Kaplan has a brand new course--so this seems like a great opportunity to test out our new Kaplan method!
Step 1: Identify the question type.
The word "Flaw" makes it clear what we're dealing with a flaw question (duh!). That means we can expect the standard argument elements of Evidence, Conclusion, and Assumption; however, we can expect that the assumption will be conspicuously weak.
Step 2: Untangle the stimulus.
Generally, the best place to begin untangling is with the conclusion. Fortunately, the word "concluded" identifies that for us; the psychologist's conclusion is that a 'veteran' thinks its blind while another personality responds to questions.
Why does the author think so? Because this so-called blind personality is responding to written questions.
So what's the gap between this stated evidence and the psychologists ultimate conclusion? Well, according to the doctor someone else is answering the questions--so why is that someone else answering the questions for the veteran?
Step 3: Predict an answer
The hole here is the question of why one personality is answer on the behalf of the blind personality. So, we can predict that we will look for an answer that says, "Why isn't the personality that's reading and writing just answering for itself?"
Step 4: Evaluate the choices.
(A) matches that prediction perfectly.