Pulling Back the Green Curtain: Data Forensics, Power BI, and Dataviz
summary
Uses R + RStudio (ggplot2) and Power BI to run Benford’s-law checks on vote tallies: scripts melt the CSV, extract first digits, count frequencies and plot actual vs expected. Finds digit "1" about 18.8% vs Benford’s ~30%, so the Russian summary data doesn’t neatly match the law yet independent significance tests are needed. Notes practical limits: Benford fails on small samples, capped or nominal fields and on aggregated vs line-level data. The R scripts and plots run and highlight where to investigate further.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:07Data analysis creativity
- 3:12Common data mistakes
- 5:19Data forensics importance
- 6:45Data quality blind spots
- 8:08Descriptive analysis basics
- 9:53Diagnostic analysis truths
- 12:21Benford's law anomalies
- 19:28When Benford's law fails
- 21:58Missing data handling
- 24:25R and Power BI integration
- 25:50Installing R
- 27:16RStudio basics
- 35:51Russian election data analysis
- 41:14R data manipulation
- 44:51Benford's law voting analysis
- 48:50R and ggplot2 visualization
- 53:50Summary
- 54:32Q&A
related talks
01h 42mRecreating, Understanding, and Visualizing FiveThirtyEight's...
01h 01mEffective Data Visualization
22mWhen statistics become stories
01h 38mEffective Data Visualization
01h 10mStorytelling through data
19mThe R integration in Power BI: Why R developers should be Interested
01h 01m50 ways to show your data
08mEvil Uses of Data