ISDS 361B Ch. 4

  1. What r the 3 different measures to describe the center of a set of data?
    • Mean
    • Median
    • Mode
  2. Mean?
    It is the average.
  3. Median?
    • Central location. gives you the middle where 50% of the numbers lie on each side.
    • *When there is an even # of observations, the median is determined by averaging the 2 observations in the middle.
  4. Mode?
    the observation with the greatest frequency
  5. Why is Mode not a good measure to use?
    • In small samples not a good measure to use
    • it may not be unique
  6. How do you compute mean, median and mode on excel using one button?
    Data Anaylsis --> Descriptive Statistics (then input the range of numbers) --> Summary statistics.
  7. Which central measure is best?
    Mean is used most often, but median could be better because it is not sensitive to extreme values.
  8. When to compute Mean?
    • Interval data
    • u want to describe a single set of data
  9. When to compute Median?
    • Interval or Ordinal data (w/extreme observations)
    • Describe a single set of data

    Ordinal data - categorized with meaning
  10. Range
    Largest observation - Smallest observation

    pg. 106
  11. What are the measures of variability for interval data only?
    • Range
    • Variance
    • Standard Deviation
    • Coefficient of Variation
  12. Is Variance a good way to explain a set of data? What could it tell us when compared to another set of data?
    No, cuz it only gives us a rough idea of the variation in the data. Comparing to another set of data it could tell us if the 1st set of data has more variation than the 2nd set.
  13. Variance interpretations are problematic cuz we squared the deviations to get the mean. How do you fix this problem?
    you calc the standard deviation. Units associated with SD is the unit of the original data set.

    pg. 110
  14. How do U interpret Standard Deviation?
    • Empirical Rule - Bell Shape
    • Chebysheff's Theorem - all shapes
  15. Empirical Rule
    Ex: histogram is bell shaped. mean and SD are 10% and 8%. Interpret
    • 68% fall in between 1 SD, so 18% and 2%
    • 95% fall in between 2 SD, so 26% and -6%
    • 99.7% fall in between 3 SD
  16. Chebysheff's Theorem
    • provides lower bounds
    • 75% lie between 2 SD
    • 88.9% lie between 3 SD
  17. Coefficient of Variation
    Indicates whether the SD (variation) is large or small compared to the number of observations.

    SD / by their mean
  18. How do you word the Chef theorem?
    • At least 75% of the numbers lie between
    • At least 88.9% of the numbers lie between
Card Set
ISDS 361B Ch. 4