This is anovice question, but I have been trying to understand it the whole day. I have got three text types and three scores representing readability indices for each text type. How can I compute a correlation to check whether there is a significant difference between the scores depending on the text? I have tried to do but didn't succeed, I got messages of this sort: one variable is constant (I cannot replicate it here anymore).
In order to attain a significant difference you need variability. In this example, if i am correct, you just have one instance of the index for each text type. I would interpret that as simply one is more readable than the other based on the single observations for each. You can’t say much more.