Cleaning Data

My studies of rotifer demography and evolution would not be possible without extensive student participation. I need three people working together for four hours every day just to get the census done. Most of the time I have six or seven students on my team. Only one student, LZ, has been working with me the whole time. She started last year as an Undergraduate Research Apprentice and this spring and summer has been working on her Senior Thesis. She decided, quite independently, that she was going to study which demographic factors affect the probability of a rotifer reproducing sexually, rather than asexually. She is a fabulous student, has become a talented researcher, and I hate to see her go.
A rough draft of her thesis is due this Friday, and as we have been trying to analyze our data, we have instead been spending most of our time fixing typos and errors in the data. We've been finding mistakes ranging from recording a rotifer as being in plate 1111 instead of 111 to the same information recorded quite differently in two different places. LZ and I spent ten hours yesterday going through our data line by line and cleaning up errors. Some sections were perfect, others awful, and we couldn't help but wonder which students had taken the clean data and which had gotten sloppy, lazy or confused.
Doing this has brought into sharp relief both the good and the bad of having such student-powered research. Relying on students is extremely useful and motivating and inspiring, but can also be frustrating and time-munching. Which way it goes depends enormously on the students I choose. I am very shortly going to be needing to find another couple of assistants, and will have this experience in mind.