construction of a key to determine recidivists from non-recidivists using the mmpi

2
A NOTE ON PROFESSOR R. B. CATTELL’S METHODOLOGICAL ADUMBRATIONS 207 ject seems fated. We long ago gave examples of its application to the study, for example, of performance(9). Here again, however, if we take a leaf from Brunswik’s(l) essays it is not difficult to show that there are six representative probing-points into behavior-those concerning (i) the internal”, (ii) the “external” and (iii) the “historical”, frames of reference, each from (a) the subject’s own, or (b) the ob- server’s vantage points. American Self-psychology deals with (i) from the stand- point (b) (to judge by Snygg and Combs(?)). Classical Behaviorism, and almost all its derivatives, deals with (ii) from standpoint (b). Psychoanalysis is largely con- cerned with (iii). It happens that alone of any techniques known to us, &-technique can be used to probe into behavior from all six of these representative possibilities. R-technique, and the various applications of the so-called “objective” procedures referred to by Cattell, concern b(ii) and b(iii) at most. So that what is regarded as limited and static by our critics, turns out to be of the widest possible scope, never even approximated heretofore in any of the systematic psychologies. REFERENCES 1. BRUNSWIK, E. Systematic and representative design of Psychological experiments. In J. Ney- man (Ed.) Berkeley Symposium on Mathematical Statistics and Probability. Berkeley: Univ. Calif. Press, 1949. 2. BURT, C. The Factors of the Mind. London: London Univ. Press, 1946. 3. CATTELL, R. B. On the disuse and misuse of P, Q, Qs and 0 techniques in Clinical Psychology. J. din. Psychol., 1951, 7,3, 203-214. 4. FISHER, R. A. The Design of Experiments. London: Oliver and Boyd, 1942. 5. KENDALL, M. G. and BABINGTON SMITH, B. Factor analysis. J. Roy. Stat. Soc. (B), (London), 1950, Id, No. 1. 6. KANTOR, J. R. Psychology and Logic. Vol. I, 11. Bloomington: Principia Press, 1944, 1951. 7. SNYGG, D. and COMBS, A. W. Individual Behavior. New York: Harper Bros. 1949. 8. STEPHENSON, W. The foundations of psychometry: four factor systems. Psychometrika, 1936, 9. STEPHENSON, W. A statistical regard of preformance. Brit. J. Psychol., 1940, 30, 230-247. 1, 195-209. CONSTRUCTION OF A KEY TO DETERMINE RECIDIVISTS FROM NON-RECIDIVISTS USING THE MMPI ROBERT A. FREEMAN AND HARRY M. MASON Washington State Penitentiary Clark@) recently published a key for the MMPI intended to differentiate re- cidivists from non-recidivists. However, he did not report applications of the key to groups other than the one used in constructing it. The present note reports ap- plications of Clark’s key to MMPI responses of recidivists and first-offenders at the Washington State Penitentiary, and attempts to develop and validate further re- cidivist keys by item analysis methods. Since August, 1949, all literate, newly-admittcd, male inmates to the prison have responded to the MMPI. Tests were administered after a 30 day orientation period. By a review of the records of inmates, those having one or more previous convictions of a felony were differentiated from those being admitted to a peni- tentiary for the first time. Clark’s key was applied to MMPI responses of 60 recidi- vists and 40 first-offenders, chosen as consecutive admittees, fulfilling these defini- tions. Line 1 in Table 1 shows the results. Clearly, the key shows no differentiation. Line 2 in Table 1 shows separation of criterion groups of recidivists and first-offend- ers used in constructing a key by item-analysis procedures from WSP records. All items showed tetrachoric r’s of .40 or greater with recidivism. This yielded 14 items, only one of which (item 234, Group form) appeared in Clark’s key. Line 3 in Table 1 demonstrates shrinkage to zero validity when this key is applied to new groups

Upload: robert-a-freeman

Post on 06-Jun-2016

215 views

Category:

Documents


1 download

TRANSCRIPT

A NOTE ON PROFESSOR R. B. CATTELL’S METHODOLOGICAL ADUMBRATIONS 207

ject seems fated. We long ago gave examples of its application to the study, for example, of performance(9). Here again, however, if we take a leaf from Brunswik’s(l) essays it is not difficult to show that there are six representative probing-points into behavior-those concerning (i) the internal”, (ii) the “external” and (iii) the “historical”, frames of reference, each from (a) the subject’s own, or (b) the ob- server’s vantage points. American Self-psychology deals with (i) from the stand- point (b) (to judge by Snygg and Combs(?)). Classical Behaviorism, and almost all its derivatives, deals with (ii) from standpoint (b). Psychoanalysis is largely con- cerned with (iii). It happens that alone of any techniques known to us, &-technique can be used to probe into behavior from all six of these representative possibilities. R-technique, and the various applications of the so-called “objective” procedures referred to by Cattell, concern b(ii) and b(iii) at most. So that what is regarded as limited and static by our critics, turns out to be of the widest possible scope, never even approximated heretofore in any of the systematic psychologies.

REFERENCES 1. BRUNSWIK, E. Systematic and representative design of Psychological experiments. In J. Ney-

man (Ed.) Berkeley Symposium on Mathematical Statistics and Probability. Berkeley: Univ. Calif. Press, 1949.

2. BURT, C. The Factors of the Mind. London: London Univ. Press, 1946. 3. CATTELL, R. B. On the disuse and misuse of P, Q, Qs and 0 techniques in Clinical Psychology.

J . din. Psychol., 1951, 7 , 3 , 203-214. 4. FISHER, R. A. The Design of Experiments. London: Oliver and Boyd, 1942. 5. KENDALL, M. G. and BABINGTON SMITH, B. Factor analysis. J . Roy. Stat. Soc. (B), (London),

1950, I d , No. 1. 6. KANTOR, J. R. Psychology and Logic. Vol. I, 11. Bloomington: Principia Press, 1944, 1951. 7. SNYGG, D. and COMBS, A. W. Individual Behavior. New York: Harper Bros. 1949. 8. STEPHENSON, W. The foundations of psychometry: four factor systems. Psychometrika, 1936,

9. STEPHENSON, W. A statistical regard of preformance. Brit. J. Psychol., 1940, 30, 230-247. 1, 195-209.

CONSTRUCTION OF A KEY TO DETERMINE RECIDIVISTS FROM NON-RECIDIVISTS USING THE MMPI ROBERT A. FREEMAN AND HARRY M. MASON

Washington State Penitentiary

Clark@) recently published a key for the MMPI intended to differentiate re- cidivists from non-recidivists. However, he did not report applications of the key to groups other than the one used in constructing it. The present note reports ap- plications of Clark’s key to MMPI responses of recidivists and first-offenders at the Washington State Penitentiary, and attempts to develop and validate further re- cidivist keys by item analysis methods.

Since August, 1949, all literate, newly-admittcd, male inmates to the prison have responded to the MMPI. Tests were administered after a 30 day orientation period. By a review of the records of inmates, those having one or more previous convictions of a felony were differentiated from those being admitted to a peni- tentiary for the first time. Clark’s key was applied to MMPI responses of 60 recidi- vists and 40 first-offenders, chosen as consecutive admittees, fulfilling these defini- tions. Line 1 in Table 1 shows the results. Clearly, the key shows no differentiation. Line 2 in Table 1 shows separation of criterion groups of recidivists and first-offend- ers used in constructing a key by item-analysis procedures from WSP records. All items showed tetrachoric r’s of .40 or greater with recidivism. This yielded 14 items, only one of which (item 234, Group form) appeared in Clark’s key. Line 3 in Table 1 demonstrates shrinkage to zero validity when this key is applied to new groups

208 ROBERT A. FREEMAN h N D HARRY M. MASON

Critterion and Group Recidivist s 1st

Admissions

1. mates' records

Clark’s Recidivism key applied to WSP in-

- . p.---__--.__-

2. ari,i,lysis tet.r;tclioric on WSP inmates.

14 item recidivist> scale developed by item-

3. plied to ricxt consecutive records.

14 item recidivist scnle developed in (2) ap-

4. (26% or more ciiffercnre) 43 items.

Recidivist scale developed by item-analysis

5. sc.rut.ivi~ (‘it.SI’S.

S d c dwelopcd in (4) spplicd to next con-

Mean S.D. N

Mean S. D. N

Mean S. D. I5

Mean S. D. N

Mean S. D. K

70.1 6 . 3

80

S.4 3 4

50

8 . 6 3 . 3

6 i

27 6 5 , .5

100

10.0 5 , 6

40

7.5 3 . 5

50

similarly drawn. A final attempt at key construction, shown on Line 4 in Table 1, employed 100 recidivists and 100 first-admittees. A difference of 13%) or more in frequency of true response between the criterion groups characterized each item selected. The final line in Table 1 demonstrates the collapse of this key’s validity when applied to the next-consecutive records.

Curetoil@) has warned against assuming the validity of a test or key until its validity has becn demonstrated on subjects other than those from which the measure was derived. The present note indicates the soundness of this advice. The possibility of constructing a recidivist key for use with M.M.P.I. seems a t this point to rest with analysis of records of recidivists and parolees who have, after ample time, demon- strated law-abiding behavior. Such a procedure might serve to provide a group of non-recidivists free from contamination by individuals who have not become re- peaters only through want of opportunity. Records now being accumulated at the Washington State Penitentiary may form a basis for such an attempt at a later date.

REFERENCES 1.

2.

CLARK, JERRY H. Application of the MMPI in Differentiating A.W.O.L. Recidivists From Son-Recidivists. J . Psychol., 1945, 26, 229-234. CURETON, E. E. Validity, Reliability and etc. Educ. Psychol. Measmt. 1950, 10, 94-96.

A SIMPLE QUANTITATIVE MEASURE O F PRESSURE FOR USE IN THE PROJECTIVE TECHNIQUES

CLAUDINE GIBSON WIRTHS

University of North Carolina

This is a brief note on the use of a measuring technique and some implications of l 8 h e scores obtained by the use of it for quantifying the variable of pressure exerted by the subject in the various projective paper-and-pencil techniques. It is applica- ble to such tests as HTP, Bender Gestalt, group administered TAT and Rorschach, and it is particularly useful in the draw-a-person technique. It is in the context of this last application that the use of this measure is discussed in this note.