What's the reason to take log of likelihood function to find maxima?

My textbook does not explain this.

Can anyone explain why?

Thank you.

Results 1 to 2 of 2

- Mar 26th 2009, 07:58 AM #1

- Joined
- Nov 2008
- Posts
- 53

- Mar 26th 2009, 02:37 PM #2
It's like a slide ruler.

It makes the calculations easier.

Log's change products into sums, so your differentiation and algebra is easier now.

Log is an increasing function so the max of L(x) occurs at the same point as the max of ln(L(x)).

In class I did the binomial with out th elog and it works just fine.

BUT most students don't remember how to differentiate $\displaystyle a^x$.