Planning a Test.

The main objective of classroom assessment is to obtain valid, reliable and useful data regarding student learning achievement. This requires determining what is to be measured and then defining it precisely so that assessments tasks to measure desired performance can be developed. Classroom tests and assessments can be used for the following instructional objectives:

i. Pre-testing
Tests and assessments can be given at the beginning of an instructional unit or course to determine:-
 weather the students have the prerequisite skills needed for the instruction (readiness, motivation etc)
 to what extent the students have already achieved the objectives of planned instruction (to determine placement or modification of instruction)
ii. During the Instruction Testing
 provides bases for formative assessment
 monitor learning progress
 detect learning errors
 provide feedback for students and teachers
iii. End of Instruction Testing
 measure intended learning outcomes
 used for formative assessment
 provides bases for grades, promotion etc
Prior to developing an effective test, one needs to determine whether or not a test is the appropriate type of assessment. If the learning objectives are of primarily types of procedural knowledge (how to perform a task) then a written test may not be the best approach. Assessment of procedural knowledge generally calls for a performance demonstration assessed using a rubric. Where demonstration of a procedure is not appropriate, a test can be an effective assessment tool.
The first stage of developing a test is planning the test content and length. Planning the test begins with development of a blueprint or test specifications for the test structured on the learning outcomes or instructional objectives to be assessed by the test instrument. For each learning outcome, a weight should be assigned based on the relative importance of that outcome in the test. The weight will be used to determine the number of items related to each of the learning outcomes.

Test Specifications
When an engineer prepares a design to construct a building and choose the materials, he intends to use in construction, he usually know what a building is going to be used for, and therefore designs it to meet the requirements of its planned inhabitants. Similarly, in testing, table of specification is the blueprint of the assessment which specifies percentages and weightage of test items and measuring constructs. It includes constructs and concepts to be measured, tentative weightage of each construct, specify number of items for each concept, and description of item types to be constructed. It is not surprising that specifications are also referred to as ‘blueprints’, for they are literally architectural drawings for test construction. Fulcher & Davidson (2009) divided test specifications into the following four elements:
 Item specifications: Item specifications describe the items, prompts or tasks, and any other material such as texts, diagrams, and charts which are used as stimuli. Typically, a specification at this sub-level contains two key elements: samples of the tasks to be produced, and guiding language that details all information necessary to produce the task.
 Presentation Model: Presentation model provides information how the items and tasks are presented to the test takers.
 Assembly Model: Assembly model helps the test developer to combine test items and tasks to develop a test format.
 Delivery Model: Delivery Model tells how the actual test is delivered. It includes information regarding test administration, test security/confidentiality and time constraint.