Mini Mult Pap

7/27/2019 Mini Mult Pap

1/7

Comparisons

of the

MMP I

nd

Mini -Mult

DAVID

L.

STREINER, CHRISTEL

A.

W OODW AR D, JOHN

T.

GOODMAN

McMaster University

N

ANTHONY McLEAN

St Joseph s H ospital

ABSTRACT

T he M M P I was compared with

two

forms

of the

Mini-Mult,

one

extra cted from

the

full

M M P I

and one

adm inistered separately. Across

six

samples

of

patie nts, correlations

be-

tween comparable scales ranged from 0.33 to 0.96,with the extracted form havinga

higher median correlation. High-point codes derived from both forms

of the

Mini-Mult

yielded poor agreement with M M P I codes, especially

for the

separately administered

Mini-Mult.

It was

concluded that

the

Mini-Mult

is not a

reliable substitute

for the

M M P I ,butcan be usedtoestimate global pathology.

Although the MMPI is the most widely used objective personality test

(Lubin, Wallis,

&

Paine, 1971; Sundberg

&

Tyler, 1962), its long length of

366 items renders it impractical for use in many situations. Depressed

patients may take as long as five or six hours to complete it, if they can

finish it at all. Patients coming to a crisis-intervention service are often

too agitated or too upset to be able to answer the questions, even if the

199 research items were eliminated. For these and other reasons, there

has been a constant, although heretofore unsuccessful, search for a shorter

form of the

MMPI.

Recently, Kincannon (1968) devised a short form, called the Mini-Mult,

in which the 71 items are administered orally. He reported a loss in relia-

bility of only nine per cent over the

MMPI,

and concluded that this loss

was not deemed sufficient to mitigate [sic] against the use of the Mini-

Mult when a standard test could not be obtained [p.319].

Further cross validation studies (Castro & Quesada, 1971; Lacks, 1970;

Lacks

&

Powell, 1970; Gayton & Wilson, 1971) have reported substantial

product-moment correlations between comparable scales of the standard

MMPI and the Mini-Mult, suggesting that the latter could predict the full

MMPI

with a high degree of accuracy. These studies used group data and

compared Mini-Mults obtained from the full scale

MMPI

with the standard

*

This study was supported in part by Crant

P . H .

167 from the Ontario Provincial

Research Council.

We would like to thank Drs PaulGrof,Arshad Majeed, and Sol Levin for making

these data available to us, and Angelo Santi and Charles Saunders for their help in

preparing the data.

Dr Streiner is also associated with the Hamilton Psychiatric Hospital, Dr Woodward

with the Chedoke Child and Family Centre, and Dr Goodman with St Joseph's Hospital.

CANAD. J. BEHAV. SCI./REV. CANAD. Sea. OOMP, 5 (1), 1973

Printed in Canada


2/7

MMPI

AND

MINI MULT 77

MMPI . Kincannon (1968) and Newton (1971) have reported that cor-

relations between standard scale scores of this internal Mini-Mult and

the MMPI are consistently higher than comparable correlations between

separately administered external Mini-Mults and the

MMPI.

AH subse-

quent studies which compared the external Mini-Mult with the MMPI

(Armentrout, 1970; Armentrout & Rouzer, 1970; Gayton, Ozmon, & Wil-

son,1972;Newton, 1971) have yielded smaller correlations than originally

obtained by Kincannon, suggesting that considerable shrinkage in ex-

plained variance has occurred with cross-validation.

Studies which compare the similarity of individual profile pairs have

indicated that the correspondence between the external Mini-Mult and

the

M M P I

is more modest than group data suggest. Armentrout and Rouzer

(1970) found very limited high point correspondence between exter-

nal Mini-Mult and

MMPI

profiles. Only 9.6 per cent of their delinquent

sample had the same two scale high points, irrespective of order. Using

a sample of college students, Armentrout (1970) reported that less than

20 per cent of the external Mini-Mult-MMPi profile pairs had the same

two highest clinical scales. In this study, the external Mini-Mult did not

predict invalid standard

MMPI

profiles well, or did it reliably predict

primed clinical scales.

The purpose of the present series of studies was to investigate further

the relationship between the MMUTand Mini-Mult high point codes, and

to compare the internally derived Mini-Mult with the separately admini-

stered form.

METHODANDBESULTS

Study 1

This study was based on 182 patients divided among four samples.

The Recurrent Affective Disorder

(HAD)

group consisted of 25 male and 36 female

patients who had been referred to the Lithium Clinic of the Hamilton Psychiatric Hos-

pital. The mean age of the males was 45.2 years S.D. = 12.90) and 43.7 years for the

females S.D. = 13.35), and they had completed an average of 10.1 years of education.

Four of the patients were diagnosed as Recurrent Depressives, 19 as Schizo-Affectives,

and 37 as Manic-Depressives, of whom 33 were tested during a free interval, two during

a depressed interval, and two during a manic interval.

The Adolescent (Adol) group was comprised of 13 male and six female inpatients

from the Adolescent Unit of the same hospital. The males had a mean age of 15.2 years

S.D.

= 0.87), and 15.5 years for the females (S.D. =

1.58).

Fifteen adolescents had

either attended or were attending regular school for an average of 8.9 years, and four

were in vocational schools. Eight were diagnosed as Behaviour Disorders of Childhood,

eight as Personality Disorders, one as Transient Situational Disturbance, one as Schizo-

phrenic, and one had no diagnosis.

The Adult group consisted of 28 fathers and 34 mothers of children who had been

referred to the Child Guidance Clinic of the Chedoke-McMaster Centre. The fathers*


3/7

78 STBEINEB, WOODWARD, COODMAN, & MCLEAN

me an age was 37.3 years (

S .D.= 8.71) and the

mo thers' was 34.9 years (S .D.

=

7.90) .

They had completedanaverageof10.7 yearsofschooling.

The Adult Inpatient

( A I )

group consisted of 16maleand 24 female patients from

the p sychiatric unitsof StJoseph's Hosp ital.Themeanage of themaleswas32.1 years

(S.D.

=

17.05)

and 32.8

years

for the

females

(S.D. =

17.94) ,

and

they

had com-

pletedanaverageof 10.4yearsofeducation. Nineteenof the patients were diagonsed

as having personality disorders, 17 assuffering from varyin g degre esof depressiveill-

ness,

andfour as schizoph renic.

In

all

cases,

the

patients completed

the

full MMPI

and the

Mini-Mult

was

derived

from the

MMPI

answer sheets.Rawscores were converted into

T

scores usingthe con-

version tab le givenbyKincannon (1 968 ) .

Pearson product-moment correlation coefficients were computed between compar-

able scales

of the

MMPI

and

M ini-Mult

for the

four groups. With

the

exception

of the

low correlationon the

Pd

scale for the Adol group (0 .33) , the correlations arecom*

parable

to

those reported

in the

previous studies, ranging from

0.61 to 0.96,

with

a

median

of

0.84.

Related

t

tests wererun foreach scale, usingthe

T

scores from the

MMPI

andMini-

Mult .For the

HAD

group,theelevationof the L scalewas significantly higheron the

MMPI

thanon theMini-Mult ( t ( 60 ) = 2.15,p


4/7

MMPI AND M INI-MULT

79

T A B L E

1

Per cent agreement

Agreement

Good

Intermediate

Discordant

TABLE

2*

Correlation;

Scale

L

F

K

s

D

y

Pd

Pa

Pt

Sc

M a

between

Group

MMI-Iand Mini-Mult high-point codes

RA D

( )

240

29.5

45.9

between the MMPI

aMMPi

versus

a M M C

0 .

0

0 .

n

0 .

0 .

0

0 .

0 .

0 .

0 .

8 5

8 3

.8 9

m

7

9 5

88

8 0

91

87

79

i

and

dMMPl

versus

dMMe

0 .

0

0

0

0

0 .

0

0

0

0

0

.83

.8 1

.90

9 2

.9 3

9 0

.86

.8 7

93

9 4

.78

Adol< )

68.4

26.3

.1.3

Mini-Mult

Adult( ) AI ( )

22.6

45.2

32.3

dMMPl

versus

dMMS

0 .

0

0 .

0 .

0 .

0 .

0 .

0 .

0 .

0 .

0 .

7 0

88

51

.68

.83

.80

6 5

7 0

93

.87

71

30.0

47.5

25.5

dMMe

versus

dMMS

0.74

0.63

0.59

0.77

0.8(5

0.78

0.68

0.78

0.93

0.90

0.85

Total{ )

29 .6

38 .4

32.1

*MMPI

= admission

MMPI; CIMMPI

= discharge

MMPI;

amie

Mini-Mult extracted from aMMPi; dMMe= Mini-Mult extracted

from dmiFi; dMMs

=

Mini-Mult given separately upon discharge.

(S.D.

11.7) and 36.7

years

(S.D. = 12.8)forthe

females. They

had an

average

of

11.0 yearsofschool(S.D.= 2.23). Four males were diagnosedashaving variousper-

sonality disorders, two hadaffective disorders, and one wasSchizophrenic. For the

females,six hadaffective disorders,onewasan Anxiety Neurotic, and sixwere given

diagnoses involving various degreesofparanoia.

Upon admission,allpatients were administered theMMPI. Just prior todischarge,

boththefull MMPIand the 71 item Mini-Mult were given. Mini-Mults were extracted

fromthe twoMMPIS.Thus, there were five "tests"foreach subject: anadmission MMPI

(aMMPi); an admission Mini-Mult extracted from it (aMMe); a discharge MMPI

(dMMPl); itsextracted M ini-Mult form (dM M e); and a discharge Mini-Mult given

separately (dMMs).

Pearson product-moment correlation coefficients were computedon theJC-corrected

scoresfor corresponding scales between theadmission MMPI and its extracted Mini-

Mult,thedischarge MMPIandbothitsextracted formandthe separately administered

Mini-Mult,andbetweentheextractedandseparate forms

of

thedischarge Mini-Mult,

andarepresented

in

T able

2. All of the

correlations were significant beyon d

the 0.05

level,

and all

except

one (K

scale

for

dMMPlvs dMMs) beyond

the 0.01

level.

The

correlations between

the

MMPI

and the

respective extracted form

of the

Mini-Mull

were higher than

the

correlations betw een

the

MMPI

and the

separately administered

Mini-Mult (Mon0.87 foraMMPiusaMMe;0.90 fordMMPlvsdMMe;and 0.71 for

d M M P l OSd MMS

).


5/7

80 STKEINEB, WOODWARD, GOODMAN, & MCLEAN

A seriesofrelatedttests wererunbetw een corresponding scalesoftheMMPIS versus

the extracted Mini-Mults.TheTscores forscalesFand

Ma

were significantly higher

onthe

aMMPi tha n

onthe

aMMe

(t(19) =

2.092,

p = 0.05 andt 19) =

2.258,

p

Mini Mult Pap

Documents

Transcript of Mini Mult Pap