As in Quality Assurance.

As far as I know, I have no outright mistakes in published papers. I made mistakes but either posted an erratum or fixed them before the paper was published. I think there are research problems which we tackled multiple times and ended up doing a better job, relative to ourselves. With the gift of hindsight I think it might be helpful to you if I point this out.

• If you want to read about robots and peer pressure, our paper in the Journal of Language and Social Psychology has the most comprehensive treatment.
• If you want to catch up on social language learning, our paper in Language Learning has a more streamlined design and fewer forking paths, part way replicating and legitimising the earlier work as well, which is good.
• If you’re interested in algorithmic learning models of the English past tense, look at our Language paper, which discusses this in detail before going into convergence effects. We did try every conceivable angle at defining verb classes and tuning the models here and I’m confident in the robustness of our results.