CogniFit has been reviewed and tested for reliability and validity. You can see the results below.
ValidityAs previously mentioned, the battery of tasks for each assessment is made up of various tasks that can measure up to 23 cognitive skills. This set of cognitive abilities has been subjected to a standardized measurement control to check validity by using several statistical measurements.
1. Cronbach's Alpha
To consider the internal consistency and check the compatibility of CogniFit's tool, the Alfa Cronbach coefficient was used. The said statistic was calculated by using the data from a sample group of over 500 participants. The data were gathered by using the general cognitive assessment tasks and the set of training programs available at CogniFit (www.cognifit.com). The Alfa Cronbach aimed to measure reliability and correlation in each of the cognitive skills.
As you can see in the table of the following page, the data gathered reached close to .8, indicating that the Alfa score had a high/good internal consistency, according to the criteria specified by George and Mallery (2003, p. 231). These results show that the correlation between the variables is high, and thus good reliability.
2. Test-RetestIn order to check the reliability of the CogniFit tool as a system, the Test-retest reliability was applied. This test was chosen to prove the stability of the measured data at two specific points in time. This statistic was calculated using the data gathered by other 500 sample users using CogniFit.
As you can see below, the data are above .8 in more than 50% of the cases, and the rest are between .6 and .7. This result shows that the CogniFit system is accurate and reliable, as the scores are near 1, and therefore there are no discrepancies among the data.
|Cognitive Skill||Internal Consistency||Test-Retest Reliability|
|Width Field of View||0,806||0,998|
|Auditory Short-Term Memory||0,915||0,698|
|Visual Short-Term Memory||0,866||0,743|
|Reaction To Change||0,571||0,88|