valencia

HOLOGRAM: A fully Interactive Environment for Grammar Teaching and Learning



Sake Jager

Groningen University



Background



At the end of 1993 a major teaching innovation project was started at the Faculty of Arts of Groningen

University. Its primary aims were to enhance student achievement through the strategic application of CALL and

to increase teacher involvement in the use and/or production of courseware materials. The project is now running

in the departments of English, German, Romance Languages and Russian. The most important offshoot of the

project is HOLOGRAM, a fully interactive environment for grammar teaching and learning. Although the use of

existing grammar packages and authoring tools had been anticipated at first (grammar is after all one of the most

frequently targeted areas for CALL), it soon became apparent that none of these could meet the specifications

drawn up in close consultation with the language teachers themselves. These specifications included the

following:



 Major exercise types: gap-filling, drag-and-drop and multiple choice;

 Extensive feedback;

 On-line reference grammar;

 Adaptability to individual student level;

 No fixed learning paths;

 Continuous assessment;

 Student-monitoring facilities through the network;

 Content fully adaptable by non-programming teachers.



Providing all these facilities in a single package is a tall order indeed, so it was hardly surprising that none of the

programs available met our specifications on all counts. Rather than accepting this and making the usual

compromises with regard to functionality and content (the suitability of ready-to-use grammar programs was

rather limited for our purposes), we decided to start developing a program which was specifically geared to our

needs.



Crucial in this decision was the explicit demand for extensive, high-quality feedback and an on-line reference

grammar. Grammar teaching in the participating departments is characterised by a cognitive approach, in which

reflection about linguistic structures and familiarity with linguistic terminology are essential. For this purpose, the

departments involved had all developed proprietary material (text books, syllabuses and exercises) before the

project was started, which they now sought to integrate into the CALL-project. It was deemed essential that any

information pertaining to the learning tasks should be made available on-line. Moreover, to reduce dependence

on outside experts and to further the dissemination of CALL within the faculty after expiration of the project, it

was decided that the language specialists should be enabled to produce their courseware independently, i.e.

without the intervention of a courseware developer or programmer.



HOLOGRAM allows the language experts to do just that. Comprising a hypertext module and an exercise module

with a feedback component mediating between the two, it is a powerful, easy-to-use tool for teachers. Preliminary

test runs have demonstrated that both students and teachers highly appreciate the flexibility and versatility of the

program. Extensive testing will start in the course of this year, when all first-year students from the participating

departments (about 400) will be using the program. During this year the interface, parts of which are still

provisional, will also be finalised and a few further modifications and extensions carried out.



How students work with HOLOGRAM



The opening screen of HOLOGRAM is a menu providing access to the sets of exercises prepared by the teacher.

For each set specifics are given, such as progress, score and a list of grammar points covered in the set. Scores

and progress from previous sessions are included. The sets are dynamic in the sense that upon completion of each

set the original selection of exercises is replaced by a new selection of exercises not yet mastered. Further details

as to the implementation of this are given in the teacher section below.



In addition to the sets, the menu holds a pre-test option which can be used to compose a personal study path

tailored to individual weaknesses and deficiencies.





1

After selection of a set from the menu, the program makes a random selection from the types of exercises

assigned by the teacher, after which the first exercise appears. The major exercise types are gap-filling, drag-and-

drop and multiple choice. A typical sequence of student actions, which is much the same for all types of

exercises, is the following: answering the question, checking the answer, getting feedback and calling up the

relevant theory from the hypertext. None of these steps is mandatory, i.e. a student can look at an exercise

without answering it; answer it without calling up feedback; answer it, ask for feedback without consulting the

hypertext, etc. Alternatively, the hypertext may be consulted before the answer is given or the student may ask for

the correct answer to be given. A typical screen is shown in figure 1, which illustrates a drag-and-drop exercise

after feedback has been provided:









Figure 1



In our example, the student has filled two gaps by choosing from the answers in the list on the right: has tried in

“He has tried for years to get a permanent job”, and the same answer in “At the time the managing director has

tried to hush up the affair.” Graphic symbols are used to indicate that the first answer is right and the second

wrong. By clicking on the second symbol the relevant feedback has become available in the form of a small pop-

up screen.



Several points are worth mentioning here. First, answers are entered right inside the sentence where they belong.

This is important in cases where the context provides important clues as to which answer is appropriate, as in the

example above where the selection of a particular tense form crucially depends on the type of time adverbial in

the sentence.



Second, not all gaps on the page have to be filled for answer evaluation to take place. Each gap is assessed

individually, which makes it possible to check the answers one by one or to check them all at the same time. As

long as a gap has not been “touched”, the maximum score for it can still be obtained.



Third, the feedback establishes a close link with the exercise to which it refers by using words from the exercise

(these words are shown underlined): “The present perfect of the verb try is not possible here because of the time

adverbial at the time.” The words occur as variables in a generic frame making it possible to create a sense of

authenticity for each feedback text, while at the same time emphasising the systematic nature of the error made:





2

for a similar mistake made in another exercise the feedback might be: “The present perfect of the verb go is not

possible here because of the time adverbial yesterday.”



Fourth, the feedback is concise and makes use of specific terminology (present perfect and time adverbial). The

idea is that students who have studied the relevant theory will recognise the problem immediately and that

students not yet familiar with these concepts are referred to the relevant sections from the hypertext, which they

can open by clicking on one of the hyperlinks given (Present Perfect: Indefinite use or Past vs. Present Perfect).



The destination of the hyperlink is a page from the hypertext containing relevant information for the grammar

problem concerned. A typical example is the following:









Figure 2



Ideally, the information given on such a page provides sufficient background for the grammar problem

concerned, but hyperlinks in the text and other navigational devices such as tables of content and linear browsing

facilities make it possible to access other parts of the hypertext as well.



Since the entry point of the hypertext is the current exercise, there is little danger of students “getting lost in

hyperspace.” Although we did consider a bi-directional system in which students could access the exercises from

the hypertext as well as access the hypertext from the exercises, we decided in favour of a mono-directional

system because accessing the exercises from the hypertext might make the presentation of exercises rather

haphazard and would offer us too little control over the learning process.



In addition to getting to the hypertext through the feedback, it is also possible to access it by clicking a button on

the exercise screen. This normally opens the hypertext at a page which relates to the exercise as a whole and

contains less specific information than the destination reached through any of the links in the feedback. This way

students are given an opportunity to explore the relevant theory before answering any of the questions.



After having their answers checked, students can try the exercise again. There are no limitations to the number of

tries, but no scores are awarded after the second attempt. Students can come back to an exercise and try it again

after finishing other exercises first. Answers given previously and the symbols signalling their correctness or

incorrectness remain intact until the answers are changed. This makes it possible for students to explore

similarities between exercises and build on information obtained earlier.



Student results are stored in a database in the network. Although more detailed reporting is possible, only the

scores obtained for each set are stored, including the number of times a set has been tried. In addition, for each





3

item every student’s first score is stored as a crude means for teachers to determine the adequacy of items.

Finally, unrecognised answers with gap-filling exercises are stored so that teachers can add them to the lists of

possible answers if they want to.



How teachers work with HOLOGRAM



In teacher mode all texts–exercises, feedback and hypertext–can be written and edited. Editing occurs in a

“What-You-See-Is-What-They-Get” environment, i.e. texts are written or edited in the format in which students

will see them.



The procedure for creating a new exercise includes the following steps:

 write the exercise;

 create a set of answers;

 provide feedback for each of the answers;

 establish one or more links into the hypertext.



More details as to each of these steps are given below.



In order to create a new exercise, the teacher-author selects the option Exercise from the menu bar and chooses

the type of exercise (gap-filling, drag-and-drop, or multiple choice). Next, a dialog window appears in which a

number of properties of the exercise are defined. These include a link to the relevant section of the hypertext and

a code identifying the grammar point covered by the exercise. Apart from arranging exercises in a logical order

for subsequent browsing and identification by the teacher, this code is used to assign exercises to sets on the

menu for students (further details are given below).



After the properties have been defined, an empty exercise screen of the appropriate type is displayed where text

can be entered either by keying it in directly or by pasting it from the clipboard. I shall use the drag-and-drop

exercise shown above to illustrate each step in the development process.



The teacher version of HOLOGRAM has three distinct modes, one for editing the text, one for entering answers,

and one for providing feedback.



In text editing mode, directions (such as “Which verb forms are correct? Choose from the list on the right. More

than one form may be correct.” in figure 3 below) and the exercise proper are entered. The length of these texts is

limited only by the space available on the screen.



Next, in order to create a gap, text is selected using the standard selection methods for MS-Windows, such as

double-clicking on a word. By choosing the menu option Create gap the selected text is converted into a gap and

becomes the correct answer by default (cf. “has been trying” in figure 3 below). There is no limit to the number

of gaps created on a page. The correct answers for all gaps together are placed in the list from which students

select answers.



Editing of answers, such as assigning more than one correct answer to a single gap or adding distractors, takes

place in answer editing mode. For each gap, there is a list of correct and a list of incorrect answers. The list of

incorrect answers is composed automatically from the answers belonging to the remaining gaps, but may be

modified if necessary. These lists are visible in the top right-hand corner of figure 3 below. When another gap is

selected in the text on the left, the list of possible answers is updated accordingly.









4

Figure 3



In feedback editing mode–the mode shown in figure 3 above–answers are selected one at a time from the lists of

answers on the right. The example illustrates the feedback for the answer “tries” from the list of incorrect

answers. Feedback may be given for both incorrect and correct answers, the latter being particularly relevant

when students have the program show them the right answers and may need further explanation as to why they

are correct.



The feedback is edited in a separate window (Edit feedback in figure 3) and may either be a text used before or a

completely new text. Variables such as “for years” are inserted here by dragging them from the list of field values

(see bottom of edit feedback) to fields in the text. The links from the feedback into the hypertext are also

established here. Each feedback text can be given a unique label which facilitates finding feedback texts used

before.



Once created, every exercise can be tested immediately by clicking a special button which allows switching

between teacher and student mode.



The final step is to make the exercises available for students by assigning them to the set menu. This is done on

the basis of the codes indicating the grammar point covered. Normally a set will consist of exercises from a

range of codes. After completion of the set by the student, the codes for areas mastered are filtered out while

those for areas not yet mastered remain in the set, so that new selections of exercises from increasingly smaller

ranges of codes can be provided as students work through the same set several times (if they want to).



What remains to be done



The description of HOLOGRAM above reflects the development of the program one year before the closing date of

the project. The final year will be used to further improve the interface (no doubt you will have noticed the rather

cluttered screen in figure 3 above) and implement functions not yet available (such as the pre-test option).



Moreover, the language experts’ work continues to meet the target of 1,500 to 2,000 exercises for each language

(with full feedback and grammar theory).







5

In addition, we are considering extensions outside the grammar domain of the project. The HOLOGRAM shell is

sufficiently generic for use in other teaching domains. The department of Art History of Groningen university is

experimenting with the program right now to assess its use for multimedia applications. Multimedia capabilities

were anticipated when we selected Asymmetrix Multimedia Toolbook 3.0 as our development tool, but adding

them is not feasible in the context of the current project, which is why we are trying to find extra funding.



HOLOGRAM will be introduced for general use in our own faculty in the summer of 1996. It is our intention to

make it available to a wider community, but through which distribution channels and at what price are uncertain

at this time of writing.









6