problems in models 7.4.2 example

[lbluefishl]

I believe I found a few discrepancies in this section.

First, the model uses delta_t, but later the explanation uses gamma_t.

Second, I believe the labels may have been put in the wrong place on the table. I'm very new to learning these models, so I may be wrong.

Specifically, the text mentions: "The intercept (control condition estimate) ... corresponds to a probability of 0.69... that estimate makes sense: 0.69 seems close to the average for the control condition" However, I believe the value 0.69 reflects the average for the experimental condition, not the control.

Similarly, it says, "The key experimental condition estimate is -2.26... which corresponds to a probability of 0.19... this estimate corresponds to the average value for the experimental condition." I believe that 0.19 corresponds to the control condition, not the experimental.

I performed the same analysis in R using the code provided. My coefficient for conditionControl was -2.26. Which should suggest that being in the control lowers the probability of success on the trial, which makes sense and follows the data. But the table 7.2 puts -2.26 coefficient for the experimental condition.

[mcfrank]

Thanks so much for catching this!