In an effort to start exploring whether the (relative) surplus of baseball players born in the months of February, June, and August is a real effect or due to normal human patterns (holidays, etc) Jared started expanding on the database by examining the birthdays collected here at the school and those of US presidents. Thus far, we cannot conclude much with any certainty, but the database is growing, as is Jared's understanding of means, standard deviations, and data collecting.

Since we completed the concepts, and her review from school, we went through other example problems to prepare her for her exam this week.

After finishing the last few minutes of the latest Hans Rosling documentary we've been watching, we returned to Jared's long-term project. I first reviewed the principle of Standard Deviation, and stressed that the more common interpretation requires the distribution of the data to be described by a bell curve. For Jared's project, this is not the case, so if we want to examine the data in terms of standard deviations from the mean, then we must invoke Chebyshev's theorem, which states that 75% of the data lie within 2s and 89% within 3s.

We continued with Hans Rosling's documentary, and - as always - I would pause when I wanted to elaborate on a particular concept, or more specifically when Jared had a question or a comment. I was very impressed today when Jared brought up multiple occasions of his noticing when statistics were presented erroneously. He specifically brought up the practice of 'cherrypicking', when only a certain subset of a dataset is presented to the public in an effort to highlight a specific philosophy or agenda (at times this may also occur completely subconsciously due to innate biases). Jared's specific example today referred to college dropouts that made it big (e.g. Bill Gates, Steve Jobs, Mark Zuckerberg) in a program that tried to dissuade kids from going to college. Jared himself recognized the flaws in this argument: that the majority of the dropouts never made it big (and were not mentioned) and that these dropouts were all at very prestigious schools.

We went through her review packet for her test on hypothesis testing with 2 samples.

We picked up on Hans Rosling's documentary, 'The Joy of Stats', where we left off several weeks ago. Today, he covered the most interesting topic of correlation analysis. Correlations can be found between almost any two random things - there are whole websites devoted to this, mostly humorous, but they illustrate how dangerous it can be if people without the mathematical skills to understand correlations start trying to draw conclusions between datasets. However, the science of correlations underlies much of statistics. As an example, I used my work from my astrophysics days to show a correlation analysis undertaken on radio, soft X-ray, and hard X-ray emission from a black hole X-ray binary, and how we were able to conclude that previously thought unrelated emission processes were, in fact, traceable to a common physical origin (for the first time ever).

We continued working on ACT practice problems, mostly algebra, but I also picked some out that had a statistics bent to them.

After consulting with Jackie, we thought it most beneficial if Jared used up our second hour today on working on ACT problems in preparation for tomorrow's test.

We finished feeding in the numbers into our 'database'. Jared then totted up, using Excel's inbuilt functions, some of the numbers. We will soon be able to start analyzing, in a statistically significant manner, the results,

We finished tabulating the data for Jared's long-term project. Now he has the full dataset to start working on. We used up the last fifteen minutes or so of the lesson to discuss what different operations can be applied to the dataset to try to extract as much information as possible. We will work on this from time to time in the weeks to come.