Without knowing whether the individual regression coefficients are significantly different from zero, or how you coded the DV (I'm presuming that 1 is the category of discontinuing exclusive breastfeeding and 0 is not, and I'm presuming that the two IVs are categorical), here's what the model suggests to me:
1. Presence of superior support increases the odds of a case being in the target category by exp(0.850) = 2.34 times (for a given value on the sufficient duration variable).
2. Engaging in sufficient duration to express breast milk increases the odds of the case being in the target category by exp(0.802) = 2.22 times (for a given value on the superior support variable).
David Morse Hi, thank you for answering. I would like to know about the constant actually (-4.259). What does the negative implies on, and what does the constant represents.?
The constant is simply a centering/starting value for the log-odds of the event. What it implies is that, for cases having a value of zero on both IVs, the odds of a case being in the target category are very low, exp(-4.259) = 0.01.
To be fair to the model it is much easier if you think of classifying a set of data into two groups rather than the odds ratio form of the model.. that was developed for assessment of risk of disease. Mostly we don't do that so it's enough to know when the IVs predict into one group or the other. Lachlan's Biostatistical Methods has all the gorey details of the risk approach. Medical reaserch wants to know that. MOST of us wish to know grouping only. HTH David Booth BTW to find the group assignment see if your observation predicts into 0 or 1.As always graphs can help..