ABSTRACT
This chapter discusses the process of collecting and processing data. The only remedies for the human errors in the data-collection process are eternal vigilance, spot rechecking, checking the data against researcher's intuition, and more checking. Experiments usually produce nice, clean, orderly data, because what comes out is a product of what is put in and the researcher has considerable control of what goes in. If the researcher cannot estimate the datum or if the datum represents too large a part of all the data in the study, then they must abandon the study. The chapter discusses discuss three difficult imputation problems: dividing combined data into parts, allocating parts of a whole, and evaluating worth in the absence of a standard. When there is no empirical source for data but the data are crucially needed, the researcher must somehow make the most sensible estimate.
