Content Analysis in SPSS and KALPHA: Procedure for a Reliable Quantitative Analysis with Cohen’s Kappa and Krippendorff’s Alpha

Abstract

During the last decades, content analysis has emerged as one of the main research techniques in communication and journalism studies. An extensive literature has described, in great detail, the procedures for its systematic and correct implementation, which has led to more empirically robust and precise research. This advance in the application of methodological protocols has not been transferred, in such a gradual manner, to the computations and statistical tests to establish reliability between coders in publications in Spanish, despite its central importance when it comes to offering reproducible scientific evidence. This has been due, among other reasons, to the need to calculate the indicators in statistical programs, which has hindered their widespread adoption. In this work we illustrate, using SPSS and the KALPHA macro, the computation of two statistical tests to measure the agreement between coders: Cohen’s Kappa (κ) and Krippendorff’s Alpha (a). The objective is, therefore, to offer and detail to the Hispanic academic community a protocol to generalize the use of computations that allow researchers without advanced statistical knowledge to report reliable and standard results in the field.