I can just answer this from a biologist student's view. I had some boring statistic courses, but one was really great: that one where we first had to take samples by our own and count them. Then all data got meaning and we were eager to learn more. We then calculated things by hand first. This was the best way to internalise all necessary steps. We did it with the computer later, but at that time we already had a feeling for the data and computation steps.
Most of this can probably also be done with social sciences. Of course, you can't take samples of animals and count them. But you could do a short census about something of interest within the course and with all paticipants. Then you write down the results and classify them, and so on.