Today we have the question type RANDOM, with which we can accomplish individual quizes. Especially for numerical questions it could be neat to have another framework for attaching individual parameters of files to questions that use the same question text.
This is a feature that my collegue statistics teachers have asked for and together we have come up with a possible model for the solution. As we are dealing with statistics, it comes natural to have individual stuff as individual measurement data on file. Each student gets its own file and then analyzes it according to the question text.
For educational reasons it could be a good idea to have the student use the very same data file for a number of questions and quizes. In statistics this gives us the possibility to show the students how the use of different statistical tools end up with different results. We have come up with a specification that contains these key points:
As data is to be reused between questions, it should be stored at category level (or simular). When an individual question is created, the individual items are chosen from the individual items that the category offers.
The first time a student takes an individual question from the category, he/she will be assigned an individual dataset from that category. Whenever the student later takes another individual question from that category, he/she will reuse the same dataset that was used the first time.
An individual dataset will contain a limited number of named items. The names and types of the items are defined by the teacher at category level. Example: The category "Some Category" can have four item definitions for its datasets: "number" as parameter; "bloodpressures" as file; "time for recovery" as file; "age group" as parameter.
I intend to have the parameter number as an automatic default, having number=1 for the first dataset, 2 for the second etc. For each dataset the teacher then needs to upload a file 'bloodpressures' and a file 'time for recovery' and set the parameter 'age group'.
If the student later gets the dataset item 'age group' for a question, the student will also be bound to always get the items 'bloodpressures', 'time for recovery' and 'number' from this dataset if he/she later have an individual question that uses any of these.
My plan is to start by constructing a page for editing datasets (datasets.php). It will be reached from the "Editing quiz" page as shown in the attached example. It will store dataset definitions and datasets in these tables:
CREATE TABLE `prefix_quiz_dataset_definitions` (
`id` int(10) unsigned NOT NULL auto_increment,
`category` int(10) unsigned NOT NULL default '0',
`name` varchar(255) NOT NULL default '',
`type` smallint(6) NOT NULL default '0',
PRIMARY KEY (`id`),
KEY `category` (`category`)
) TYPE=MyISAM COMMENT='Options for multianswer questions'
CREATE TABLE `prefix_quiz_dataset_items` (
`id` int(10) unsigned NOT NULL auto_increment,
`number` int(10) unsigned NOT NULL default '0',
`definition` int(10) unsigned NOT NULL default '0',
`value` varchar(255) NOT NULL default '',
PRIMARY KEY (`id`),
KEY `definition` (`definition`)
) TYPE=MyISAM COMMENT='Options for multianswer questions'
I continue by creating the quesiton type INDIVIDUAL. When editing the question the teacher provide a question text and also pick the relevant dataset items by names.
It is thereafter necessary to define the actual question type and the answers for each one of the datasets in the category. It would be useful to use some kind of file import to define these answers. I cannot offer any format for this but I hope others can come with ideas here. When the questions have got their answers they will be ready to use in quizes.
I will go ahead with this now. If you wish I can have things checked in as I proceed or have a huge bunch of code submitted when all is done.