QCing blinded SUBJIDs

[Dregerator]

Hi, I'm trying to QC blinding of the data to ensure it is done correctly. I have the base data where the SUBJID's are not scrambled and the scrambled datasets where the SUBJIDs are scrambled. I do not know what the algorithm was used so how do i go about checking to see if they're done correctly? For example:

The Normal SUBJIDs are as follows

10045001
10055001
10055005
10055004
10055003
10055006
10055002
10075001
10075002

Blinded:

50021001
50021002
50031001
50031002
50041001
50041002
50051001
50051002
50061001

So would I need to create a unique sequence in one of the normal dataset and trace that sequence to the blinded dataset to ensure the Blinding is done properly? Or would I need to ask the principal programmer what the algorithm is for blinding...or would that defeat the purpose of double programming?

Thank you in advance.

[ChrisNZ]

1. Am I dreaming or were the IDs just split in half, and the two halves' positions reversed?

2. how do i go about checking to see if they're done correctly? 

What does "correctly" mean? What's the expected outcome?

2. Scrambling IDs is normally done with a function such as SHA256.