st: Logistic regression with a log transformed variable How to determine‏ economic significance

[Robert Teunissen <r.a.p.teunissen@tilburguniversity.edu>]

Hi all,

I am using a logit regression model 1succes 0failure with six
continuous independent variables and two log transformed size
variables, total assets and total deposits.

My question is how to interpret the results and measure the economic
impact, magnitude, size? The normal step in interpretation I take is
as follows, I firstly run a stata test on summary statistics (sum),
then a logistic regression (logit) and than the marginal effects for
the economic significance (mfx).

These results can be explained by sentences as: a one standard
deviation increase in X (IV) increases the probability of succes by X
percent. The results are derived by multiplying the dy/dx of the
marginal effects by the Standard deviation to get this percentage
economic impact.

So far so good, but how do I interpret the results with the two
ln(size) variables? And also, how to report them? To clarify my
problem I will show you a part of the results:

Summary statistics
ln total assets    mean-1.469624    stdD2.796627

Logistic results
ln total assets    coeff-0.2381355    stERR0.133577    Z-1.78    PZ0.075

Logistic results with odds ratio
ln total assets    oddsR0.7880959    stERR0.105275    Z-1.78    PZ0.075

Marginal effects
ln total assets    dydx-0.2381355    stERR0.0165    Z-1.78    PZ0.075
  X-0.7886

In a normal situation with independent variables I would multiply
ln(total assets) of the marginal effects -0.2381355 by the standard
deviation of the summary statistics 2.796627 which results in an
economic impact, after a one standard deviation increase, of -8.22
percent.

Although this holds for the other non log variables, this does not
sound correct for this ln variable. But how do I interpret the
economic significance than with these ln variables? And how do I
present them in text? like, a one standard deviation increase,
increases the probability. Or something like that?


I look very much forward to tour responses and thanks in advance! best regards,


Peter Teunissen
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/