1. What kind of information did it give you about the usability of the prototype?

I think both techniques helped gather information about the emotions/feelings/thoughts risen in the user while performning the tasks or interacting with the protoype. Moreover, they help rise positive and (specially) negative features and issues of the prototype. The user behavior was more useful to gather non-verbal information because you focused on the physical reactions of the user, while the think-aloud helped the observer understand better which actions/steps of the task  and which features where the ones that caused those physical reactions. By complementing both techniques, the observer can build a good picture of when, why and how the user felt with the prototype. It gave clues of which aspects annoyes the user, which ones enable it to perform the task, which features were memorable, useful or confusing, etc…so both techniques were very helpful to help the observer realize how the prototype could be improved based on the user experience.

2. What aspects of the techniques worked well or were frustrating?

In the user-observation it was difficult to concentrate on both the verbal and non-verbal behaviors of the user. In the think-aloud sometimes it was difficult to process the information, write it down and concentrate on what the user was saying next (specially collecting the data of the interface part/location was frustrating). I think that having at least 2 observers (one that focuses mainly on the verbal and one that focuses on the non-verbal or physical expressions can be useful). Even more,  recording the evaluation could help the observers rise more issues that could have gone unnoticed in the real-time observation (in the tutorial we were not able to record because we could not record with the camera, observe and write down at the same time, we were only 3 per group).