Case Story: E-gramm

Summary	E-gramm is a grammar checker for students of English as a foreign language
Group / workshop	Formative e-Assessment	Status	seed
Project details...	A team led by Dr. Jim Lawley and Dr Rubén Chacón Beltrán in the Facultad de Filología at the Universidad Nacional de Educación a Distancia (UNED) has developed a program - e-gramm - to support learning of English as a foreign language. Source of the account: http://ocde.p4.siteinternet.com/publications/doifiles/962008011P1T011.pdf (pp. 11-14). The majority of text below is drawn verbatim from this report with some modifications and simplifications. Software available at: http://www.innova.uned.es/

Slideset

slideset for this case study on slideshare (please use the S.T.A.R template: http://www.slideshare.net/yish/star-case-study-template)

more help on embedding slidesets... (Ctrl+click to open in new window)

Situation

What was the setting in which this case study occurred?

Facultad de Filología at the Universidad Nacional de Educación a Distancia (UNED), a university characterised by a population of students aged 25 and over with no formal academic qualifications, has developed a computer program, e-gramm, specially designed for students of English as a foreign language with a very low level of proficiency.

Task

What was the problem to be solved, or the intended effect?

• Students of English as a foreign language with a very low level of proficiency.
• Adults learning English in an autonomous or semiautonomous learning situation.
• Requirement for more extensive writing practice than is possible in conventional classrooms with large groups of students.
• Desired method of teaching is a task-based methodology in which students are encouraged to write freely and abundantly on many different subjects so that their mistakes emerge naturally, since these mistakes are, naturally, learning opportunities. Reaping the benefits of this approach requires that the mistakes are duly corrected – preferably by the student.

Actions

What was done to fulfil the task?

Development of e-gramm which is a grammar checker, similar to that in Microsoft Word, but specially prepared for students of English as a foreign language. It enables them to detect and, in the light of corrective feedback, correct the mistakes they make when they write in English.
The programme offers two complementary filters: incorrect sequences and problem words. In the case of incorrect sequences, the computer programme highlights in colour the sequences of words in which it has detected a mistake and provides on-screen feedback which explains the errors, offering examples of correct usage and explaining how they differ from what the student has written. The student processes this information and makes the appropriate changes to what has been written. By clicking again on the incorrect sequences button the student will discover if that section of the composition is now correct. Clicking on the other button - problem words - causes the programme to highlight words which are often associated with error, and to provide feedback which will enable the student to decide if in this instance the word has been used correctly or not.
The project team analysed many hundreds of student compositions, detected the mistakes, encoded them computationally, fed them into the database, and wrote corresponding feedback, which enables students to modify their written production. The existing prototype of e-gramm is designed to detect 500 errors which accounts for approximately 60-70% of the mistakes made by Spanish-mother-tongue students at level B1 in the Common European Framework. The programme is capable of expansion to detect up to 80,000 errors and the aim is for the team to continue to feed its database. With e-gramm, correction is not automatic but instead requires
students to understand the grammatical and lexical explanations and then write the correct version themselves; the expectation is that this cognitive process will produce learning.
At times only the student can decide if what (s)he has written is correct. A phrase like this, for example: Actually, he is working in London is grammatically well formed, and if the student wishes to say, in effect, „In fact, he is working in London‟, it is also correct. However, if the student means „At present, he is working in London’, it is incorrect. The student, as the author of the phrase, knows better than the teacher which meaning is intended, and it is the student who should decide whether correction is required.

Results

What happened? Was is a success? What contributed to the outcomes?

No formal evaluation has been carried out. The authors make the following claims:

• E-gramm fosters just-in-time learning. It provides the student with the required information to determine if a phrase is correct or not at the most opportune moment; just when the student is most interested and receptive.
• E-gramm provides students with highly motivating and useful work because they are working on their own written production.
• By assuming responsibility for correcting their own writing, students become more autonomous.
• E-gramm allows students to progress at their own speed.
• Although intended for lower-intermediate level students it has also proved popular with more advanced students who are anxious to avoid embarrassing „howlers‟ in their written English.

Lessons Learned

What did you learn from the experience?

The authors of e-gramm claim: "The program entails the active participation of students in their own continual assessment and serves therefore as a good example of formative assessment."

The definition we are adopting for this project is: An assessment functions formatively when evidence about student achievement elicited by the assessment is interpreted and used to make decisions about the next steps in instruction that are likely to be better, or better founded, than the decisions that would have been made in the absence of that evidence.

Now, it seems likely that e-gramm would result in improved learning, but it is not clear that on its own it would enable better decisions about the next steps in instruction to be taken - though one could readily imagine a wider instructional context in which it could be embedded that would do this.