1) Another measure of centrality is the midrange: It is the average of the minimum and the maximum values in a data set.
a) How sensitive is the midrange to outliers, compared to the mean and median?
b) How, if at all, is the midrange affected by additive shifts (e.g., each value in the data set increases by 1)?
c) How, if at all, is the midrange affected by multiplicative shifts (e.g., each value in the data set is multiplied by 12)?
2) Consider sound levels, measured in decibels (db):
• 0 db indicates the softest sound that the human ear can hear unaided (which is amazingly close to total silence)
• The decibel scale is a logarithmic scale (e.g., a 20 db sound is ten times as loud as a 10 db sound)
This doesn't fit neatly into our four data type boxes (nominal / ordinal / interval / ratio). What do you think would be the best way to describe sound levels? (Hint: Would a change in scale help?)