * no matter where you cut the plane, the angle is the same
  * walk through GIF -- break this up into steps and different slides. May require building my own figures
  
lab-4.html

  * no changes needed

lab-5.html

  * no changes needed

9-m_regression.html

  * I already updated the example 

10-m_regression.html

  * don't use factorial design for example
  * liked interpreting coefficents when controlling for variables
  
11/12-diagnostics.html

  * better example for homoscedasticity (it wasn't very clear)
  * add code for multicollinearity
  * explain independence plots more -- give strong examples of each end (dependence, independence)

lab-6.html

  * this is a boring lab. How can we spice it up?
    * murder mystery mini-hacks (something's wrong with this model)

13-dag.html
* add paper by Grosz, Rohrer, & Thoemmes (2020) "The taboo against explicit causal inference in non experimental psychology." As optional reading. Describe during class?
* clarify "causality"

14-interactions.html
* use name of variable instead of Y-hat

homework-2.html
* make due last week of class
* weights are 1/predicted values, not predicted values!!!

lab-8.html
* is it worth keeping in contrast codes? 

15-interactions.html
* write out equations on the screen 

16-18-factorial.html
* keep venn diagrams in sums of squares
* give examples for when to use different SS
* switch order of simple slope examples

19-interactions.html
* blank slides

20-poly-boot.html
* keep -- when do I use this!

lab-9.html
* short and straightfoward -- good for this lab.