For this assignment, I have chosen School Motivation and Learning Strategies Inventory test and the Attention Test Linking Assessment and Services test. The reliability of the SMALSI test can be measured through retest reliability by repeated testing on the same sample of subjects (standardization samples). It is noticed that with increasing correlation, indicators tend to decrease. This is due to the fact that the measured property is unstable, and age-related changes may occur, as well as events that affect the state of the studied qualities. However, in some cases, a repeat test is carried out after a long time interval, for example, in order to assess the prognostic validity. Thus, the SMALSI test requires repeated observation at certain stages of life in order to provide the most reliable results.
The reliability quality of the ATLAS test can be observed through internal consistency reliability. It is determined by “splitting” the test into two parts and calculating the correlation coefficient between the estimates obtained for each of these parts. The rationale for such a procedure is the provision that, with a normal distribution of scores on a complete test, performing a random set of tasks from parts of the test gives a similar distribution. It can be concluded that the ATLAS test is more reliable than the SMALSI one due to the more prominent mechanism of proving the results valid.
In order to carry out the pragmatic validation of the methodology – to assess its effectiveness, efficiency, and practical significance, an independent external criterion is used – an indicator of the manifestation of the studied property in everyday life. In the case of the SMALSI and ATLAS tests, I would rate their validity as construct and content validity types, respectively. Construct validity of SMALSI is related to the theoretical construct itself and includes the search for factors that explain the behavior when performing the test. Content validity of ATLAS requires that each task or question belonging to a particular field has an equal chance of becoming a test task. It evaluates the compliance of the content of the test with the measured area of behavior. Tests compiled by two development teams are conducted on a sample of test subjects. The reliability of the tests is calculated by splitting the tasks into two parts, as a result of which an index of meaningful validity is obtained. I think that the reliability of the mentioned tests will provide valid scores for the population.