Task Analysis Comparison Paper

According to Dix (1998), “Task analysis is the process of analyzing the way people perform their jobs: the things they do, the things they act on and the things they need to know” (p.260). There are three different approaches to a task analysis:
• Task decomposition
• Knowledge- Based techniques
• Entity- relation-based analysis

All three approaches have some common characteristics; however, they emphasize slightly different areas. Task analysis serves many purposes. It helps to research on requirements for new systems and procedures, and it helps to produce technical documentation for training materials.

Task Decomposition
Task decomposition is focusing on the technique where a task is split into subtasks and the order they are being executed. In task decomposition the most typical approach is to use hierarchical task analysis where the tasks and subtasks are written in the order they are performed. The tasks are given the numbers in which they should be executed. In addition to the order, the plan is also stating which tasks should be performed and which one can could be omitted. The task hierarchy can be shown diagrammatically or textually. When designing the task hierarchy, subtasks may need to be expended and might require additional information or a stopping point.


Knowledge- based analysis
Knowledge- based task analysis refers to creating task analysis by listing the names, object and actions and then constructing a list of related words. This approach is based on taxonomies of the required tasks. When building such hierarchy it is good to start with listing all related words, and later removing items that are unnecessary. Often the taxonomy already exists, therefore, when using this approach, we may ask the user to sort and group related objects. Task analysis for knowledge description (TAKD) uses task descriptive hierarchy, which is a special kind of taxonomy. In the simple taxonomy the branches are either/ or branches. Producing a simple taxonomy for action and objects is very similar. There is always a relationship between taxonomy of actions and the hierarchical task analysis descriptions of tasks.


Entity- relation-based analysis
Entity- relation-based techniques are usually connected with database design and object-oriented programming. This model is based on:
• Entities – we seek information about them
• Attributes- the data that we collect about entities
• Relationships- information needed to build the structure

The entities used in this approach are the ones that will be represented in the consequential computer system. The emphasis in this approach is put on the relationship between actions and objects; therefore it can be called object-based methodology. The relationship between objects, actions and events falls into a few categories:
• Object-object
• Action-object
• Action-event

The idea is to describe the relationship between objects in human and computer tasks. To develop the entity- relation-based diagram we need to:
1. Identify entities
2. Identify relationships
3. Add attributes to the relations
In the entity- relation-based analysis it is often common to list more items, and drop any unneeded later.

References
Dix, A., Finlay, J., Abowd, G. & Beale, R. (1998). Human- Computer Interaction. Prentice Hall Europe.