一般在经济类的实证分析中,分统计上的显著性与经济上的显著性,往往在经济上的显著性的判断上比较迷茫,标准是什么?
如下例:
In Panel A, in each of the four audit fee regressions (1), the coefficient on the management forecast attribute is positive
and significant at less than the 1% level. The coefficients on FREQUENCY, SPECIFICITY and HORIZON are 0.378, 1.684 and
0.731 with t-statistics of 49.85, 20.33 and 22.55, respectively.
These results also are economically significant. For instance, the coefficient of 0.378 on FREQUENCY implies that a one-standard deviation increase in the number of forecasts is associated with a 4.6% increase in audit fees relative to the mean. Similarly, one standard deviation increases in SPECIFICITY, HORIZON and ACCURACY are each associated with 8.4%,4.9% and 2.5% increase in audit fees, respectively.
这里的自变量回归系数分别为0.378、1.684、0.731,算出来的经济显著影响,对审计收费的影响分别为4.6%、8.4%、2.5%。
Mean | Std dev | |
FREQUENCY | 0.66 | 1.55 |
SPECIFICITY | 2.89 | 0.64 |
HORIZON | 4.79 | 0.86 |
ln-fees | 12.75 | 1.52 |
这种经过标准化处理的经济显著性判断是不是按以下方式
FREQUENCY的Beta=1.55/1.52*0.1275=4.9%,
specificity:=0.64/1.52*0.1275=9%
但是这与上文中计算的4.6%与8.4%都有一定的距离,想知道为什么?如何计算的?