Regression models: Causal relationship vs. a good explanitory model? (1 Viewer)

[moll.]

Doing the ECON2206 Introductory Econometrics project on regression models and I'm really confused by two question which seem to be asking the exact same thing:

(6) Does the model in (1) provide a good explanation of median house prices ? Explain.
(7) Does the model in (1) capture a causal relationship between the air pollution and median house prices across suburbs ? Explain your reasoning.

What's the difference between the two? I've been staring at my screen for 15 minutes trying to work it out, to no obvious avail.

Just fyi, it's a simple regression model between house prices and air pollution in certain suburbs.

Thanks.

 

[22944]

While I feel reluctant to help anyone who thinks a video of a cat being kicked into a wall constitutes humour, it is my duty to help others, who like myself, are subject to the horror that is econometrics.

Anyway, so from what i’ve figured so far, Q6 is hinting towards R squared while Q7 is more to do with ZCM etc.

Tell me what you think cos i’m not 100% sure either.

p.s. i’m not actually pissed off about the vid or some crazy animal activist or anything, i don’t even like pets but i just think it’s a bit much.

That is all.

 

[moll.]

I got bored of waiting for a reply so I just answered 6) as a question about R-square and statistical significance and answered 7) as economic or practical significance and whether d(nox) will genuinely result in d(price).

EDIT: d = delta

 

[moll.]

So I asked my tutor about 6 and 7 and he said that really it was about statistical significance versus economic significance. For 6 you use r-square to prove explanatory power. For 7 you use t-stat to prove beta1 =/= 0 or 1 and hence is economically significant.
Just fyi.

 

[22944]

Cool, thanks for that.

 

[whoisurdaddy]

what are u guys getting for the stardard error of prediction for q16?