Question Tools Logo ™

www.questiontools.com

RSS Logo
About Company People History Credits Contact
News Stories Articles International Use Reviews & Comments Customers Awards Press Mailing List Competition
Products Editor Suite Editor Chooser Editor Edition Table NetSeries NetSeries Chooser NetSeries Edition Table Elan
Services Publishers Hosting Snapshot Archiving Conversion End-to-End Examinations Training Support
Prices Products Hosting Elan Educational Products Educational Hosting
Buy Products Hosting Education: Products Education: Hosting
Downloads Installers Updaters Guides and Factsheets How to Guides Language Scenarios Examples Discussion Group
Support Feedback Examples How to Guides Common Problems Known Issues Changes & Fixes Email Support Discussion Group
FAQs Demos SimpleSet Editor NetSeries Taking Tests

Browser Errors

The problem

There have been infrequent reports of tests with incorrect maximum marks, as well as tests where all screens were displayed to a candidate, instead of just a random selection.

Why does this happen?

Question Tools software gets the browser to do quite a bit of the work, and in particular selecting the questions and calculating the marks. The advantage of this approach is that the server does not have as much to do, and so the whole thing runs more quickly.

However, browsers protect themselves against scripts that are too long or infinite by timing out scripts that run over a certain amount of time. If the machine is particularly slow, or if there are other processor-hogging applications running (such as antivirus software), there is a chance that the script in the exam will have no processor time and will be timed out.

In short, these problem occur because of slow machines, lack of memory, and a bit of bad luck (i.e. if something else is hogging the processor).

Safeguards

Question Tools does not just collect the percentage correct and the marks from the browser. It also collects every selection the user makes. In NetSeries it is possible to open a completed test and step through it a screen at a time. While the main results pages in NetSeries show the result as reported from the candidate's browser, the results displayed when you step through a candidate's test are recalculated, and in this way you can compare the recorded results, with one that has just been recalculated in your own web browser. In effect, this is one of the safeguards — multiple sources of information are collected and presented to enable comparison.

Dealing with the problem

With version 3.0.16 of Editor we have moved the randomization code to an earlier point in the process. In particular, the questions a candidate will see are calculated when she or he first sees the pre-test screen. There are then attempts after this to calculate the random selection of screens if the first did not succeed.


Home   •   Reviews & Comments   •   International Use   •   Customers   •   Links   •   Contact   •   Legal & Refunds   •   Privacy