2

I fitted an ordinal logistic regression but I'm unable to interpret the coefficients. Can anyone assist in this regard? Here is the output generated: 

Call:
polr(formula = factor(grade) ~ factor(Month) + Day, data = myData, 
    Hess = TRUE)

Coefficients:
                  Value Std. Error t value
factor(Month)4 1.405114    0.51547  2.7259
Day            0.007672    0.01944  0.3947

Intercepts:
    Value   Std. Error t value
1|2 -0.6785  0.7019    -0.9667
2|3  1.6767  0.7162     2.3412

Residual Deviance: 333.602 
AIC: 341.602

The grade is factored:

1 = good
2 = very good
3 = excellent

Month is factored:

3 = March
4 = April

The grade is the response while month and day are my explanatory variables.

mdewey
  • 16,541
  • 22
  • 30
  • 57
liz
  • 21
  • 3
  • 1
    This might help you: [Negative coefficient in ordered logistic regression](http://stats.stackexchange.com/q/38087/). – gung - Reinstate Monica Oct 21 '15 at 00:03
  • I don't get you. can you explain please? – liz Oct 21 '15 at 00:43
  • There's a lot of information at the link that can help you interpret OLR models. Can you be more specific about what you need to know? Eg, do you know what a standard error is? – gung - Reinstate Monica Oct 21 '15 at 01:07
  • Yes, I know what standard error is which is the standard deviation divided by the square root of n. My question is "How do i interpret the coefficient values for Day and Month with respect to the response variable which in this case is the grade? – liz Oct 21 '15 at 02:04
  • the standard error you reported in your question is **not** the standard deviation divivded by the square root of n. – Maarten Buis Oct 21 '15 at 08:54
  • The standard error there is the error of the regression model right? – liz Oct 21 '15 at 13:24

0 Answers0