For the convenience of the reader, in the following table we include a comparative between the original Krippendorff’s notation, the Atlas notation for the □ coefficient and the notation used in this tutorial.Įmpirically, we have discovered that the semantics that the software Atlas applies for computing the coefficients □ □□□□□□/□ □□ □□□□□□ and □□-□/□□-□ are the ones explained in this tutorial. In order to clarify these relations, in this tutorial we always use a notation based on the ones provided by Atlas, but some slightly differences can be appreciated. Now, a variety of notations co-exists in the literature that may confuse the unfamiliar reader with this ICA measure. The development of the different versions of the □ coefficient has taken around fifty years and, during this time, the notation for its several variants has changed. However, it may be difficult at a first sight to identify the prompted results since the notation is not fully consistent between Atlas and some reports in the literature. Krippendorff’s □ in AtlasĪtlas also provides an integrated method for computing the Krippendorff’s □ coefficients. For this reason, in this tutorial we will focus on the application and interpretation of Krippendorff’s □ coefficients. However, they should not be used for drawing any conclusions about the validity of the coding. In any case, Atlas provides us these measures that allow us to glance at the results and to get an idea of the distribution of the codes. Nevertheless, as in the case of the percent agreement, it ignores the agreement by chance so it is not suitable for a rigorous analysis. On the other hand, the Holsti Index is a variation of the percent agreement that can be applied when there are no pre-defined quotations and each coder selects the matter that considers relevant. It is well reported in the literature that simple percent agreement is not a valid ICA measure, since it does not take into account the agreement that the coders can be attained by chance. Simple percent agreement and Holsti Index Coefficients in Atlas for the computation of ICAĪtlas provides three different methods for computing the agreement between coders, namely simple percent agreement, Holsti Index, and Krippendorff’s □ coefficients. Finally, we discuss the calculation, interpretation and validity conclusions that can be drawn from the ICA coefficients provided by Atlas. Then, we describe briefly the protocol for analyzing case study research involving qualitative data analysis in software engineering and we introduce a running example on the topic that will serve as a guide along all the tutorial. First, we describe the different operation methods provided by Atlas for the computation of ICA. We will assume that the reader is familiar with the general operation of Atlas and focus on the computation and evaluation of the different ICA coefficients calculated by Atlas. See the W3C color names for a full list of color names or look for (W3C) in the table below.In this tutorial, we describe how to use the ICA utilities provided by Atlas.ti v9 (from now on, shortened as Atlas) as well as a guide for interpreting of the obtained results. For example, in HTML tags and CSS that use color codes, you could use "red" instead of "#FF0000". With these colors, you can also use the color name.
0 Comments
Leave a Reply. |