Keywords are key when it concerns blogging. I explained my initial ideas on how to approach keywords for your blog in a previous post. That post gives some basic ideas on how to use them to get and keep an overview of the competition in your blogging sphere.
In this current post I am going a bit more in depth and taking the concept of “Keyness” and “Dispersion” into consideration when it concerns keywords in the context of blogging.
Important: keyness and dispersion are calculated between a specific text in relation with a set of other texts. For example, a blogpost compared to all blog post of a blog. But of course more variants of comparing are possible, hence this blogpost.
Concepts of Linguistic Research: keyness
Keyness is a concept used in linguistic research. Keyness is expressed in a number. In an informal way, you could say: It tells you how key your keyword really is. If the keyness number is positive, a keyword that is measured by the keyness occurs more often in a text in comparison with a reference text then can be expected by chance. If the keyness is a negative number, the opposite is true: the keyword occurs less often in a text in comparison with a reference text then can be expected by chance.
The keyness can be calculated in different ways, with the classic chi-square test or with the log-likelihood test. The details for this calculations and how to apply them to texts can be found in the book Statistical Methods in Language and Linguistic Research. The book gives very nice examples of statistical methods that can easily be translated into the blogging sphere.
If we calculate the Keyness with the Log-Likelihood method (LL), the following holds.
The higher the the LL-value, the more significant the difference.
Concepts of Linguistic Research: dispersion
Where keyness can reveal a lot of interesting information for your blogging strategy, combined with measuring dispersion of keywords between blogs or blog post the idea of using a statistical approach for analyzing your blog competition becomes even more valuable. Calculation dispersion can reveal how keywords with a high keyness are distributed among blog posts. This you can use to analyze if the keyness that is found is more or less relevant for your blog.
The statistical methods very briefly described in this post are just two simple but powerful ways to get a better overview on what is happening in the field your blog is operating in. When combined with a couple of scripts that collect and do the statistical analysis as described above you have a nice tool that can help you to create a more relevant blog.
If you want to participate, give me a comment, and I will invite you as a beta tester as soon as this becomes relevant. Perhaps you want to know how the keyness is of your keywords? The question is, in relation to what? If you have that answer let me know, and we can look at it together.