# How can I interpret binary logistic table?

I´m beginner with SPSS and I have on problem on interpreting binary logistic results. So I have this table:

Variables in the Equation

                          B      S.E.  Wald     df   Sig.   Exp(B)
Step 1ais_first_timer   -1.582  .129  149.593   1   .000     .206
trip_duration       -.301   .047  41.114    1   .000     .740
back_to_country      .583   .130  20.030    1   .000    1.791
back_to_event       1.275   .255  25.059    1   .000    3.578
visiting_places     -.179   .120   2.203    1   .138     .836
Constant            1.072   .145  54.862    1   .000    2.922


So my problem is how can I write an analysis based on that? I know I have to watch Sig and B. But what do they show me? My main problems is what does B shows me? For example if my B is -1.582 and sig is 0.000 and my dependent variable is nationality and covariates are, as you can see in the table, first timer, trip duration etc. Can I say watching this table, for example English people (0) are more likely to be first time visitors than Americans (1), and this is statistically important? Or can i say that English people have longer vacation than Americans Or i can`t say that?

I hope somebody can help me :)

• If I've well understood your question the interpretation you are looking for has nothing to do with your $\beta$. You must first clearly define your dependent variable, that is, do you want to predict the probability that each nationality have of being first time visitors or do you want to predict the probability of which nationality will have longer vacation periods. Then, you can proceed with calculation and interpretation of your log odds.
– Vara
May 8, 2013 at 6:33

## 1 Answer

Your dependent variable is nationality? I'm not really sure that makes sense. Your DV should be something you want to explain/predict/describe using the covariates. Sounds like you want a model where vacation length is the DV and is explained by nationality.

More generally, you can use results from this table to interpret direction and statistical significance of coefficients, but not to speak about their absolute size or size relative to the effects of other covariates. You'll have to do some addition analyses (e.g., showing predicted probabilities for different combinations of covariate values) to say something more specific about size of these relationships.