# statistics - median

• Aug 12th 2006, 05:10 PM
bennyya
statistics - median
How do I work out how to find the new sample mean when you have a sample size of 20, the mean is 12, the data is bimodal with modes of 10 and 17 and the median is 14. After this has all been calculated a new item of data has been found, being of value 17, which needs to be added to the sample. So now the sample size is 21 and I was able to work out the new mean is 12.24 by using the equation - but I can't work out what the new median would be.
• Aug 13th 2006, 12:29 AM
Glaysher
How many forums have you posted this in?

Sample size of 21 would make the median the 11th observation

If so, put it into order and stick the new 17 in the appropriate place
• Aug 13th 2006, 02:10 AM
CaptainBlack
Quote:

Originally Posted by bennyya
How do I work out how to find the new sample mean when you have a sample size of 20, the mean is 12, the data is bimodal with modes of 10 and 17 and the median is 14. After this has all been calculated a new item of data has been found, being of value 17, which needs to be added to the sample. So now the sample size is 21 and I was able to work out the new mean is 12.24 by using the equation - but I can't work out what the new median would be.

Half the observations should be above and half below the median.

Without knowing some values about the old median you can't work out
the new median.

Here the old median is between the 10th and 11th largest sample values,
but that 11th largest sample could be as small as 14, and presumably as
large as 17. Also there could be a number of sample values between 14
and 17.

RonL
• Aug 13th 2006, 07:55 AM
bennyya
Without the raw data it is not possible to find the median?
• Aug 13th 2006, 08:24 AM
CaptainBlack
Quote:

Originally Posted by bennyya
Without the raw data it is not possible to find the median?

Without sufficient information to allow you to deduce what are the central-ish
data values; yes you can't find the new median.

RonL
• Aug 13th 2006, 06:57 PM
bennyya