Creating A Psychological Test

Creating A Psychological Test

Item analysis

Research question(s) or hypotheses

The main research question is addressing reliability as well as the validity concerns which are on the level of the item. it entails the way these two notions can be used in guiding in decisions pertaining the test items that will be kept, the items that need to be changed and the items that will be substituted or abandoned altogether.


There are three main constructs that will be used and they are item difficulty, item consistency and item discriminability. These three construct deal with aspects of the item that are of great importance in their own right and they are also connected to each other in one way or another. However assessing these three constructs of tests alone can not give sufficient information to make a decision on the fate of individual items. Therefore the balance of difficulty, dicriminability and consistency of n item is difficult task. These three constructs were arrived at in different ways. The concept of difficulty is a very common one hence this was an easy construct to come up with. This aspect was chosen since every item is designed with specific preferences, beliefs, attributes and behavioral self reports in mind and therefore there can be difficulty arising pertaining items (Key,2010).

The second construct is consistency which involves assessment of how well an item is able to hang together. This construct was arrived at since the approach can be used as a test for other items as opposed to a hypothesis that the overall reliability of an instrument is not in any way affected by a particular item. The third construct is item discriminability, and just as item reliability this construct pays for researchers that are beginning to try and ensure that they familiarize themselves with the literature that pertains to the research and therefore they can maximize on the validity of each item that is being tested.

Tests for measuring constructs

It is very important to note that each construct contributes to the entire scene and therefore they are worth being tested independently. There are various tests that are used to measure the three different constructs for item difficulty, the level of difficulty can be tested and expressed as an average based on the criteria of response that have been received pertaining the difficulty of the item. Item reliability can be test using two approaches incase an item has been deleted; are recomputing an instrument reliability in an undefined number of times and each time deleting different items from the instrument. The second approach is of estimating how reliable a particular test is at the item level. Therefore the consistency can be tested through obtaining an interitem consistency and it is necessary to obtain an average correlation with each item in a particular instrument (Shuttleworth ,2009).


Key.P.J.(2010).Reliability and validity.Retrieved April 4,2013 from,M.(2009).construct validity.Retriebed april 4,2013 from