Consider the following regression results:
Call:
lm(formula = log(y) ~ log(x), data = d)
Residuals:
Min 1Q Median 3Q Max
-6.6119 -1.4477 0.1735 1.5365 4.8160
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.1264 0.2520 0.501 0.618
log(x) 0.2870 0.2279 1.259 0.212
Residual standard error: 2.251 on 79 degrees of freedom
Multiple R-squared: 0.01967, Adjusted R-squared: 0.007263
F-statistic: 1.585 on 1 and 79 DF, p-value: 0.2117
Describe how the response y
depends on the regressor x
.
The presented results describe a log-log regression.
The mean of the response y
increases with increasing x
.
If x
increases by 1 percent then a change of y
by about 0.29 percent can be expected.
However, the effect of x
is not significant at the 5 percent level.