What is offered here is similar in many ways to a multiple choice presentation package for a stand-alone computer. The difference is that these pages make available roughly the same degree of functionality normally expected from stand-alone packages, but on the internet.
'Dynamic multi-choice' is not a static test, it is :
| Top of Page | Operational Requirements | General Information | Design notes | User notes |
To view this software you need a fairly up-to-date computer and web browser (the pages were developed on a Macintosh PowerPC using Netscape Navigator 3.01).
If you have difficulty receiving these pages you may find that your browser options for Java Script and cookie reception need to be enabled.
| Top of Page | Operational Requirements | General Information | Design notes | User notes |
(a) Automatic test item selection: when the user loads the 'Multi-choice' page, the Java Script code automatically puts together a ten-question test, selecting ten different questions from a 50 item test bank.
(b) Item-by-item presentation: the questions are presented one by one; each time the user clicks a radio-button to indicate what he thinks is the correct answer, and then calls for the next question.
(c) Random sequencing: the order in which the questions and the choices are presented is randomized, so a different set of questions should appear in each test. Even if exactly the same questions appear in subsequent tests, their sequence within the selections will almost certainly be different. Where the details of individual question repeats are concerned, the choices (distractors) are more likely than not to appear in a different order. Checks are maintained to ensure that the same question never appears twice in the same test (i.e. the same selection of ten items).
(d) Performance scoring: within each test a count is kept of the number of (executed*) attempts, the number of items answered correctly, and the choice made where the answer is incorrect. A current-test scoreboard is displayed and updated each time the user calls for a new item.
(*executed attempts: the 'final designated choice' per question -- while deciding on which distractor is the right one, the user may click on any of the four radio buttons any number of times without penalty.)
(e) Automatic level assignment: on completion of a ten-question 'test' the user is shown the correct choices for all the questions presented, and if he or she wants to proceed to another test, a decision --based on the user's performance-- can be automatically made as to whether or not the next test will be harder (if 80% or above scored in the current test), easier (if less than 60% scored in the current test), or at roughly the same level as the one just done. (For an on line demonstration of this see the "Multi-choice" pages accessible from www.luxfield.co.uk/Resources/PQT/intro.html).
(f) User activity record (under development): on completing a test, details of user activity and performance are copied to the server; this data is not made directly accessible to the user --i.e. the person taking the test-- but can be accessed and analysed by the test provider.
(g) Operational environments: these and similar internet pages are being developed and tested in Mac PowerPC+ / Netscape 3+ environment, and with a modernish PC and Internet Explorer browser.
(h) Intention: The pages available here are experimental. The experiment is to see if the technical aspect works satrisfactorily. Content, although every attempt has been made to use items that are useful, is in this context of secondary importance.
(i) Ranges: See the "Multi-choice" pages accessible from www.luxfield.co.uk/Resources/PQT/intro.html for a demonstration of a five level presentation --easy, average, hard, really hard and horribly hard. In fact there need be no practical limit to the number of levels offered.
(j) Potential: The software has uses beyond the domain of EFL practice and testing: it could be used for practice tests in many other subject areas.
(k) Modifications: As with most software, feature and function are different issues: contents, formats, page-layouts, colours, etc. can all be modified to suit specific needs. If these pages run satisfactorilly on the internet, LAN versions as well as off-line stand-alone versions can be developed.
| Top of Page | Operational Requirements | General Information | Design notes | User notes |
Coding: I am using HTML code (version 3.01) and Java Script (version 1.1). Most of the functionality and effects result from relatively heavy use of the random numbers generator and a 'shuffle' function. Display formats derive largely from plain HTML code for tables,embedded in the Java Script code but only dependent on it for the varying colour schemes
Interactivity: In HTML pages interactivity can be engineered through the use of 'Forms' or 'Cookies'. Luxfield's Dynamic HTML Pages make use of both.
Cookie details: Each of these dynamic .html pages contains Java Script code, which does two main things: (a) it generates data to make specific internet-pages, and (b) it examines and reacts to a short string of application-generated data embedded in the document headers (the "cookie").
For the dynamic and fixed multiple-choice applications, the cookie initially contains the following data :
(a) a number indicating the number of times a request has been made from a specific user station, during the current session.
(b) the ten numbers of the questions selected.
(c) after each number, a '*', standing for 'not attempted'.
When the user clicks a choice button, code in the 'Multi-choice' page immediately works out whether or not the choice is right. If it is, the '*' in the cookie is changed to 'a' --for 'right'; otherwise 'A', 'B','C', or 'D', to indicate which of the 'wrong' distractors was chosen.
When the user clicks on the 'next question' link, a request, with the adjusted cookie, is sent to the server for the next question. From the request for the second question onwards, when Java Script looks for the cookie it finds one containing details of the questions selected for the current test, the number of items attempted, and the number answered correctly. On extracting this data it updates the score, and presents the next question.
The procedure is repeated until the tenth question, when the request from the user is for the 'Report' (a different .html page). Here a summary is produced, all the items selected --and their correct answers-- are collated. Arrangements are made, depending on the user's performance, as to which link will be activated if 'another test' is requested.
Design: I make minimal use of graphics or sophisticated page design. This is done to ensure users don't waste time waiting for things to happen. In all pages except the one for the easiest level, what might look like graphics are in fact tables generated by plain HTML code.
Feedback Very grateful for any feedback, especially on the technicalities. Please send to: ed@luxfield.co.uk
| Top of Page | Operational Requirements | General Information | Design notes | User notes |
Luxfield's 'Dynamic Multi-choice' offers a series of multiple choice test and practice session presentations (for English Language).
Each presentation takes questions from a bank of about 50 items.
When you load 'Dynamic Multi-choice', the experimental Java Script code will make up a ten question test for you, i.e. the computer will choose for you any ten questions (all different) from the 50 item test bank.
You can take as many tests as you like. They will (should) all be different -- though of course if you take more than a few tests you will begin to see the same items appearing again and again. (This might be good --more chances to try again, and achieve a higher percentage.)
Each question comes with four choices. You make your choice by clicking on the small round button (the 'radio' button) to the left of the choice. If you change your mind, you can click on a different button. You can do this as many times as you like.
When you have decided which of the four choices you think is the right one, you should ask for another question --click on the 'Next question' link near the bottom of the question display (on the right, for 'Very Easy English Words').
When the next question appears, the scoreboard will show if your last choice was correct. If it was, the figures in the centre of the score board will show some improvement (unless, of course, they previously showed all correct at 100%).
If you look to the right of the score board, you should see a list of ten numbers. These are the test bank item numbers of the ten questions selected for your test. They are coloured red, white and black. The colours mean:
If you want to tell me about your experience of using this sample write to ed@luxfield.co.uk Tell me what hardware and browser you are using.
DF Salthouse Oct 99
| Top of Page | Operational Requirements | General Information | Design notes | User notes |