Analysing and Reporting ordinal logistic regression in R was wondering if anyone could help. Sorry if this question has been asked before - cannot find it anywhere 
I am running multiple ordinal logistic regressions using R, using the "ordinal" package, and CLM function. It produces and output table like this one:

Output wihtout the interaction:

From this, and a few additional follow up tests, this is how I have interpreted this: 
There was no main effects of timing on juror confidence in their verdicts (p= 0.740) (before, M=6.31 , SD= 2.08) (after, M= 6.45 , SD= 2.15) or complexity of the expert testimony (p= 0.329) (standard: m=6.10, SD=2.22) (scientific: m=6.70, SD =2.01). However, there was an interaction between complexity and timing (p= 0.018). Further analysis found that confidence would go up by 1.306 if the expert testimony was presented before the eyewitness account, regardless of the complexity of the testimony.
However, I am unsure how to report this results for a statistical paper, using this output. For example, I do not have any degrees of freedom? 
Any help would be much appreciated.
 A: My suggestions:


*

*Always interpret the model without interactions first. Then describe the effect of the interaction. For instance, the stopping distance of a car interacts with the weight of the car with the speed it's traveling. If you fit an interaction between these, the "main effects" are 0, but surely you can't report something silly like "speed and weight were not associated with stopping distance."

*"Main effects" and "interactions" are jargon for a non-statistical audience.

*Don't "accept the null hypothesis" (you write "there were no main effects"). Your study is probably underpowered. 

*Use a contextually correct description of the predictors. "After/before" means in relation to...? I'm assuming the delivery of the verdict or sentencing?

*Report the actual coefficients and their 95% CIs instead of the qualitative p-value. 

*Exponentiate the "Estimate" column for a cumulative odds ratio. For instance, "afterbef" = exp(-1.251) $\approx$ 0.29. So you can say, "The odds ratio for a higher juror confidence was 0.87 before sentencing was delivered."
