Boxplot GLM with binomial errors - interpret summary My apologies, I am not even sure how to phrase this question clearly.
I am essential trying to get meaningful statistics to present tomorrow.
All I have managed is this. However, I am very confused how to phrase the descriptive statistics or whereabouts they even are. 
I am trying to come up with something like: The wasps are negatively affected by the chemical treatment. The control significantly differs from 2ppb and 10ppb. I am after F and p values and maybe n values I think?



 A: Factor Significance
For interpretation, you want to perform an ANOVA on the fit model, rather than simply interpreting the significance of the coefficients, which is what summary is showing you.  The F test is not appropriate for binomial response family, instead you use Chi-squared.  For example,
> anova(imdglm, test = "Chisq")
Analysis of Deviance Table

Model: binomial, link: logit

Response: cbind(Correct, Incorrect)

Terms added sequentially (first to last)


          Df Deviance Resid. Df Resid. Dev  Pr(>Chi)    
NULL                         35     79.636              
Treatment  2   24.123        33     55.514 5.779e-06 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

More Appropriate Levels
You may want to reconsider the way you are currently processing the treatments.  Your code treats them as categorical, whereas they are dosage levels, which are cardinal.  A more appropriate approach would be to either treat the treatment levels as ordered:
imd$Treatment <- factor(imd$Treatment, levels = c("Control", "2bbp", "10bbp",
                        ordered = TRUE))

or introduce a new column for the dosage amount (0, 2, or 10), and use that as the explanatory variable in the glm.
