ASSESSMENT AND EVALUATION OF EXCEPTIONAL CHILDREN

Assessment and Evaluation of Exceptional Children

Assessment and Evaluation of Exceptional Children

Reliability

Reliability is the consistency of your assessment, or the point to which an tool evaluates the matching way each time it is employed under the matching relative standing with the matching subjects. In short, it is the repeatability of your measurement. A evaluate is deliberated consistent if a person's count on the matching investigate bestowed double is similar. It is valued to consider that reliability is not evaluated, it is estimated. There are two means that reliability is commonly estimated: test/retest and inside consistency.

 

Test/Retest

Test/retest is the more very protected manner to assess reliability. Simply put, the thought behind test/retest is that you should get the matching count on investigate 1 as you perform on investigate 2. The three principle elements to this manner are as follows:

1) realize your assessment tool at two split times for each subject; 2) compute the relationship between the two split measurements; and 3) presume there is no change in the implicit relative standing (or trait you are attempting to measure) between investigate 1 and investigate 2.

 

Internal Consistency

Internal consistency assesses reliability by grouping queries in a questionnaire that evaluate the matching concept. For instance, you could draft two collections of three queries that evaluate the matching fundamental thought (say class participation) and after accumulating the replies, run a relationship between those two gatherings of three queries to ascertain if your tool is reliably evaluating that concept.

One universal way of computing relationship ideals amid the queries on your tools is by employing Cronbach's Alpha. In short, Cronbach's alpha separates all the queries on your tool every probable way and computes relationship ideals for them all (we use a computer program for this part). In the end, your computer end wares engenders one number for Cronbach's alpha - and just like a relationship coefficient, the nearer it is to one, the higher the reliability assess of your instrument. Cronbach's alpha is a less very protected assess of reliability than test/retest.

The main divergence between test/retest and inside consistency assesses of reliability is that test/retest includes two administrations of the assessment tool, where the inside consistency manner includes only one authorities of that instrument.

 

Validity

Validity is the might of our resolutions, inferences or propositions. More formally, Cook and Campbell (1979) delineate it as the "best obtainable approximation to the certainty or falsity of a bestowed inference, proposition ...