Day 04: Statistics for Business Applications

Last Time:
- qualities of histograms
  - symmetry
  - unimodal, bimodal, or multimodeal?
  - What are typical (central) values?
  - What are the largest and smallest values?
  - Are there unusual or atypical values?
  - wildcard: bin size!
- measures of central tendency
  - mean - average value (center of mass of histogram)
  - median - middle value
  - mode - most common value (or class midpoint)
Today:
- If you have problems with data for 2.3, #25, I've obtained a corrected data file (ASCII or Excel format only!)
  This used to be a problem, and will be if you use the data from the CD: I'm not sure if it's been fixed on the server....
- Collect Homework #1; Homework #2 due next Tuesday; homework #3 (associated with chapter 4) due Tuesday, 2/3
- Do you have any questions about previous stuff?
- Chapter 4: Measure of central tendency, and measures of spread
  - Example: M&Ms (Revisit data from last time)
  - Calculation of the mean (summation notation)
  - range
    Difference between largest and smallest values
  - variance
    mean of the squared deviations (differences from the mean) of the data.
    Notation - population variance (sigma squared) versus sample variance
  - standard deviation
    square root of the variance - it has the same units as the mean
  - Computing means, standard deviations, etc. with Minitab:
```
Stat -> Basic Statistics -> Display Descriptive Statistics
```
  - There are two rules based on the standard deviation that you'll want to be aware of:
    - The empirical rule - the 68/95/99 rule.
      - The standard deviation serves as a "typical deviation"
      - Works well when the distribution is bell-shaped
      - May work okay in any event.
    - Chebysheff's Theorem - the 1-(1/k^2) rule.
      - Works all the time, but is extremely conservative: it leads to a 0/75/89 rule!
Next time:
- Homework #2 due (chapter 2)
- finish measure of spread (standard deviation), start normal distribution (the bell-shaped curve)

Links:

Website maintained by Andy Long. Comments appreciated.