New MMPI-2 Scales by David Nichols

But you won’t find the keys to his scales in his own book! (Nichols, D. S., 2011. Essentials of MMPI-2 Assessment, 2nd Edition. Hoboken, NJ: John Wiley & Sons.)

University of Minnesota Press wouldn’t let him publish his keys. I can understand an argument for not publishing the keys to the main validity scales. Attorneys have been known to try to coach their clients on how to take the MMPI-2. (Although most fakers are not geniuses so it is very unlikely to help. I and my co-auther’s have shown that this can backfire and produce even more clearly faked MMPI-2s). Unv. of Minn. Press however argued “ownership,” not the fear of coaching. How is that for encouraging science? How can psychologists continue the tradition of developing new scales for particular needs if they only allow their own people to develop scales (such as the MMPI-RF which are some of the worst MMPI scales ever…)?

I am publishing the keys to Dr. Nichols “Cognitive Stability Scales” here to hopefully encourage more research on these scales, which I have found very useful. I have found his “Cognitive Problems” also a good measure of possible ADD- something the MMPI-2 could use.

Cognitive Stability Scales (Nichols, 2008)

The Cognitive Stability Scales, Cognitive Problems (CogProb) and Disorganization (DisOrg), seek to maximally separate item content referencing the normal-range cognitive difficulties (memory, concentration, judgment) that are commonly seen in non-psychotic disorders, especially depression, from that referencing explicit non-paranoid psychotic processes.

CogProb – Cognitive Problems (12 items)

TRUE

31 147 233 299 308 325 475 482 533 565

FALSE

165 561

Males: Mean 2.10; S.D. 2.31 Females: Mean 2.19; S.D. 2.46

DisOrg – Disorganization (11 items)

TRUE

32 60 72 96 198 298 307 319 508 551

FALSE

427

Males: Mean 1.17; S.D. 1.50 Females: Mean 1.20; S.D. 1.47

Revised D Subscales (Nichols, 2009a)

“The Harris-Lingoes subscales for Scale 2 are a mess. They are extensively overlapping, with 55 total overlaps (M = 11 overlaps/subscale), far more than for the Harris-Lingoes subscales for Scales 3, 6, 8 & 9. For any given pair of Scale 2 subscales, there will be 4-5 overlaps (range 0-12), about 27% of average subscale length. The worst offender, D5 is completely contained in D1. Eight (14%) of the Scale 2 items appear on none of the Harris-Lingoes subscales for Scale 2. The extent of overlap among these subscales therefore compromises their distinctiveness and discriminant validity.

The set of five alternative subscales (Nichols, 2009a) for Scale 2 presented below largely remedy these problems. The total number of overlaps has been reduced to six, averaging just over one per subscale, such that the average overlap for any given pair of scales is only .6 (range: 0–3), or 10% of average scale length. Moreover, each Scale 2 item appears on at least one subscale. A possible disadvantage is that two of the revised subscales are only 8 items in length (vs. a minimum of 10 items for briefest of the Harris-Lingoes Scale 2 subscales). However, it is believed that users will find that the item content of these revised subscales is much more homogeneous and distinctive than that of the Harris-Lingoes subscales.

(N.B. I removed the scoring keys of these scales on 1/30/12 after I received a letter Pearson’s attorney demanding that I remove them. Please contact Dr. Nichols for more information on these scales.)