Hi Mr.Enda

I recently have read the article from official IELTS site about memorizing. I think it good to share and paste some part of it here.

We address the challenge of assessing performance when IELTS (Academic Writing Task 2) candidates may have memorized and reproduced lengthy chunks of text that potentially disguise their true proficiency. Our profiling procedure separates out text that is more and less likely to reflect the candidate’s genuine linguistic knowledge. The procedure was applied to 233 retired scripts by Chinese candidates, and the results are analyzed by band and test centre.

As expected, errors decreased as band increased. Similarly, the quantity of non-generic nativelike text increased with band. But the use of material copied from the question and of ‘generic’ nativelike text (text that can be used in most essays) remained constant across bands for all but one test centre. Using the mean profiles as norms, a script known to be problematic was examined, to demonstrate how profiling can isolate the nature of differences. To assist examiners, a simplified version of the profiling procedure is offered, that can be used as an informal diagnostic. The profiling procedure recognizes the legitimacy of producing some pre-memorized nativelike material in a writing test, by contextualizing it within the broader pattern of the candidate’s written performance overall. The procedure requires further refinement than was possible within this modest project, but already suggests potential strategies for IELTS examiners to recognize memorized material in writing tests.

6.1 Recommendations to IELTS examiners
This study is able to make some first recommendations for the future training of IELTS examiners, regarding scripts that appear to have excessive memorized material. Firstly, the strong negative correlation in our sample between band score and the level of errors, and the positive correlation between band score and non-/topicgeneric nativelike language, indicate that the banding procedures are robust. The reason why potentially memorized material is problematic is precisely because it does not correlate with proficiency (though it did for Hong Kong scripts—see earlier). This means that examiners should have confidence in their intuitions regarding suspicious scripts.

Secondly, the evidence that memorization can be the path to effective learning, in both a first and second language, coupled with the fact that native speakers legitimately internalize useful turns of phrase as part of their own preparation for tests and exams, jointly create a dilemma in relation to whether it is appropriate to reward apparently memorized material. Examiners need not, therefore, feel that it is up to them to solve the problem of a suspect script: the difficulty is inherent and essentially insoluble, since there is no independent way to tell what the candidate truly ‘knows’ (nor any uncontentious way to define ‘knowledge’ in this regard).

Thirdly, if faced with a perplexing script, the examiner can adopt the simplified profiling approach described in this report, by highlighting continuous runs of linguistic material falling into the category ‘nativelike non- or topic-generic’: that is, material that is nativelike but not copied from the question, and that would not be worth memorizing for generic use across all written tasks. Isolated words, ie, words that are surrounded by material that would fall into another category, should not be counted as non-/topic-generic (see earlier description of buffer material). It is recommended that the procedure be carried out not only for the suspect script but also for a handful of uncontentious others, as a means of gauging the reliability of the coding relative to the norms provided here.

By counting the total number of words highlighted, and comparing them to the norm for the band the script appears to fall into, it should be possible to ascertain whether there are grounds for identifying the suspect script as abnormal (and the others profiled at the same time as normal). The present study suggests the following norms (based on a lower threshold of 2.5 standard deviations from the mean), though further research should be done on much larger samples to confirm these values. In particular, the Band 5 threshold, as determined in this study, seems possibly a little high relative to the others.

 Band 3 scripts contain no less than 22% non-/topic-generic nativelike material.
 Band 4 scripts contain no less than 26% non-/topic-generic nativelike material.
 Band 5 scripts contain no less than 37% non-/topic-generic nativelike material.
 Band 6 scripts contain no less than 40% non-/topic-generic nativelike material.
 Band 7 scripts contain no less than 50% non-/topic-generic nativelike material.
 Band 8 scripts contain no less than 59% non-/topic-generic nativelike material.

As the analyses in Section 5 showed, scripts falling below the threshold are not necessarily irregular—something that can be ascertained by examining other features of the profile. A truly problematic script, such as the one profiled in Section 4, will be strikingly different in regard to the distribution of the profile components. The purpose of the profiling should never be construed as that of ‘proving’ that material in a script has been memorized. That simply is not possible. Rather, profiling offers a means by which the examiner can offer a justification for his/her disquiet, as part of the case for a review of the script.

6.2 Recommendation to IELTS

We have argued in this report that a certain amount of memorized material in a script is not only acceptable but an indicator of task proficiency. We have also shown that in a given sample, such as the 233 scripts

....

**** To download the article : http://www.ielts.org/PDF/Vol9_Report5.pdf