Re: tictech: DWA comments

From: Kurt Sahl & Kathleen Mertens (bluesky@scn.org)
Date: Fri Mar 14 2003 - 14:46:22 PST

  • Next message: Graham Ford: "Re: tictech: DWA comments"

    -tictech message:

    Mark--You make a valid argument for the use of spell check by students
    who use computers to complete their DWA. I hope some of the tictech
    subscribers forgive me as an aspiring PhD candidate, but I'd like to
    forward to the list a study released today on this subject. Then, I'd
    like to respond to your post with possible reasons why a decision to
    prohibit use of this function might be made.

    A study on this subject was announced today and picked up by CNN.com.
    (See story @
    http://channels.netscape.com/ns/news/story.jsp?floc=FF-PLS-PLS&id=2003031415
    165001000002&dt=20030314151600&w=CNN&coview=)

    Below is the press release:

    =======================================================================
    Do Spell Checkers Do More Harm than Good?

    Just about everybody uses a computer spelling and grammar checker. Do
    they truly help improve documents or are there any "side-effects?" A
    recent study by University of Pittsburgh Katz Business School I.S.
    professor Dennis Galletta demonstrated that spelling and grammar
    checkers can diminish rather than enhance the skill level of writers.

    According to Galletta, when computer spelling and grammar checkers were
    turned off, people with high verbal scores on the SAT did better than
    people with lower Verbal scores. However, when the spelling and grammar
    checkers were turned on, the performance of each group dropped.

    Says Galletta, "The spelling and grammar checkers made high-verbals
    perform worse than usual, or just like low-verbals."

    Galletta's research examined three types of errors; errors flagged
    correctly by the computer, errors found by the computer that were not
    truly errors; and errors missed by the computer because perhaps the word
    is spelled right but used incorrectly, as in "to bare arms."

    Participants were graded on their performance on finding each of the
    three error types that were planted in a business letter. The goal of
    each participant was to find all errors and take appropriate action
    thereby leaving the document with zero errors.

    When errors were correctly flagged, all participants performed well.
    However, trouble began when the grammar and spelling-checking software
    itself had problems, a common situation. Both groups did better without
    the software when correct items were erroneously flagged as errors by
    the software; they seemed to believe what the software was telling them
    even though it was incorrect. They ruined perfectly good sentences.

    It was even more interesting when errors were missed altogether by the
    software. Without the software, the high-verbals performed about three
    times better than the low-verbals. However, with the grammar and
    spell-checking software, the performance of the high-verbal subjects
    sank to equal that of the low-verbal subjects.

    "This study shows that when the spelling and grammar checker were on,
    both the low and high verbals got lazy, and ignored the fact that errors
    were still present, despite what the checker told them," says Galletta.

    Galletta stops short of recommending warning labels to be placed on
    office software, but says "when spelling and grammar checkers make
    obvious errors, it does us a favor because it puts us on alert that the
    software is not infallible. The trouble is, we don't take those
    infallibilies seriously enough."

    For more information on Galletta's study, e-mail him at
    Galletta@katz.pitt.edu
    =======================================================================

    I e-mailed him and he sent me the actual study in .pdf format. If
    anyone would like a copy, send me an e-mail.

    Now, what to make of this? First off, the study took place in a
    university setting and, if it makes any difference, was conducted by a
    business school prof. It would be inappropriate to generalize the
    findings to an elementary setting. Second, it is just this kind of
    study that policy makers read to make their own generalizations.

    Mark said:
    > I suggest that in 2003 this is totally unrealistic and is indeed
    > punitive. My kids write better on a computer than on paper - almost
    > all of them. Their writing on a computer is more complex, longer,
    > better organized, and easier
    > to read. They are more at home in front of a word processor than with
    > a piece of paper. I have 14 computers in my classroom, and all my
    > kids write on a computer daily. Spell check is ALWAYS ON. They become
    > better spellers as a result of having it on. They LEARN from it.
    > It's as accepted as an eraser on a pencil.

    One issue that emerges when one compares the results from research with
    those of a teacher's classroom observations is that results from a
    teacher's classroom observations are extremely hard to replicate. For
    example, you said, "They become better spellers as a result of having it
    on. They LEARN from it." As you know Mark, I am aware of your experience
    using computers in the classroom and do not question it here. You'll
    forgive me for a second if I play devil's advocate. How do you really
    know if your students have _learned_ from using spell check? Have you
    conducted a study of your own? If so, what were the results?

    The point is if I have to make a policy decision about the use of
    computers when a student takes his/her DWA, I need information regarding
    the fairness of the tool. In the absence of solid information, I must
    rely on experience. If the DWA is used to determine students' ability
    to compose prose, and one student has access to a spell checker and one
    does not, that may not be fair. Fairness is germane because I may be
    unsure about the spell check's effect on correct spelling. The use of
    the dictionary may level the field because I can assume (due to
    tradition) that _all_ students possess a skill set--dictionary
    skills--but I cannot assume _all_ students can choose among word choices
    in a dialogue box, an untraditional skill set (emphasis on "all"). In
    summary, I want to make sure that computer-using students do not have an
    advantage when taking the DWA.

    Now, according to this study, it appears that the use of a spell-checker
      does IN FACT level the playing field. Of course, more research in an
    elementary school setting is needed to shed light on the spell check
    issue. This study turns the table on conventional wisdom and raises the
    question of whether the spell check function hinders rather than
    advantages its users.

    Great conversation! This is an issue that deserves attention because
    knowledge about when and where, how and why computers may be superior to
    paper and pencil challenges long-standing notions of student learning.
    While the process is indeed slow, the marble chips are making louder
    sounds when they fall to the ground. Keep it up the good work!

    Kurt Sahl
    University of Washington, College of Education
    PhD candidate and father of 2 SPS students
    bluesky@scn.org

    -end tictech message. To join, leave, or visit
    the message archive, go to tictech on the Web:
    http://www.earthdaybags.org/tictech/



    This archive was generated by hypermail 2b29 : Fri Mar 14 2003 - 15:02:47 PST


    Learning Space Development Server
    This page under development for The Learning Space
    Copyright ©1996-2000 by the Authors - All Rights Reserved
    Unauthorized use prohibited.
    This site was whacked using the TRIAL version of WebWhacker. This message does not appear on a licensed copy of WebWhacker.