Merge learning objectives slides

[estebangongora]

The first learning objectives slide (slide 4) does not include all the objectives and topics covered in the wiki. Later on, slide 17 does cover the learning objectives better so it might be useful to merge both slides or remove slide 4 altogether. Additionally, slide 17 may cause some comprehension issues as it suggests a flow diagram in which, for example, after performing a simple linear regression, you must move to a multiple linear regression by adding more variables. If time permits it, making a new version of the flow diagram might help to explain when to use each type of regression.

[dschoenig]

I think there are two problems to deal with:

1. The learning objectives themselves do not qualify as objectives that promote active learning
2. The flow chart is inconsistent (sometimes decisions are separated vertically, sometimes horizontally), and it gives the impressions that linear, regression, t-test, ANOVA are somewhat separate methods, while they're actually all facets of linear regression and use OLS.

[LaurieMaynard]

Currently, I merged the slides, but also adjusted the objectives to better fit the presentation. I also changed the title of slide 17 to Linear models, to emphasize to the fact that these are just variants of linear models.

I decided not to change the flow chart, due to time constraints, but as a further suggestion to improve this chart, I would remove the arrow between ANCOVA and ANOVA.

[dschoenig]

Sounds good! I agree, some work is needed to improve the flow chart. The problem is that I'm not completely sure what the aim of the flow chart is, a decision graph? a structure of the workshop?

I personally think the following things should be changed in a future round of revisions:

• In the chart, the linear regression needs a continuous predictor, this is not true, but categorical and continuous predictors are permitted.
• If the chart is to be udnerstood as a decision graph, then "2 variables" cannot follow form a method for "1 variable", instead there should be a fork: "1 variable" goes to one side, "2 variables" to the other. The same should apply to the number of levels or groups