QuBAS analyses data simultaneously in two separate analysis engines. These engines are independently developed by two different statisticians in different programming languages. On completion, the results from the two engines are compared using a DSCPP process, and only if all agree is a report produced.
Since the engines are independent of each other, the chances of an error in one branch producing results which match those from the other are extremely small. If the results match, therefore, it can be safely assumed that those results are statistically robust.
This process validates that Engine A is installed, performing and operating correctly by comparing with Engine B, and vice versa. With CrtV, validation is completed with each and every analysis, so there is no need for IQ, OQ or PQ.
As well as validating calculations, QuBAS uses a similar DSCPP process to validate audit trail, user inputs and report generation from the results.