Which version? The current, I believe, is the MMPI-II (which I've taken). I don't have my T-Scores, but my Welsh Code was 0'8+25-74/613:9#K-LF/ with a 'Fake Bad' T-Score of 56. There's more to it, but the Welsh Code order the clinical scales (numbers) and validity scales (letters) in order from high T-Score to low, with clinical and validity scales separate.
Yeah, it's basically correlations - people with x-y score tend to have a,b, and c characteristics. But a large part of the test is also designed to detect elements such as lying (to the test taker and to yourself), strange inconsistencies, people trying to get a certain diagnosis dishonestly, and a bunch of other stuff (that's why some of those questions seemed so simple: they're designed to find people randomly ticking off answers or trying to appear psychotic even when those with the diagnosis would not [generally] agree).
As such, it's meant to be a guide to diagnosis (not definitive), and to help the examiner determine/confirm if something fishy is going on. Unfortunately (in some ways), the test can now be scored electronically, which means a psych may take the easy route of just believing the computer output instead of properly fitting it with the other patient data (eg, someone may have a high 2 score ['depression'] because of an illness, not a psych issue).
-------------------------------------------------------------------------------------
I did some work and decoded mine to make it easier to understand:
Code:
Welsh Code
===============================================
No. T d<1 Short Long Name
0: 70-79 - Si Social Introversion
8: 65-69 - Sc Schizophrenia
2: 62-64 - D Depression
5: 60 | MF Masculinity/Femininity
7: 59 | Pt Psychasthenia
4: 50-57 - Pd Psychopathic Deviate
6: 40-49 | Pa Paranoia
1: 40-49 | Hs Hypochondriasis
3: 40-47 - Hy Hysteria
9: 30-39 - Ma Hypomania
K: 60-64 - n/a Defensiveness
L: 58-59 | n/a Lie
F: 56 | n/a Infrequency
===============================================
T-Scores:
Mean: 50
SD: 10
Note that the names are not quite accurate, and are remnants from the origins of the test. The No. and Short Name versions are correct to use in practice.
You can read more on it
here.