test - xsmle and interpretation

[Marcos Paucar]

Hey listers,

I am using the "xsmle" command to run fixed effects models on panel data. I am particularly interested in computing the indirect effects of my covarietes so I added the "effects" option. All my variables are in logarithms - dependent ln(x) and independent ln(y)-, but I'm not sure about the interpretation of the coefficients the "effects" option retrieves (LR Direct, Indirect and Total Effects).

For example, if LR Direct retrieves a coefficient of 0.312331. Should I go: a 1% change v1 is correlated with a 0.312331% change in y or a 31.2331% change in y.

Best regards,

Marcos

[Marcos Paucar]

Hey listers,

I am using the "xsmle" command to run fixed effects models on panel data. I am particularly interested in computing the indirect effects of my covarietes so I added the "effects" option. All my variables are in logarithms - dependent ln(x) and independent ln(y)-, but I'm not sure about the interpretation of the coefficients the "effects" option retrieves (LR Direct, LR Indirect and LR Total Effects). In the example I list below my coefficient is -0.632061, it would be correct to interpret then: a change in 1% in prec is correlated with a change in -0.0632061 in y?

      ------------------+----------------------------------------------------------------
      LR_Indirect       |
                ln_prec |  -.0632061   .0473598    -1.33   0.182    -.1560296    .0296173

Best regards,

Marcos