Basic Concepts

From siette.wiki.en
Jump to: navigation, search

Contents


no7hOo <a href="http://phftwxnqleso.com/">phftwxnqleso</a>, [url=http://pumsukaqyuuh.com/]pumsukaqyuuh[/url], [link=http://ifdnisjzhwpc.com/]ifdnisjzhwpc[/link], http://lkhawymiwijq.com/

http://testing.alexandriaarchive.org/entry-humornsex.com.com entry humornsex.com.com free book audible.com http://testing.alexandriaarchive.org/www.audible.com www.audible.com np 3715.pdf http://testing.alexandriaarchive.org/answer-www.elf.wwme-org answer www.elf.wwme org lamescia.com http://testing.alexandriaarchive.org/entry-www.lbpoints.com entry www.lbpoints.com post big canal96.com http://testing.alexandriaarchive.org/www.aol.games.com www.aol.games.com tqjunior.thinkquest.org http://testing.alexandriaarchive.org/www-1913-intel.com www 1913 intel.com libery.thinkquest.com http://testing.alexandriaarchive.org/reply-www.bissettmags.com.au reply www.bissettmags.com.au piano sheet music.com http://testing.alexandriaarchive.org/www.audible.com-rca-3freeprograms www.audible.com rca 3freeprograms boyne highlands.com http://testing.alexandriaarchive.org/topic-mrs.-deauxma-videos topic mrs. deauxma videos corinne labbe noretud.com http://testing.alexandriaarchive.org/www.audible.com-rca www.audible.com rca john s. steel

Athos

Athos is the new editor used in Siette, available from version 3.3. It is recommended to use this editor rather than the previous one, Tedi.

Test

A test is an exam created and delivered using Siette. It consists of a set of items and a group of parameters that define which and how many items can be included in the test, how the results of the test are assessed, etc.

Item

An item is a question which appears in a test. Siette works with the item types described below, as well as with their evaluation mechanism. Furthermore, items can be siettlets, generative, timed or external. These categories are not mutually exclusive. For instance, an item can be multiple (of basic type), siettlet, generative and timed.

Basic items

Basic/Internal items are those supported directly by the Siette assessment engine. That is the atomic pieces of evidence that can be processed by Siette to estimate the student knowledge level. Internal types are:

Multiple choice, single answer questions

This kind of item presents more than two possible answers (choices). Students may select one (or none) of these answers.

Multiple choice, multiple answers questions

These items are multiple-choice items with more than one correct choice. Students must select all correct choices in order to score points for the question. This kind of item can be further classified into:

  • Items with independent choices. Choices are mutually independent. This type of item is equivalent to a set of true/false items.
  • Items with dependent choices: The correct answers are combinations of the set of possible answers. For the answer to be correct, all members of a combination must be selected. This kind of item is equivalent to a multiple-choice item where the answers are all possible combinations. That is, if the item has n possible answers, it is equivalent to a multiple-choice item with a factorial of n possible answers. (Not yet implemented)

Short answer questions

When working with this type of item, students must write one or more answers which satisfy certain conditions. Possible answers are stored in the knowledge base as response patterns which identify correct as well as incorrect answers. In addition, Siette automatically includes a default pattern which corresponds to an incorrect response which, in turn, does not satisfy any of the given patterns. For each of the answers written by the student an evaluation procedure is applied, depending on the open-answer type (see below).

Depending on the way in which the written answer is corrected, there are four types of open-answer questions:

  • Matching: Answers supplied by the student are directly compared with the response patterns.
  • Siette Regular expressions: During the item construction process, the patterns stored are regular expressions. The item correction procedure consists of checking, for each response pattern, whether the set of given answers belong to the language generated from the corresponding regular expression.
  • Java regular expressions: These are analogous to the latter, but allow the construction of more complex patterns. These patterns are generated according to the Java regular expression class. This class is available from the 1.4 version of Java language.
  • Openmath patterns: They are designed to deal with mathematical functions of a single variable. (Under development)

The four types of open-answer items have been implemented as plugins. Consequently, it is easy to include new types of items. It suffices to extend an abstract class with the correction mechanism desired. The abstract class simply requires the implementation of a method which determines whether a given string belongs to a certain pattern or not. This type of item allows the system to store incorrect response patterns. Such patterns are particularly useful in self-assessment tests since they allow the identification of misconceptions and generate adaptive feedbacks. Incorrect patterns can be inferred by applying bootstrapping techniques during the item calibration process. These patterns can be obtained by analyzing the performance of students that have previously taken tests containing the item. In tests where the item correction is shown immediately after the student's response, it is necessary to include an example of a correct response which will be displayed if the answer given by the student was incorrect. This example must be added during the item construction stage. Once the plugin has been programmed, Siette should be recompiled, reinstalled and restarted.

From an assessment perspective, this type of item can be considered a collection of multiple-choice items with a size equal to the number of boxes shown within the item.

External links

All other types of items, like those containing Java applets, or external items, are finally converted into one of these internal items for evaluation purposses.

Siettlets or autoevaluated items

Generative items

Template library

Two examples. Pastime items

Timed items

External items

Scoring

The final scoring of a test depends on the evaluation method, which can be one of the listed below:

Personal tools
Namespaces
Variants
Actions
Navigation
SIETTE Manual
Toolbox
In other languages