2

I have a regression that is of the following form:

$$\text{salespc} = \beta_1 + \beta_2 \text{realprice} + \beta_3 \text{realincpc} + e_i $$

My output from R is as follows:

           Coef     S.E.   t     Pr(>|t|)
Intercept 253.5044 9.2183 27.50 <0.0001 
realprice -18.4374 2.6315 -7.01 <0.0001 
realincpc  -0.0035 0.0004 -8.45 <0.0001

I'm supposed to find price elasticity of demand and its standard error. I think that the elasticity is $$\beta_2 \frac{\text{realprice}}{\text{salespc}}$$ at given values of $\text{realprice}$ and $\text{realincpc}$. But I'm not sure how to find the SE. I know this isn't an econ forum but I thought it may be my best shot. Thanks.

I'm supposed to do a similar question with

$$\log(\text{salespc}) = \gamma_1 + \gamma_2 \log(\text{realprice}) + \gamma_3 \log(\text{realincpc}) + e_i $$

but I think once I get the first part I can do the second.

tshauck
  • 479
  • 2
  • 4
  • 11
  • 4
    The first question is tough: one tends to think of [parametric bootstrapping](http://stats.stackexchange.com/questions/10727/how-to-get-standard-error-of-a-function-delta-method-vs-simulation/10729#10729) or the [delta method](http://en.wikipedia.org/wiki/Delta_method). Are these the sorts of calculations that would be expected of you? The second, in comparison, is [almost trivial](http://stats.stackexchange.com/q/9913), so consider addressing it first. – whuber Nov 03 '11 at 05:09

0 Answers0