Stats - Problematic Formulae

[Alchemist4lyf]

Salam everyone.
Alright. So as you all might have noticed in the statistics books, (pakistani ones), the formulae given are usually wrong, or not explained well enough. I'm having a lot of trouble with a particular problem. Median for continuous data.

The formula given is : L + [(N/2 - Cf)/ (f)] + i
F is frequency of median class
N is cummulative frequency
Cf is cummulative frequency of the class before the median class
i is the width of the median class

Now. L is, according to the book, the lower class LIMIT of the median class. But in a past paper book, L is taken as the lower class BOUNDARY. Now there's only 0.5 unit difference between the two, but it renders the whole answer inaccurate. So, could you guys please help me decide, which one is it.

Oh, and if you guys know about any other of such formula, please inform me.

 

[destined007]

Its class boundary. Technically, class limit is also correct because for eg there are classes 0-10, 10-20, 20-30, etc, the class limit and class boundary are the same. So don't be confused on that. For mismatched classes calculate boundaries first.
Have a look at this to have better understanding:
http://web.grcc.cc.mi.us/Pr/tutoring/BA ... UTLINE.PDF

 

[SalmanPakRocks]

Here are more easy ones
Median (group date) = L+h/f(N/2-C.F)
where L= lower boundary
h= class width
f= Frequency
N= total of frequency
C.F= Cummulative Frequency of Previous class boundary.