I'm working with logit model and OR to interpret results.
After run regress one of the models specifications, the table with raw result shows:
Code:
. logit sick tp t2m ro, or ------------------------------------------------------------------------------ sick | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- tp | 1.000355 .0000864 4.10 0.000 1.000185 1.000524 t2m | 1.000981 .000862 1.14 0.255 .9992928 1.002672 ro | 1.000226 .0001402 1.61 0.106 .9999516 1.000501 _cons | .1293345 .0021047 -125.69 0.000 .1252744 .1335262 ------------------------------------------------------------------------------
I use "esttab" to import table to Latex, then the option "eform" let obtain the exponentiated coefficients, but the table presents odds ratio with 3 digits and they can be interpreted as the variables have no effect. (The table below shows a variety of model specfications)
Code:
.# delimit ; . esttab m1b_3 m1b_2a m1b_2b m1b_2c, eform se(%9.4f) > mtitles("tp _t2m_ro" "tp_t2m" "tp_ro" "t2m_ro") nostar; .# delimit cr ---------------------------------------------------------------- (1) (2) (3) (4) tp _t2m_ro tp_t2m tp_ro t2m_ro ---------------------------------------------------------------- sick tp 1.000 1.000 1.000 (0.0001) (0.0001) (0.0001) t2m 1.001 1.001 1.001 (0.0009) (0.0009) (0.0009) ro 1.000 1.000 1.001 (0.0001) (0.0001) (0.0001) ---------------------------------------------------------------- N 310623 310623 310623 310623 ---------------------------------------------------------------- Exponentiated coefficients; Standard errors in parentheses
Any suggestion to obtain a table result (e.g. through "esttab") that can be imported into Latex which allow over 3 digit format for odds ratios?
0 Response to Command "esttab": Adding more digits to odds ratio coefficients
Post a Comment