Real news articles and other resources to motivate discussion in undergraduate statistics courses.
New news cite: https://realworlddatascience.net/
- NYTimes What's Going on in This Graph? More specifically:
- 2019 - What effect does what we eat, and the food we waste, have on the environment? -- Bar plot and stacked bar plot
- Recent #TidyTuesday tweets (a different dataset each week)
- Georgia's Top 5 Counties with the Greatest Number of Confirmed COVID-19 Cases -- Clustered bar graphs, time series plots, misleading plots, misleading data.
- The issue with pie charts and in general the Data to Vis website is useful for exploring which plots to use for what
- 2017 - Alternatives to pie charts
- 2019 - How does Mike Soroka do it? The last plot shows the relative frequency of the different pitches for each count. The experiment is Mike Soroka throwing a pitch and the relative frequency of a specific pitch is the number of those pitches that he threw divided by the total number of pitches. You can imagine that if he threw an infinite number of pitches we could get the exact probabilities.
- 2018 - When does the hottest day of the year usually occur? In the 2nd through 5th plot, the relative frequency is plotted in space. In figure 2, for example, we see the number of years where the hottest day was in June divided by the total number of years. The experiment is seeing when the hottest day is each year.
- 2015 - The most distinctive words in online dating profiles, by state. They compare the relative frequency of words in dating profiles in the different states. The experiment would be each word and the relative frequency of a specific word would be the number of times that word was used divided by the total number of words.
- 2017 - Combinations? Permutations? Those words don't mean what you think they mean. "Suppose you are performing clickstream analysis for a company and there is a large number of ways in which a customer can navigate through the website. Assuming your data set is large, and there are many visits to the website, you're likely to apply machine learning (ML) in your investigations. A crucial point to consider very early on is whether you are interested in customers taking specific routes through the site (permutations), or just visiting groups of pages together (combinations) because that can significantly impact the choice of ML algorithm you might use."
- Where did Bayes' Theorem come from? "Thomas Bayes (1701 — 1761) was an English theologian and mathematician."
- 2020 - Using the Hypergeometric distribution in the Magic game
- 2009 - One place the Hypergeometric distribution is used is in representative drug sampling. Basically, the police seize a lot of packages of drugs, but testing each package to see if it contains controlled substances is time consuming and expensive. Instead they randomly sample some and look at the probability of all the packages in the sample containing a controlled substance given certain values of the parameter r. If you are extremely dedicated, there is a section in this document which described the method in detail.
- 2019 - The Poisson Distribution and Poisson Process Explained
- 2019 - Predicting the Frequency of Asteroid Impacts with a Poisson Processes
- 2019 - Statistics in court: Incorrect probabilities includes a brief summary of the Sally Clark case and others where the product rule was misused.