and

essential for quality of questionnaire

. You need to provide

for creating a new test. Whether it is specification and/or

of old/new constructs, a current test becomes

, or the constructs utilised are

for clinical use. The next thing is to conceptualise and

the construct/s you want to measure. A conceptual definition explains the construct on a

level,while an

definition shows exactly how your test will define or

your construct.

range for the measurement, (can affect features to use), as well as describe the testing

. There are many types of testing formats - you will need to decide the

used - eg visual cues? verbal?. After you've chosen, write out

items - always write

as many as you need!

Studies.

Specify a sampling plan. Describe the target

for the test. The standardisation sample for a test determines the

and forms of the

group to whom the particpants are compared to. During this phase you must also outline the reliability and validity studies to be taken and your rationale

Consider the results of your validity and reliability checks. Be willing to make

based on your results. When it all checks out, begin to prepare your

manual; the manual includes a description on the test

process, so others can replicate your results and evaluate your findings

over repeated measures - test-retest ability.

consistency refers to the

of a specific single test form

- the degree to which multiple items on a test assess the

construct

. There are several different ways to test/measure validity.

-related validity is used to establish whether a measure is valid by comparing with a measure that is already known to be

. When a criterion measure (one thats already been proven) is measured at the same time as the potential measure, this is testing for

validity. When the criterion measure is taken later, the goal is to establish

validity.

When you test for

validity you are making sure your item doesn't accidentally measure another construct by seeing if it realtes to measures of other constructs

validity refers to whether the actual content of a test or item is representative or matches with each domain of the represented construct