QUANTitative Analysis & Statistics

Passion Driven Statistics

"A multidisciplinary, project-based curriculum that supports students in conducting original research, asking original questions, and communicating methods and results using the language of statistics." See the website for an E-Book with videos, code, and assignment

The following videos linked here (to YouTube, or see the website) do not demonstrate any software (used for SPSS). See videos for SAS, Python, R, and Stata (video lengths vary). Uses Base R for data management.

Note: Explanatory Variable = Independent Variable (IV), Response Variable = Dependent Variable (DV)

UK Data Service

Data skills modules

Interactive text and video tutorials

• Survey data (with SPSS) - Learn about how to find survey data from the UK and around the world, what you need to know about it before you can start analysing it, and how to produce simple tables and graphs for your research or reports.
• Aggregate data - Learn where to find these data, how to produce summary statistics using them and how to create a choropleth map.

These suggest completing the surveys module first.

• Longitudinal data - Learn what longitudinal studies are available, key features and issues with using longitudinal data and how to start some basic analyses.
• Exploring crime surveys with R (Beta version) - Learn where to find survey data to explore crime and how to access it. This module also covers getting started with survey data using R Studio including basic functions, exploratory analysis, visualisations and conducting weighted analyses.

YouTube Channel

These include the videos from the above DataSkills modules.

Eva Witesman

In management

Data Management

Data Cleaning

• Data Cleaning basics (~20 min) - Incorrect and missing values
• Deidentifying data (~7 min)

Data Structure

• Unit of analysis revisited (~13.5 min)
• Long vs wide data formats, plus unit of analysis (~8.5 min)
• Orientation to long formatted dataset (~4 min)

Statistics

Fundamentals

• Descriptive statistics (~16 min)
• Two purposes for data description (~2 min)
• Level of measurement (~17 min) - Interval, Ordinal, Categorical, Binary
• Gestalt principles and data visualization (~12.5 min)

Causality

• The four criteria of causation (and how correlation and causation are related) (~6.5 min)
• Question types revisited (especially predictive) (~8 min) - Single Cause, Greatest Cause, Predictive
  • Older: Three question types (~13 min)
• Interpreting regression statistics for each question type (~9 min)
• Statistical conclusion validity (~10 min) - Not enough data, Break the rules, Fishing, Restricted range, Inconsistent measurement, Insufficient controls (environment and units), Biased statistics

Less-Unique Videos

Regression

• Correlation, graphing, lines, and slopes (~4.5 min) - Positive, Negative, and No Correlation
• Lines, regression lines, and formulas (plus null hypothesis test!) (~16 min) - y = mx+b vs y = B0 +B1X + E in a Single Cause model. Reviews positive, negative, and no correlation, and the 4 rules of causality.
• R-squared statistic (~15 min)

Others

Using R

• Why I teach statistical programming in R and RStudio (pros and cons list) (~8.5 min)
• Analysis is a creative process (and how to get help) (~6 min)
• text analysis fun in R (~14 min)

Other Topics

• Five header rows codebook method plus how to save as .csv (~18.5 min)
• Analysis types for evaluation (~8 min) - Uses the pen writing
• Sampling Review (~11 min)