ISA Statistics, Prof. Janeba

• Section 4.1
1. Name two measures that give the center of a set of data.
2. Name two measures that give the spread of a set of data.
3. Which example shows that knowing the center for a dataset doesn't tell you all about the dataset?
4. Which example shows that knowing both the center and the spread doesn't tell you all you'd want to know?
5. If center and spread aren't enough to know, what's another method of summarizing data that tells us much more?

 
• Section 4.2
1. Why oughtn't we conclude from figure 3 that as people age, their heights go down steadily starting in their 20's, while their weights go up and then down?
• Section 4.3
1. What do you look for in a histogram to estimate the average?  What about estimating the median instead?
2. Which is more greatly affected by a single datum much larger than the rest - the average or the median? (see figure 6)
3. For histograms with long tails on one side:  which is farther out on the tail - the average or the median? (see figure 7)
• Section 4.4
1. Expand the phrase Root mean square into a larger expression that describes more fully what RMS means.
2. Make sure you can actually compute the RMS of a list of numbers.
• Section 4.5
1. Write a careful sentence that gives the precise relationship between the Standard Deviation and RMS.
2. What, roughly, does the SD tell us about a list of numbers?
3. (From class) If we want the "typical deviation size", why not just average the deviations?
4. What is the "rule of thumb" about SD's - how much of a list is typically within one SD of the average?
5. Continue the "rule of thumb" from the previous question (see box on p. 68, and note third part given in class).
6. How often does this "rule of thumb" actually work?
• Section 4.6,4.7
1. Make sure you can actually compute the SD of a list of numbers, both by hand and with your calculator.
2. For help using many popular calculators, see my help files.
3. When is the SD of a list equal to zero?
4. If all the numbers in a list are exactly 5 units away from the list's average, then the SD will be ...