Centre (middle) of a skewed histogram

[Cel126]

Hey Guys,
I was just going over some statistics work and I wasn't 100% sure on how to find the centre (middle) of a a skewed histogram (either positive or neg).  I was told by someone that you have to estimate, or is there a particular way?
thanks 
 

[REBORN]

Graphically? It's just a rough estimate. The centre will be closer to the skewed side!

[Cel126]

Graphically? It's just a rough estimate. The centre will be closer to the skewed side!

Yes Graphically, okay thanks for that !

[Daenerys Targaryen]

I don't exactly remember, but I'm pretty sure I use to add up all of the entries. Divide by how many there where. That would give me the middle number of entry. Then I'd count that many entries from left to right (or which ever way) and what ever value that entry had would be the centre.

[Stick]

Hey Guys,
I was just going over some statistics work and I wasn't 100% sure on how to find the centre (middle) of a a skewed histogram (either positive or neg).  I was told by someone that you have to estimate, or is there a particular way?
thanks 
 

Use the median, if it's possible.

[Daenerys Targaryen]

Right - that's what I was explaining. Lol.

[REBORN]

Yeah...the median is a numerical measure. My answer is graphical.

If the y-axis is in percentages, the centre is given by the bar at 50% (median by definition)

[Daenerys Targaryen]

Also note that with symmetrically shaped graphs, mean is the best measure of the middle
However when it's skewed median is the best.

[Yacoubb]

Find the Median of the distribution using the (n+1/2)th position rule. After locating the position of the median, count through the frequencies until you get to an approximate of that; that would be your median interval

[Cel126]

Thanks for the help everyone