In basic terms, when thinking about the consistency of the data collected the researcher needs to consider if, for example, the same question asked of the same person in similar circumstances, will produce the same answer.

If it doesn't then the researcher would have reason to doubt the consistency of their data...