Click each link to expand or contract the details for that week.

By the end of the week, learners will be able to:

- Describe how to ensure your success in this class
- Get to know your classmates
- Describe what Self Regulated Learning means to you
- Create a reproducible workflow for data analysis

- Familiarize yourself with this website and required course materials.
- Review the syllabus and start on HW 0.
- Read PMA6 Ch3 and ASCN Ch 1 (Before Wednesday)
- Refresh on linear regression (Before Fri) PMA6 Ch 8, ASCN Ch 7,9

- No school. Strike day

- Welcome to the class. Checking in on HW00/logistics
- Reproducible workflows (whiteboard).
- Explain DM check ins

- Group Quiz on class logistics & Data Preparation
- Jump start - write down everything you know about LinReg (LJ)
- Work in groups to organize ideas into topics (Whiteboard)

- Practice asking questions
- Interpret different types of predictors
- Identify moderating and confounding variables
- Fit and interpret an interaction model

- Read ASCN Ch 9.6, Ch 8
- Review HW1 to see what is expected of you
- Convert your class folder to an R Project (video in Canvas)
- Review the example data management files in Canvas. Start on your personal dm files.

- QFT: Model Building & Variable Selection
- Recap on the purpose of linear regression models, assumptions, interpretation of predictors.

- Interpretation of Categorical Variables (9.4)
- Presenting results (9.5)
- What does it mean for a variable to be a confounder (9.6)

- Stratification & Moderation (8.1-8.6)
- Interactions (10.1, 10.2)

- Perform various variable selection techniques
- Identify pros and cons for each method

- PMA6 CH9, ASCN 10
- HW1 draft due

- Introduction to Peer Review
- Check in on interactions with categorical variables.
- Testing interaction terms (10.2)
- Multicollinearity (10.3)
- Automated Variable Selection (10.4)

- Comparing between models (10.5)

- Group Quiz on Model Building
- General advice, what to watch out for.

- Build and interpret a Logistic regression model on binary data
- Use measure of model fit to compare between models.

- PMA6 CH12.1-12.8, ASCN Ch11.1-11.3
- Check your LJ for completion
- Work on your dm files

- HW1 final
- QFT/LJ: Non-continuous outcomes
- Fitting and interpreting Logistic Regression models. (ASCN Ch 11.1, 11.3)

- DM File check in
- Odds Ratios are always the odds of an event for one group compared to another group. (11.4)
- Logistic regression worksheet

- LJ Check in
- Use logistic regression to classify observations into two groups. (Ch 12)

- Use Logistic Regression to classify observations into two groups
- Identify the optimal cutoff point for a binary classifier
- Create and interpret a ROC curve
- Create a confusion matrix
- Calculate and explain terms such as Sensitivity, Specificity, and Accuracy

- Read ASCN Ch 12
- Install packages: caret, ROCR

- Confusion matrix.
- Sensitivity, specificity, accuracy
- ROC curves
- Changing the cut point -default is not always best

- Open work day to finish homework 2

- Group quiz on Logistic Regression & Classification
- Exam 1 review session

*NA*

- Review all prior materials. Homework, quizzes, discussion boards, QFT.
- Draft a 1 page set of notes for each of the 3 topics covered so far.
- Write 1 serious exam question (don’t cheese this)

- QFT on Missing Data
- Effects of non-response

- Exam 1 (MLR, Variable Selection, LogReg, Classification & Prediction)

- Exam1 cont.
- What are the mechanisms in which data can be missing?

- Explain the effects of missing data.
- List and define the different missing data mechanisms.
- Explain the typical methods of handling missing data and the problems with each.
- Explain the mathematical model behind two imputation methods

- Watch the Seminar on Missing Data: https://media.csuchico.edu/media/0_tgnydpgf and reflect on what seemed to be the most important ideas in that talk in your LJ

- What strategies are available for handling missing data? (18.4)
- What are some methods for imputation? (18.5)

- Multiple imputation is the gold standard of how to analyze data with missing values. (18.6)

- Discuss specific details for one specific method called MICE: Multiple Imputation using Chained Equations

- Explain the mathematical model behind multiple imputation using chained equations
- Conduct multiple imputation on a data set and analyze the results.

- install packages: mice, VIM

- Flex day

- Career Panel

- Group quiz on Missing Data
- Special Topics Working session

- Explain how PCA can be used as a dimension reduction technique
- Explain the difference between multivariate and multivariable
- Conduct a PCA using both the correlation and covariance matrix
- Use visualization techniques to identify the number of PC’s to retain

- Read ASCN Ch 13 and 14.1 (PMA6 14.1 & 14.2)

- QFT on Dimension reduction
- Introduction to PCA (14.1)
- Basic Idea of PCA, how it connects to linear algebra. (14.2)

- DM check in 2
- Generating PC’s using R (14.4)
- Data Reduction (14.5)
- Standardizing (14.6)

- Use in multiple regression (14.8)
- What to watch out for (14.9)

- Use visualization techniques to identify the number of PC’s to retain
- Explain the difference between PCA and FA
- Create a latent factor model, visualize and interpret results.
- Use latent factor scores as a predictor in another model

- Read ASCN 15, and PMA6 15

- No school - Cesar Chavez Day

- Introduction to Factor Analysis
- Factor Model (15.2)
- Factor Extraction (15.4)

- Rotating Factors (15.5)
- Factor scores (15.6)

*NA*

*NA*

- Group quiz on Correlated Models
- Flex day

- Exam 2 review

- Exam 2 (Missing Data, PCA, HLM)