Loading ...
Sorry, an error occurred while loading the content.

Web ER and grading

Expand Messages
  • Gordon Luster
    Hello, I have a couple of items related to the recent discussions of web ER and grading schemes. First, I ve been working with a small group in western Japan
    Message 1 of 3 , Feb 1, 2005
    • 0 Attachment
      Hello,

      I have a couple of items related to the recent discussions of web ER and grading schemes.

      First, I've been working with a small group in western Japan to put some graded reading
      and listening content on the web (including cell phones). I presented some early results
      from the project in the ER forum at the JALT 2004 national conference, so some of you
      may have seen the site there. We are planning to make our prototype site, with about fifty
      articles in three levels, available free to Japanese college/university users and a few others
      starting in April, in the hope of expanding our community and getting some feedback on
      how we can improve. If anyone is interested in trying out the site or maybe participating
      in the project, please email me at: gordon@...

      Second, the recent posts by Rob Waring (number 2164) and Brett Reynolds (number 2174)
      were interesting to me, because one issue that has troubled us from the beginning is how
      to grade our content. (I posted a question about grading to this list back in July of 2003,
      before we started.) Currently, it seems to me that the grading of ER materials is rather
      chaotic, with different publishers adopting different criteria, which they are keeping largely
      secret. The result is that we cannot readily compare levels from different publishers and
      cannot understand in detail how levels are defined even within the output of a single
      publisher. EPER has made an effort to grade content from different publishers on a
      uniform system, but we are still left with the reality that the materials from different
      publishers were graded according to different criteria, so there are still some mismatches
      within levels.

      In our own project, each of our stories appears at (currently) three levels, which
      correspond roughly to the 1000 and 2000 levels flagged by the vocabulary profiler on Tom
      Cobb's site at http://www.lextutor.ca/vp/eng/ (also cited by Brett in his post), plus an
      "original" level in unsimplified English. In the earliest days of our project, I asked writers
      to use the profiler's output as a rough guide in adjusting the vocabulary of each level and
      also added some loose criteria concerning sentence complexity. However, I found that
      there was still more variability among the products of different writers than I had hoped,
      so recently we have decided to switch to a dedicated rewriting staff who can be trained
      (eventually) to produce more consistent levels.

      It would be helpful, I think, to have a detailed publicly available set of level criteria that
      anyone producing graded materials would be able to use easily. This could be something
      like Rob's proposal of a system for rating web content, except that it could be used for
      production, not just for hanging numbers on previously published materials. With such a
      system, it would be possible to form a community of materials providers all working to the
      same standards, making exchanges of materials more feasible and enhancing the usability
      of everyone's output. We might start with software similar to Cobb's profiler, merging his
      word lists with lists of common multi-word items taken from corpus studies. We might
      also suggest some fuzzy guidelines regarding sentence length and complexity and maybe
      some common readability statistics. From my own experience, I wonder whether strict
      grammatical criteria would be appropriate, because the need for certain structures is
      dictated too much by the genre and subject matter of the writing. I also wonder if lists
      above the first 2000 or 3000 word families would be useful, because at higher levels
      learners mainly need to acquire words related to their particular situations and interests.

      I think if a comprehensive public grading system could be made available, it might become
      a useful standard for both web-based and paper-based ER materials. It would be a
      significant project, though, and might need the input of several people from different
      institutions to achieve wide acceptance. This group would be a good place to start. Would
      anybody be interested?

      Gordon Luster
    • Rob Waring
      Hi Sorry for the local nature of this broadcast. It s that time of year again. As you all know Jalt Shizuoka will be held from the 8th to 10th of OCTOBER
      Message 2 of 3 , Feb 2, 2005
      • 0 Attachment
        Hi

        Sorry for the 'local' nature of this broadcast.

        It's that time of year again. As you all know Jalt Shizuoka will be
        held from the 8th to 10th of OCTOBER (not November as usual). The
        theme is 'telling stories' and there will be a heavy influence of ER.
        Marc Helgesen and I will be co-conference chairs and as such I don't
        think we can be directly involved in setting up the symposium we
        always have. Marc and I have been working to get some big ER types to
        present.

        Would anyone care to be involved this year? If you do please send me
        a mail on waring_robert@... and I'll add you to the JALT ER
        Symposium discussion list. Any controbutions are welcome.

        Our deadline is 28th February.

        We need a co-ordinator, a moderator and some presenters. Hands up
        please.


        best regards

        Rob
      • gradedreading
        ... Hi Regarding Rob s proposal for a standardised grading scheme for web based reading materials, I d be happy to submit my own site to be analysed. At
        Message 3 of 3 , Feb 4, 2005
        • 0 Attachment
          >
          Hi

          Regarding Rob's proposal for a standardised grading scheme for web
          based reading materials, I'd be happy to submit my own site to be
          analysed. At present I use a rudimentary scheme combining the
          readability stats available with Word (Flesch Reading Ease, Flesch
          Kincaid Grade Level) with a personal judgement based on my experience
          as a teacher and writer.

          Kieran McGovern
          Editor, EFL Reading
          http://www.gradedreading.pwp.blueyonder.co.uk/index.html
        Your message has been successfully submitted and would be delivered to recipients shortly.