The 3 Ms of Central Tendency

                                     Mean, Median, and Mode
                                             Which one should you use?




Image retrieved 3/25/2012. From http://m-m.seebyseeing.net/
Why Does It Matter?
   You live in a town of 1,000 people where the
    mean income is $25,000 and the median
    income is about the same.




                         Retrieved from http://en.wikipedia.org/wiki/Damascus,_Virginia
                         4/7/2012
Why Does It Matter?

   Now the mean income is $1,023,976, but
    the median income is still about $25,000.




                Retrieved from http://genehoyas.com/ 4/7/2012
The Town of Millionaires

   It is a wonderful story about this
    whole town of millionaires who
    are so frugal they live in
    ordinary homes. What a great
    example of America!
   Why is this misleading?
Why Does It Matter?
   Most people in this town are still pretty
    poor.
   The mean does not give a good picture of
    the “typical” resident.
   This could be a big deal if the town loses
    government services because of its
    income level…but we will hope whoever is
    in charge of those numbers does a little
    better research.
Which M, M, & M would be best
for…
   The favorite movie of college students?
Which M, M, & M would be best
for…
   The favorite movie of college students?
   Mode: good for non-numerical data items
    and for frequent occurrences.
Which M, M, & M would be best
for…
   The distances students in your class travel
    to school.




                                     Retrieved from
                                     http://openclipart.org/detail/21980/toy-
                                     car-by-nicubunu 4/7/2012
Which M, M, & M would be best
for…
   The distances students in your class travel
    to school.
   Median: one student may live much further
    than everyone else. When an outlier may
    significantly influence the mean, we use
    median.


                                     Retrieved from
                                     http://openclipart.org/detail/21980/toy-
                                     car-by-nicubunu 4/7/2012
Which M, M, & M is best?
   The daily high temperature during a week in
    July.




              Retrieved from
              http://activerain.com/blogsview/1205564/too-hot-to-blog-
              dog-days-of-summer 4/7/2012
Which M, M, & M is best?
   The daily high temperature during a week in
    July.
   Mean: since daily temperatures are not likely to
    have an outlier, mean is best. When data have
    no outlier, use mean.




                                      Retrieved from
                                      http://activerain.com/blogsview/1205564/too-hot-to-blog-
                                      dog-days-of-summer 4/7/2012
Summary
   When data have no outliers, we typically
    use mean.
   When data have outliers, we choose
    median.
   When we are interested in the frequency
    of particular values or we want to
    summarize non-numeric data, we use
    mode.
Measures of Central Tendency
   Go to the Khan Academy site whose web
    address appears on the next slide and work with
    the data to achieve the requested values for the
    mean and median.
   This will help you determine the effect changes
    in data have on those measures.
   Knowing this will help you figure out which M
    works with the data best!
http://www.khanacademy.org/math/statistics/e/exploring_mean_and_median


   Below is a sample problem from the Khan
    Academy website above.

Mmm which one

  • 1.
    The 3 Msof Central Tendency Mean, Median, and Mode Which one should you use? Image retrieved 3/25/2012. From http://m-m.seebyseeing.net/
  • 2.
    Why Does ItMatter?  You live in a town of 1,000 people where the mean income is $25,000 and the median income is about the same. Retrieved from http://en.wikipedia.org/wiki/Damascus,_Virginia 4/7/2012
  • 3.
    Why Does ItMatter?  Now the mean income is $1,023,976, but the median income is still about $25,000. Retrieved from http://genehoyas.com/ 4/7/2012
  • 4.
    The Town ofMillionaires  It is a wonderful story about this whole town of millionaires who are so frugal they live in ordinary homes. What a great example of America!  Why is this misleading?
  • 5.
    Why Does ItMatter?  Most people in this town are still pretty poor.  The mean does not give a good picture of the “typical” resident.  This could be a big deal if the town loses government services because of its income level…but we will hope whoever is in charge of those numbers does a little better research.
  • 6.
    Which M, M,& M would be best for…  The favorite movie of college students?
  • 7.
    Which M, M,& M would be best for…  The favorite movie of college students?  Mode: good for non-numerical data items and for frequent occurrences.
  • 8.
    Which M, M,& M would be best for…  The distances students in your class travel to school. Retrieved from http://openclipart.org/detail/21980/toy- car-by-nicubunu 4/7/2012
  • 9.
    Which M, M,& M would be best for…  The distances students in your class travel to school.  Median: one student may live much further than everyone else. When an outlier may significantly influence the mean, we use median. Retrieved from http://openclipart.org/detail/21980/toy- car-by-nicubunu 4/7/2012
  • 10.
    Which M, M,& M is best?  The daily high temperature during a week in July. Retrieved from http://activerain.com/blogsview/1205564/too-hot-to-blog- dog-days-of-summer 4/7/2012
  • 11.
    Which M, M,& M is best?  The daily high temperature during a week in July.  Mean: since daily temperatures are not likely to have an outlier, mean is best. When data have no outlier, use mean. Retrieved from http://activerain.com/blogsview/1205564/too-hot-to-blog- dog-days-of-summer 4/7/2012
  • 12.
    Summary  When data have no outliers, we typically use mean.  When data have outliers, we choose median.  When we are interested in the frequency of particular values or we want to summarize non-numeric data, we use mode.
  • 13.
    Measures of CentralTendency  Go to the Khan Academy site whose web address appears on the next slide and work with the data to achieve the requested values for the mean and median.  This will help you determine the effect changes in data have on those measures.  Knowing this will help you figure out which M works with the data best!
  • 14.
    http://www.khanacademy.org/math/statistics/e/exploring_mean_and_median  Below is a sample problem from the Khan Academy website above.