Since we chatted in our first class today, we started straight away on discussing the importance of statistics in scientific settings, and in everyday life. Without the rules of statistics, large datasets become unmanageable, and cannot be represented in a logical, coherent, and relevant manner. One example is the large amount of data used to map the Universe in 3D (i.e. measuring the distances to galaxies).

Emma was absent.

I helped with a linear regression assignment involving the use of StatCrunch and Microsoft Excel.

We covered a myriad of subjects as part of this review, including probability theory, dependent vs independent events, complementary events, probabilities under a uniform probability distribution and finding a sample mean from a frequency table. We also reviewed using the TI-83 Graphing Calculator in several statistics applications.

We discussed basic definitions, including "statistics," "parameters," "samples," populations, etc. We reviewed random samples and went over one-variable statistics routines on the TI-83-Plus graphing calculator.

We worked problems from a Final Exam worksheet assigned by the professor.

We spent the session working on a lengthy problem set. The set began with uniformly distributed times and problems with area of part(s) of a rectangle equated to probability. Then we moved on to standard normal distribution with probabilities equating to areas under the normal curve.

We reviewed Chapter 8, hypothesis testing using P-value and Critical Value techniques. We also went over Confidence Intervals.

We spent the session doing a problem in which we examined the claim that shoppers with higher denominations of cash tend to spend less than shoppers with lower denominations. (Two proportions, two samples.) We used the P-value and Critical Value methods to investigate the claim and then used a confidence interval to explore the difference between the two proportions. The evidence suggested that shoppers with higher denominations do spend less. But the sample consisted only of undergraduate students, so it would be risky to attribute this propensity to all American shoppers.

We explored problems in which two populations (e.g., husbands and wives) were expected to have different heights and used the null hypothesis that the mean heights were the same. The example problem gave data for husbands and wives' heights. We tested the null hypothesis and, based on the data, rejected it.