# 3SLS method

#### mimilafleur

##### New Member
Dear members:

I am reading a report on an econometric analysis that has been carried out for fish industry.

I am not familiar with 3SLS. I would like to know how to interpret the results and how to compute the confidence intervals.

The only thing I got from the report is the model itself (y = a + bx) the estimated coefficients (a and b), STD errors, t-value and one sided p-values.

I do not know the size of the sample (N).

One of the models is: ln(Y) = -2.290 + 0.436 * X + 0.061 * X-1 + 0.360 * Z

where Z is a dummy variable equaling 1 between 2000 and 2003 and X-1 is X lagged one year.

There are actually 12 models in total and 67 parameters.

I would like to know how to compute de confidence interval for 0.436 (for example) if I had the total size.
Does the coefficient of X, β1, (0.436) follow a t-student statistic with N-67-1 parameters, or a t-student statistic with N-3-1 parameters?

β1 ε [0.436± t-student (#parameters) * STD error] is this right?

I have another question about elasticity. Is there a website where I could look up the theory behind elasticity?

Elasticity of the X variable = β1* average(X) = derivative of Y * average(X)/average(Y)
It has been determined that a 1% change in X would lead to 1.39% of Y.

Thanks for any hint.