This is the website for the course PSYCH101-D, Data Science for Research Psychology, taught at UC Berkeley in Fall 2019.

Course Description

Experimental and data science abound with models. Models of data can be used to simulate, as in models of the climate, to explain simply, as in paper airplanes, and to predict, as in prototype models; all of these are forms of inferential thinking. In this course, we will learn to use Python to describe, create, manipulate, and interrogate models of data. With these new skills, we will simulate, explain, and predict phenomena and data, drawing examples from research psychology. As one application of these tools, we will develop classical statistical approaches, like null hypothesis significance testing and linear regression.

Course Materials

Category Date Links Title
lec 08/28 datahub Binder Lec 00 - Welcome
lab 08/30 datahub Binder Lab 00 - Getting to Know Jupyter and OK
lec 09/04 datahub Binder Lec 01 - Python and Pandas and Plots
lab 09/06 datahub Binder Lab 01 - Visualize Munge Repeat
hw 09/09 datahub Binder Homework 01 - Python Drills
lec 09/09 datahub Binder Lec 02 - Probability and Statistics
lab 09/13 datahub Binder Lab 02 - Bootstrap Estimation for Personality Data
hw 09/16 datahub Binder Homework 02 - Descriptive Statistics and Bootstrapping
lec 09/16 datahub Binder Lec 03 - Models and Random Variables
hw 09/20 datahub Binder Homework 03 - Models IRL
lab 09/20 datahub Binder Lab 03 - Models & Monsters
lec 09/23 datahub Binder Lec 04 - Parameters and Priors and Posteriors
lab 09/27 datahub Binder Lab 04 - Comparing Bootstrap and Model-Based Sampling
lec 09/30 datahub Binder Lec 05 - Null Models
lab 10/04 datahub Binder Lab 05 - Modeling Null Hypothesis Significance Testing
lec 10/07 datahub Binder Lec 06 - Bayesian Inference
lab 10/11 datahub Binder Lab 06 - Bayesian Inference for Differences of Means
hw 10/14 datahub Binder Homework 04 - Parameterized Models and t-Tests
lec 10/21 datahub Binder Lec 07 - Categorical Effects Models
lab 10/25 datahub Binder Lab 07 - Modeling Categorical Effects
lec 10/30 datahub Binder Review of Modeling Concepts
lec 11/06 datahub Binder Lec 08 - Multiway Modeling
lab 11/18 datahub Binder Lab 08 - Multiway Modeling
lec 11/18 datahub Binder Lec 09 - Regression
proj 11/18 datahub Binder Project Proposal
lab 11/25 datahub Binder Lab 09 - Bayesian Linear Regression
lec 11/25 datahub Binder Lec 10 - Formulas and Linear Models
lec 12/02 datahub Binder Lec 11 - Over-Fitting and Cross-Validation
hw 12/04 datahub Binder Homework 05 - Linear Models
lec 12/04 datahub Binder Lec 12 - Under the Hood of pyMC

How to Use the Materials

The table above has links to all of the materials for the course. Either one will drop you into a cloud-computing environment in your browser, where you can review lectures, complete homeworks and labs, and get scores on autograded sections.

Berkeley Users

The links which look like interact, are for use by Berkeley-affiliated folks. After logging in with your CalNet ID, you’ll be dropped into, where a copy of the course materials, plus any file you add, will be maintained for your own use. If you’ve never used this service, check out this video.

Non-Berkeley Users

The links which look like Binder, are for use by non-Berkeley-affiliated folks, i.e. anyone without a CalNet ID. After clicking on these, you’ll need to wait through a short (<2 min) build period. If you encounter an error during this build, click the link a second time. If the build still fails, contact the course staff.

Once the build succeeds, you’ll be dropped into a temporary cloud environment, where you can look through and edit the materials, but any changes you make will not be saved. You can still complete homeworks and labs and get scores on autograded sections. If you stop interacting with the material (viewing pages, clicking, typing) for an extended period (~20 minutes), your environment will be deleted. If you need to step away while working, you can download assignments to your local machine, using the file menu in the Jupyter notebok, and then reupload them using the Upload button in the Jupyter file browser.

If you’re not Berkeley-affiliated and would like a persistent copy of the course materials, you’ll need to perform a local installation of the environment using git, GitHub, and pip. See instructions below.

Local Installation

For instructions on local installation, see the GitHub repository where the materials are stored.