Reliability coefficients are designed to measure the degree to which data warrant serious considerations in subsequent analyses. They call for reliability data that consist of several independent replications of [Page 725]the process by which phenomena of scholarly interest are converted into analyzable data. This process may be embodied in mechanical measuring devices, performed during standardized tests, or enacted by human coders who are formally instructed to interpret selected phenomena, distinguish among the units that are relevant to a research question, and assign well-defined categories or numerical values to them. Written coding instructions are provided to standardize the process across all coders employed, and it is these coding instructions that need to be reproducible elsewhere as well for the resulting data to be considered ...
Looks like you do not have access to this content.