Hello,
we are trying to calculate negative keywords in a target corpus, however, were are not sure what we should look for in the output to identify the negative keywords (as they appear not to be clearly marked).
Are the keywords extracted with the ‘show negative keywords’ option flagged all negative, or is there something we are missing?
By running the analysis using Dice as an effect size measure, we do not get any negative results (the lowest keyness effect is 0), this changes if we use MI or M2, in which case we do get positive and negative effect sizes in the output.
Thank you very much,
Valeria, Silvia and Sara
--
You received this message because you are subscribed to the Google Groups "AntConc-Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to antconc+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/antconc/45594679-213c-4e1c-aaa7-6b2747b3dbddn%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/antconc/CAEjwtaYrwPTSNAMS%3DfdQ_PyL5eC8fq1HkvvRr0S_ZQd%2BiNRQhw%40mail.gmail.com.