Logistic Regerssion

[tg128]

hi

I am trying to construct a logistic regression model. However, when I typed in the equation in the Equation Estimation Window, I have following message 'unmatched parenthesis'.

I selected the Method as Binary and my equation looks like below

offended_12m_after c ((1+e^(-(d_0_time + d_6_to_10_times + d_less_than_5_times + d_more_than_10_times + number_of_offence_before + offended_12m_before + onset_change + total_onset_pre)))^(-1)

Could you please tell me what method I should be using for logistic regression and maybe some ticks about how to express the equation?

Thanks a lot!

[trubador]

The number of open parenthesis "(" should equal to the number of close parenthesis ")". In your case, it seems one close parenthesis is missing and that's why you are getting the error message. Besides, exponential function is expressed as "exp" not "e" in EViews. However, if your dependent variable is already binary, then you can simply enter the following into your equation editor and select the Binary estimation method as "Logit":

Code: Select all

offended_12m_after c d_0_time d_6_to_10_times d_less_than_5_times d_more_than_10_times number_of_offence_before offended_12m_before onset_change total_onset_pre

[tg128]

The number of open parenthesis "(" should equal to the number of close parenthesis ")". In your case, it seems one close parenthesis is missing and that's why you are getting the error message. Besides, exponential function is expressed as "exp" not "e" in EViews. However, if your dependent variable is already binary, then you can simply enter the following into your equation editor and select the Binary estimation method as "Logit":

Code: Select all

offended_12m_after c d_0_time d_6_to_10_times d_less_than_5_times d_more_than_10_times number_of_offence_before offended_12m_before onset_change total_onset_pre

hey, first of all, thanks!

I have tried the equation u posted, and I have different message this time, see below

'Quasi-complete seperation: D_0_time>=0 (quasi) perfectly predicts binary response failure'

I suppose it's probably nothing to do with your equation, however, I still don't understand what it means...

Hope someone can help answering me. Thanks!!

[EViews Glenn]

EViews 6 User's Guide II, p. 215-216.