Viewing 2 reply threads
  • Author
    Posts
    • #35378
      Anastasios Asimakopoulos
      Keymaster
        @anastasios

        If you have any feedback or comments at the end of the fourth unit, we’d appreciate them. These can be comments about the content itself or about the site and the ease of use. However, please don’t feel under obligation to add anything here if time doesn’t permit.

      • #37224
        Rhian Webb
        Participant
          @rmwebb

          @anastasios Hi!  I have been thinking about the terminology I use in class with my students when teaching using a corpus-based language approach. The following words are still a little unclear to me and I was wondering if you have a set of definitions that you use with your learners for: frequency, common and typical. (I know I could do WSD on [common] and [typical] and I might just do that now!)

          Also, some course books introduce terminology, such as, strong collocation or weak collocation. I was thinking about what these terms mean and how we, as teachers, might help students grasp the statistical significance of ‘strong’ and ‘weak’ as I imagine it is connected to statistics like ‘strong/weak’ correlation. My Chinese students could almost definitely get their heads around the numbers which pop up when we do corpus queries.

          Any thoughts or guidance on where to go to get/find some easy-to-grasp definitions for these terms would be much appreciated. :mail:

           

        • #37244
          Anastasios Asimakopoulos
          Keymaster
            @anastasios

            Hello @rmwebb. I tend to avoid the term ‘common’ because it has several meanings (happening often; being shared; ordinary). I don’t think there’s a set of definitions out there but I can help a bit. The terms frequent/frequency are usually easy for students to understand because they refer to the actual occurrences of a word e.g. x occurs 5 times in the corpus, y occurs 20 times in the corpus = y is more frequent than 5.

            Typicality, on the other hand, refers to the statistically significant co-occurrence of words in the corpus, the fact that they are not together purely by chance. Since we are talking about Sketch Engine, I will refer to the association measure that it is used to determine collocation (in Word Sketch and Sketch Diff). This is called LogDice and it identifies collocations that are exclusive and frequent. So, in a way you could turn off frequency in Word Sketch, and forget about it. Now, LogDice has a maximum value of 14. So a high score (closer to 14) indicates a strong collocate, and a low score indicates a weak collocate. A weak collocate is not a bad thing; it just means that the collocate is found with many other words in the corpus. SE uses the term ‘candidates’, I like the term ‘partners’. Which ones are useful for students, it really depends on their needs. For example, ‘good’ + ‘soldier’ is a strong collocate in BAWE and all occurrences refer to The Good Soldier: A Tale of Passion. Would I advise students to learn it? No. Would I tell English lit students not to paraphrase it? Yes. Again, it really depends on the level of the students. Someone might be already familiar with the collocation ‘team members’ (LogDic 11.4 / 119 occurrences) and might be more intrigued by ‘team morale’ (LogDice 7.9 / 3 occurrences). So, you don’t necessarily need to refer to the scores and frequencies in class, but it’s good to be able to know what they refer to if the students ask. I hope this shed a bit of light and it didn’t make things more complicated haha

        Viewing 2 reply threads
        • You must be logged in to reply to this topic.