My tight list was: 7, 28, 57, 68:
Narcissistic: +2
Negativistic: +1
Depressed: + 1.
I've also done a looser list: 5, 7, 9, 12, 21, 22, 24, 27, 28, 34, 35, 39, 40, 41, 56, 57, 61, 63, 65, 68, 69, 71, 77, 81, 83, 86, 88, 97.
I picked out lots of ones related to social isolation that Dolphin did not (maybe that says more about me than the quiz!). These were distributed:
Paranoid: +1
Histrionic: n/a
Anti-social: n/a
OC: +2
Sc (schizoid or schizotypal): +4
Narcissistic: +3
Avoidant: +2
Negativistic: +5
Sch (schizoid or scizotypal): +2
Borderline: +1
Dependent: +4
Depressive: +3
There were lots of other questions where you'd expect CFS to influence your answer too... but a line had to be drawn somewhere. I'm sure that many people with CFS would not answer true to a lot of these questions (I know I wouldn't) - but you would expect someone suffering from CFS to be more likely to answer positively.
There's not a really clear correlation between those sections with many questions which we would expect to be influenced by CFS, although the two sections with no questions which would distort results were also two of the personality disorders which were not found to feature disproportionately with CFS patients.
With the Paranoid section we only marked 1 question as being possibly distorted, but really all but the question on your partner cheating could be affected by bad experiences with the way the medical community has treated CFS:
eg 37: Others will use what I tell them against me.
Seeing as the patients filling in this questionnaire had their answers rather misleadingly presented in a way that could create the impression that CFS is the result of personality disturbance, it may well be reasonable for them to have answered 'true' to that one. Did they know Reeves was involved when filling out the form?
I'm not sure much more can be gained from looking into this without knowing exactly which questions were answered positively for the study.