Determining which estimator to use (stats)

[jasper90]

Consider a uniform distribution on the interval 0≤ X ≤ θ. We are interested in estimated θ from a random sample of draws for the PDF. Two potential estimators are:

θ1 = (2/n) Ʃ Yi

and

θ2 = (n/θ)(y/θ)^(n-1)

which estimator would you prefer and why? What statistical properties did you use to decide?

Uniform distribution f(x)= 1/(B-A) for alpha < X < Beta

We use method of moments estimator and max likelihood estimator

 

[jasper90]

anyone?

 

[Ray Vickson]

anyone?

Sure. Show us what you have done so far. Those are the Forum rules, and are also the means of mastering the material and passing the course.

RGV

 

[jasper90]

I really don't know. Every problem we have done in class was done the reverse way.

Like, I know for max likelihood estimator, we take the Ln of f(x) and then derive it. Then we set to 0 and solve for our estimator. But I have never had to choose one. I tried reversing the process, but it is definitely wrong.

I know I would be replacing B with θ1 and θ2.