Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Another Look at The Mini-Mult

Download as pdf or txt
Download as pdf or txt
You are on page 1of 39

Western Michigan University

ScholarWorks at WMU

Masters Theses Graduate College

8-1972

Another Look at the Mini-Mult


Dianne S. Umansky
Western Michigan University

Follow this and additional works at: https://scholarworks.wmich.edu/masters_theses

Part of the Psychoanalysis and Psychotherapy Commons

Recommended Citation
Umansky, Dianne S., "Another Look at the Mini-Mult" (1972). Masters Theses. 2831.
https://scholarworks.wmich.edu/masters_theses/2831

This Masters Thesis-Open Access is brought to you


for free and open access by the Graduate College at
ScholarWorks at WMU. It has been accepted for
inclusion in Masters Theses by an authorized
administrator of ScholarWorks at WMU. For more
information, please contact wmu-
scholarworks@wmich.edu.
ANOTHER LOOK AT THE KI3I-KULT

ty
Dianne S. Umansky

A Thesis
Submitted to the
Faculty of The Graduate College
in partial fulfillment
of the
Degree of Master of Arts

Western Michigan University


Kalamazoo, Michigan
August 1972

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOVi LEDGEMENTS

In writing this thesis, I have benefited from the

encouragement, advice, and constructive criticism of Dr.

Malcolm Robertson, Dr. E. Huitema, and Dr. George Sidney.

My thanks go to them as to the staff at William Upjohn

DeLano Memorial Clinic, especially Mr. Gordon Gradwell

and Mr. John Gallagher, who have given much needed help.

Special thanks also, to my husband whose encourage­

ment, support, and advice helped make this thesis a reality..

Dianne Sue Umansky

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
INFORMATION TO USERS

This dissertation was produced fro m a m icrofilm copy o f the original docum ent.
W hile the most advanced technological means to photograph and reproduce this
document have been used, the q u ality is heavily dependent upon the quality of
the original subm itted.

T he follow ing explanation o f techniques is provided to help you understand


markings or patterns which may appear on this reproduction.

1. The sign or “ target" fo r pages apparently lacking from the document


photographed is "Missing Page(s)". If it was possible to obtain the
missing page(s) or section, they are spliced into the film along w ith
adjacent pages. This may have necessitated cutting thru an image and
duplicating adjacent pages to insure you com plete con tin u ity.

2. When an image on the film is obliterated w ith a large round black


mark, it is an indication th a t the photographer suspected th a t the
copy may have moved during exposure and thus cause a blurred
image. You will find a good image o f the page in the adjacent fram e.

3. When a map, drawing or chart, etc., was part o f the material being
p h o to g ra p h e d the photographer follo w ed a d efin ite m ethod in
"sectioning" the material. It is custom ary to begin photoing at the
upper le ft hand corner o f a large sheet and to continue photoing from
left to right in equal sections w ith a small overlap. If necessary,
sectioning is continued again - beginning below the firs t row and
continuing on until com plete.

4. T h e m ajo rity o f users indicate that the textual content is o f greatest


value, however, a som ewhat higher q u ality reproduction could be
made from "photographs" if essential to the understanding o f the
dissertation. Silver prints of "photographs" may be ordered at
additional charge by w riting the O rder D epartm ent, giving the catalog
number, title , author and specific pages you wish reproduced.

University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 48106

A Xerox Education Company

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MASTERS THESIS M-3657

UMANSKY, Dianne Sue


ANOTHER LOOK AT THE MINI-MULT.

Western Michigan University, M.A. , 1972


Psychology, clinical

University Microfilms, A XEROX C om pany, Ann Arbor, Michigan

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PLEASE NOTE:

Some pages may have

ind ist inet print.

Filmed as received.

University Microfilms, A Xerox Education Company

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS

Page
List of tables ii

Introduction 1

Method 7
Results 9
Tables 13
Figure: Mean T Scores 20

Discussion 21

Appendix . 29
References 31

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES

Table Fa^e
1 Froduct-moment Correlations between
Comparable Scales for Three Groups:
Total, Male, and Female 13
2 Mean and Standard Deviations of the
Scale Scores for the Three Groups:
Total, Kale, and Female 14

3 Mean Differences between Comparable


Scale Scores of the MKFI and Mini-
Mult for Three Groups: Total, Male,
and Female 15
4 Values of t_ and Estimate of Standard
Error of Differences between Comparable
Scales of the MKFI and the Mini-Mult
for Three Groups: Total, Male, and
Female 16

5 Comparison of the MMFI and the Mini-


Mult with Respect to the Ordinal
Position of the Three Clinical Scales
Highest in Hank 1?
6 Comparison of the MKFI and the Mini-
Mult with Respect to the Ordinal
Position of the Three Clinical Scales
Highest in Rank for Males 18

7 Comparison of the MMFI and the Mini-


Mult with Respect to the Ordinal
Position of the Three Clinical Scales
Highest in Hank for Females 19

ii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
INTRODUCTION

The purpose of this study was to evaluate the rela­

tionship between the standard Minnesota Kultiphasic

Personality Inventory and the recently developed 71-item

Mini-Mult (Kincannon, 1968),.

Since its introduction by Hathaway and McKinley in

19^0, the Minnesota Multiphasic Personality Inventory

(MMFI), has been established as a clinically useful and

popular self-report inventory. However, many investiga­

tors agree that the length of the test, both in terms

of number of items and time required to complete the

inventory, seriously restrict its application to clinical

and research projects.

To overcome the restrictions imposed on the MMFI

by its length, several attempts have been made to develop

an abbreviated form from which the standard MMFI scale

scores could be reliably predicted (e.g. Poulds, Caine,,

and Creasy, I960; and Jorgenson, 1958). Lacking predic­

tive accuracy, these forms failed to achieve acceptance.

However, a 71-item short form of the MMFI, the Mini-

Mult, introduced in 1988 by J. Kincannon, was specifi­

cally designed to predict all the standard validity and

clinical scales, and may prove to be more acceptable and

have greater potential usefulness than earlier abbre­

viated forms. Initial investigations indicated a high

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2

degree of correspondence between the Mini-Mult and the

MMFI.

Lacks (1970) scored MMPI answer sheets of 9^- in­

patients at an acute, urban, intensive treatment center

for both the standard MMFI and the scales of Kincannon* s

7 1 -ltem Mini-Mult.. The results of correlational and

clinically relevant comparisons supported Kincannon*s

findings that the Mini-Mult predicts results of the

MMFI with a high degree of reliability.

In another study, Lacks and Powell (1970) compared

the Mini-Mult and the MMPI for a group of hospital atten­

dant applicants. Data were collected from the personnel

files of 20 males and 20 females who had applied for em­

ployment during 1967 -6 8 , as psychiatric attendants at a

Mental Health Center. The results of the study supported

previous work that the Mini-Mult is a reliable substitute

for the standard MMFI*. For the total sample, statistically

significant differences were found for scales Pa, Pt,

and Ma. Correlations between the two forms for the dif­

ferent scales ranged from ..65 to .-9 0 .

Armentrout and Rouzer (1970) investigated the accu­

racy with which the Mini-Mult could predict features of

the standard MMPI in a nonpsychiatric population of in­

stitutionalized adolescents. The subjects were 100

male and 25 female delinquents between the ages of 13

and 19 who were tested at a residential diagnostic center*.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The Mini-Mult and the MMPI were administered to each

subject and were completed in that order with a 2k-48

hour intertest interval. Group results showed good cor­

respondence between the Mini-Mult and MMFI scale scores*

However, an examination of individual profile pairs sug­

gested that, for these subjects, the Mini-Mult showed

little correspondence regarding validity, high points,

or general elevation* Therefore, they suggested that

the results of their study were inconclusive*

The problems inherent in creating shorter versions

of longer tests have interested many researchers (Borgatta,

196^; and Silverstein, 1965)• The major problem in

creating a shortened instrument is the assumption, gen­

erally accepted by many investigators, that a longer test

is significantly more reliable, and therefore, poten­

tially more valid than a short form. As illustrated by

Kincannon (1968) this assumption is mathematically ex­

pressed in the Spearman-Brown formula which is based on

two further assumptions. These assumptions, that all

items in a scale are equivalent and that any deletions

of items would be made on a random basis, need not be the

case, according to Kincannon (1968). Indeed, Kincannon

proceeds to illustrate that these assumptions need not

be applied in this case by citing references which have

demonstrated that the MMPI scales are quite heterogen­

eous and by proposing a systematic procedure for item

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
k
reduction, which he followed in developing the Mini-

Mult. Ey comparing the 28% average loss in reliability

predicted by the Spearman-Brown formula with the 9%

average loss in reliability predicted by the results of

his test-retest comparisons, Kincannon was able to sup­

port his proposition.

Kincannon*s study consisted of Mini-Mult and MMPI

comparisons with three sample populations. The first

comparison was conducted on the MMPI answer sheets for

a sample of 50 male and 50 female recent admissions to

the psychiatric service of a city-county general hospi­

tal* Product-moment correlations between these two sets

of scores ranged from ^80 to .93• A second comparison

was carried out on the MMPI answer sheets of 25 male and

25 female consecutive admissions to the local community

mental health center. The product-moment correlations

between these two sets of scores were essentially identi­

cal to those of the inpatient group and ranged from *70

to .96. For the third and most comprehensive comparison,

he used a test-retest design with 30 male and 30 female

new admissions to the acute psychiatric service of a gen­

eral hospital. Each subject in this group was requested

to complete a retest of the standard MMPI and take an

independently administered Mini-Mult in an alternating

sequence with the MMPI retest. The product-moment cor­

relations between the first standard administration of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
the MMFI and the Mini-Mult extracted from these MMPI

answer sheets ranged from .82 to .96. Comparisons be­

tween comparable scales for all combinations of the two

administrations of the two forms revealed a high degree

of correspondence between the independently administered

Mini-Mult, the extracted Mini-Mult, and the standard

administration of the MMPI. The £ tests for paired com­

parisons of means for the Mini-Mult and the MMPI showed

statistically reliable differences for scales F, Hs, and

Ma in the comparison of the first standard administration

of the MMPT and the Mini-Mult which was extracted from

those answer sheets.

Research for the present study was conducted on

MMPT answer sheets for outpatients of an adult outpatient

psychiatric clinic. In this study, scale scores of the

standard administration of the MMPI and the Mini-Mult

extracted from the MMFI answer sheets were compared.

The following hypothesis was postulated:

There are nc statistically significant


differences between the corresponding scale
scores of the standard MMFI and the Minl-
Mult extracted from the MMFI answer sheets.

This study was not intended to replicate Kincannon*s

(1968) research. Although some parts of the research

were similar, the material was, in.general, different.

Mo attempt was made to employ a test-retest design.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Also, this study employed an actuarial method for in­

vestigating profile correspondence, rather than the

ratings of clinical psychologists. Further, reliability

of individual scale scores was not evaluated. However,

a comparison of the results of this study with a recently

completed study on a similar sample will provide some

measure of scale reliability.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
METHOD

The subjects for this study were obtained from the

1968 files of outpatients at an adult outpatient psychi­

atric clinic. Subjects were consecutively selected ac­

cording to alphabetical sequencing. Selection began at

the beginning of the alphabet and continued until files

were obtained for 50 males and 50 females who had com­

pleted an MMFI. The age range for this sample was 16

to 60 with a mean age of 30.71 for males and 3 2 .2 ^ for

females. However, 5&% of the male group was clustered

within a 16 to 30 age range and 5 of the female group

was clustered within this same 16 to 30 range. There­

fore, based on age distribution, this is not a truly

representative sample.

All standard MKFI’s were machine scored and K cor­

rected. The Mini-Mult for all subjects was then scored

with templates derived from Klncannon's Mini-Mult and the

raw scores obtained were converted into standard scale

raw scores using Kincannon*s conversion table. Two sets

of standard K corrected raw scores for eleven validity

and clinical scales (L, F, K, Hs, D, Hy, Fd, Fa, Ft,

Sc, and Ma) were obtained for the standard MMFI and the

Mini-Mult extracted from the MMFI answer sheets. The

scores of the two forms were then compared to determine

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8

the degree of similarity.

Since contemporary use relies heavily on the con­

figuration or profile of scores, an additional analysis

focused on the stability of the three highest scores,

in comparing the short and long forms. For each subject,

K corrected profiles were plotted on standard profile

sheets for the standard'MMFI and the Mini-Mult. Three

point codes based on these profiles were tabulated in

the manner described by Lichtenstein and Eryan (1 9 6 6 ).

Further analysis of profile correspondence evaluated

similarity in elevation of profile pairs by comparing

the occurrence of clinical scales above or below a T

score of 7 0 .

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RESULTS

For this study two sets of scores on the validity

and eight clinical scales were obtained for each subject,

the standard MMPI (St) and the Mini-Mult extracted from

the MMPI answer sheets (MM),

Product-moment correlations between the comparable

soales of the standard MMPI and the Mini-Mult ranged

from *26 to .34 with a median of .75* For this total

group all correlations were significant at the .01 level.

For males the correlations between comparable scales for

both forms of the test ranged from .1 3 to .8 5 with a

median of .77* For this group correlations were signi­

ficant at the .01 level for all scales except Hs and

Hy. Neither of these scales was significant at the .0 5

level and only the Ey scale was significant at the .10

level. For females the correlations ranged from .33

to .8 5 with a median of .66 (see Table 1). These cor­

relations were significant at the .01 level for all scales

except the Hs scale which was significant at the .02

level.

The means and standard deviations for each scale

are presented in Table 2. In most cases, the standard

deviation was smaller for the Mini-Mult than for the

standard MMPI.. When t tests of mean differences for

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10

comparable scales on the MMPI and Mini-Mult were cal­

culated, several significant differences were found

(see Tables 3 and *0. For the total group, significant

differences (p<01) between means occurred for the F,

K, Hs, Pa, and Pt scales. The female group showed mean

differences significant at the ,01 level for scales K,

Hs,. D, Pt, and Ma, Scales F and Hs were the only scales

found to have significant differences at the ,01 level

for the male group.

Using K corrected T scores, the degree of corre­

spondence between comparable scales of the MMPI and

the Mini-Mult was assessed in a manner more relevant for

contemporary clinical judgment. Profiles of mean

K corrected T scores for males and females are presented

in Figure 1, Evaluating profiles as invalid when either

scale L, F, or E was equal to or above a T score of

70, of the total group had invalid MMPI’s and 23 of

these had invalid Mini-Mults*. There were only 2 cases

where the Mini-Mult was invalid and the MMPI was valid.

For males, 16 of 50 (3%%) bad invalid MMPI*s. Of these

16, only 6 had invalid Mini-Mults. This group had 2

cases where the MMPI was valid and the Mini-Mult was not.

Of the 50 females, 26 (52%) had invalid MMPI*s. Seven­

teen of these 26 also had invalid Mini-Mults, There

were no cases in the female group where the MMFI was

valid and the Mini-Mult invalid. Bevaluating the pro­

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
files as invalid when scale L was equal to or above

a T score of 60,. or scale F was equal to or above a

T score of 80, and leaving K unrestricted, invalid MMFI*s

were obtained for 29 of the total group. Of these 29*

there were 15 invalid Mini-Mults.. Also, the Mini-Mult

was invlaid in 7 oases where the MMFI was valid. In­

valid MMFI*s occurred for 13 {26%) of the 50 males. Of

these 13, only 4 Mini-Mults were invalid. In 3 {&%)

of these 50 cases, the MKFI was valid and the Mini-Mult

invalid. Of the 50 females, l6 {32%) had invalid MMFI*s.

Among these 16, 11 had invalid Mini-Mults.. In 4 cases

of the 50, the MMFI was valid but the Mini-Mult was not.

For further analysis, K corrected profiles were

plotted on standard profile sheets for the standard

MMFI and the Mini-Mult for each subject.. Three point

codes based on these profiles were tabulated in the man­

ner described by Lichtenstein and Bryan (1966). These

results are presented in Tables 5» 6, and 7» The pro­

bability that the top ranked score will remain among

the top three is 75% for the total group, 7%% for males,

and 72% for the female group.. The probability that the

second or third ranked score would fall outside the three

point code is b$% and b6% respectively for the total

group* Equivalent probabilities occurred in both male

and female groups..

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12

Additional analysis of profile correspondence con­

sidered similarity In elevation of profile pairs by com­

paring the occurrence of clinical scales above or below

a T score of 70.. For the total group, 13 of the profile

pairs had no scales above a T score of 70 on the MKFI.

Of these 13 pairs, only 2 had no scales above 70 on the

Mini-Mult. There were 5 cases with no scales above a

T score of 70 on the Mini-Mult and only one of these had

no scale above 70 for the MMFI. Only 12$ of the total

group had the same scales above 70 on both tests. Eight

profile pairs for the male group had no scales above a

T score of 70 on the MKFI and only one of these had no

scale above 70 on the Mini-Mult. In this group, 8 pairs

had the same scales above 70 on both tests. Profile

pairs having no scales above a T score of 70 on the

MMFI occurred in 5 of the 50 cases in the female group.

Of these 5» only one had no scales above 70 for the

Mini-Mult. There were 5 cases for this group where no

scale above 70 occurred on the Mini-Mult, only one of

which had no scale above 70 for the MMFI.. In this group

there were ^ pairs which had the same scales above 70

on both tests*

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13

TAELE 1

Product-moment Correlations "between


Comparable Scales for Three Groups:
Total, Male, and Female

Scale Total Male Female

L .76 .85 *66

F .30 .85 .79


K *84 .84 -85
Hs .2 6 -13** -33*
D .78 .77 .85
By .48 .2 6 * *60

Fd .74 *68 .62

Pa •75 .82 .74

Pt .82 .84 *80

Sc -67 .74 -59

Ma -55 .48 -51

Median r Median r Median r

.75 .77 .6 6

p < .0 1 p < .0 1 p < .0 1

*p< .1 0 ♦p < •02

♦♦not sig­
nificant

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

TABLE 2

Mean and Standard Deviations of the Scale Scores for the


Three Groupsi Total, Male and Female

Total Male Female

Standard Mini--Mult Standard Mini--Mult Standard Mini-Mult

Scale M S.D.- M S.D.. M S.D . M S.D. M S »D — M S.D.

L 4.02 2.2? 4-02 2..17 3.82 2-36 4.12 2-44 4.22 2-18 3.92 1.89

F 10.68 7*31 8.82 4.96 11-88 8-44 9.72 5.12 9.46 5-81 8.26 4.41

K 11.87 5-08 11.07 4.40 11-74 5.54 11.44 4.80 12.00 4.63 10.70 3.96

CM
CM

CM
HS 6.30 17*42 21.10

0
18.56 21.65 3.59 5.90 3-41 19.70 6.53 3-71

.
D 29*88 7.42 29.33 5.87 26.78 6.99 28.46 5.48 32.98 6 .5 6 30.20 6-18

Hy 28.18 6.42 28.29 4.15 26.45 6.24 27.46 5*68 29.92 6.18 28.58 4.25

Pd 29.18 6 .0 6 28.21 5.12 29.38 6.09 28.58 5.35 28.98 6.01 27-39 6.25
Pa 14.06 4.45 14.93 4.15 13.62 4.99 14.24 4.43 14.70 4.07 15.62 3-76

Pt 36.25 7*64 34.57 7.14 34.48 8-05 33-72 7.41 38.02 6.83 35.42 6.82

So 38.05 10.B9 37.47 8.79 37.10 12.55 37.86 9.30 39.00 8.98 36.96 8-21

Ma 20.29 4.47 20.94 3.29 21.22 4.39 20.54 4-16 1 9 .3 6 4.38 20.94 3.62
TABLE 3

Mean Differences between Comparable Scale


Scores of the MMFI and Mini-Mult for Three
Groups: Total, Male, and Female

Scale Total Male Female

L *00 *30 .30

F 1*86 2 .1 6 1*20

K *80 .30 1.30

Hs 3.09 3*68 2 *5 0

D -55 1.68 2*78

1*34
O

Hy .11
t

Fd .97 .80 1*59


Fa *87 .62 *92

Pt 1*68 .76 2 .6 0

Sc *58 .76 2.04

Ma .6 5 *68 1.58

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16

TABLE 4

Talues of t_ and Estimate of Standard Error of


Difference between Comparable Scales of
the MMPI and the Mini-Mult for Three
Croups: Total, Male, and Female

Total Male Female

Scale t sdiff t sdiff £ sdiff

L 0.00 .173 1.76 .17 M ►66

F 4.13* .^5 3.13* .69 2 .40** .50


00
K 3.01* .265 *77 -0 3.94* .33
.

Hs 83* .64 4.04* .91 2 .81 * .89

D 1.17 .47 2 .67 ** .63 5 .68 * .^9

Hy *19 *57 *99 1.02 1*34 .71


Pd 2.37** .41 1.21 .66 2.12** .75
Pa 2.72* .32 1.95 .4-1 2.24 .41

Pt 3.73* M 1.01 .75 4.19* .62

Sc .71 .82 ,6k 1.196 1.84 1.11

Ma 1 .6 7 .39 1,10 .62 2.77* .57

df = 99 df = 49 df = 49
* p < .01 * p < .01 * p < .01
**p<.05 **p<.05 **p<.05

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE 5

Comparison of the MMPI and the Mini-Mult with


Respect to the Ordinal Position of the
Three Clinical Scales Highest in Hank

Rank on Bank on Mini-Mult


MMPI
j . or
1 2 3 4 more

1 i*6 15 lb 25

2 12 20 23 ^5

3 8 2** 22 ±6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18

TABLE 6

Comparison of the MMPI and the Mini-Mult


with Rfespect to the Ordinal Position
of the Three Clinical Scales
Highest in Hank for Males

Hank on Hank on Mini-Mult


MMPI

1 2 3 4 So:

1 26 8 5 11

2 5 13 10 22

3 k 12 10 2k

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19

TABLE ?

Comparison of the MMPI and the Mini-Mult


with Hespect to the Ordinal Position
of the Three Clinical Scales
Highest in Hank for Females

Hank on Hank on Mini-Mult


MMPI

1 2 3 4 85re

1 20 7 9 14

2 7 7 13 23

3 4 12 12 22

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20

90

80
T Scores

70

60
Mean

50 MMPI
MM

Hs

Mean T Scores of Males


for the Standard MMPI
and the Mini-Mult

80

70
T Scores

6o

50
Mean

^0 MMFI

MM

Sc Ma

Mean T Scores of Females


for the Standard MMPI
and the Mini-Mult

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DISCUSSION

Considering the popular concept that error should

be kept to a minimum, the use of the Mini-Mult will not

be justified until the accuracy of its prediction of the

MMFI is established with a great degree of confidence*

Let us first consider the range of product-moment

correlations. In Klncannon*s comparison of the standard

MMPL and the Mini-Mult (S1M 1 ), the correlations ranged

from ,30 to ,.96 with a median of ,37, The adult out­

patient psychiatric clinic from which the files for this

study were obtained, was the population from which a

sample was selected by Gerald Hartman for a study com­

paring the MMPI and the Mini-Mult, In his study, cor­

relations for the standard MMPI and the Mini-Mult ex­

tracted from the MMFI answer sheets (MM2 ) ranged from

,.66 ta .8 9 with a median of .8 3 , Correlations for this

study ranged from ,26 to .8*4- with a median of .75* Al­

though this range is considerably broader than those

obtained in the above two studies, the results are simi­

lar in all cases. However, the median obtained in this

study is considerably below the median obtained in the

other studies, see Appendix A* When the male and female

groups are considered separately, the results are slightly

different. For the male group, the correlations ranged

21

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22

from .13 to .85 with a median of .77* The Hs scale for

this group was not significant and the Ey scale was sig­

nificant only at the .10 level. The correlations ob­

tained for the female group ranged from .33 to .8 5 with

a median of .66. All scales for this group were sig­

nificant (p<.01) except for scale Hs which is signifi­

cant at the .02 level. However, the median for this

group is slightly lower than the medians for either the

total group or the male group.

Hext consider the mean and standard deviations

obtained in this study and those obtained by Kincannon

in his comparison and Hartman in his MMPI and MMg

comparison. For the total group in the present study,

the standard deviations for comparable scales were lower

on the Mini-Mult than on the MMPI. The standard devia­

tion was greater on the Mini-Mult than on the MMPI in

only two cases, on scale L for the male group and scale

Pd for the female group. In Kincannon*s comparison, the

standard deviation was lower for the Mini-Mult than for

the MMPI for all scales. Hartman obtained only one case,

scale Pt, for his comparison, where the standard devia­

tion was higher for the Mini-Mult than for the MMPI.

The next point to consider is the mean differences

for comparable scales obtained in these three studies.

Kincannon*s study revealed significant differences between

the means for scales F, Hs, and Ma in his SjMj compari­

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
son. He attributed the differences for scales P and

Ma to the tendency of the Mini-Mult to underestimate

extreme scores for these scales. The difference for

scale Hs, he indicated, was probably reflected in sampl­

ing error. Hartman obtained significant differences be­

tween the means for scales F, K, Pd, and Pa (p<.01) for

his MMPI and MM2 comparison. The differences for scales

P and Pa were attributed to the tendency of the Mini-

Mult to underestimate the scale scores for these scales*

He attributed the difference for scales K and Pd to the

tendency of the Mini-Mult to overestimate scale scores

for these scales. In the present study, significant

differences were found for scales P, K, HS, Pa, and Pt

for the total sample (p<^01). At the .05 level of sig­

nificance, the mean difference for scale Pd was also sig­

nificant. Considering the scales for which mean differ­

ences were found, the results of this study are quite

similar to the results of Hartman's study. However,

when the reason for these mean differences are considered,

the only similarity found was for the P scale. For this

scale, the mean difference was attributed to the tendency

of the Mini-Mult to underestimate extreme scores for this

scale, in both studies* In the present study, differ­

ences for scales K, Pd, and Pt were also attributed to

a tendency for the Mini-Mult to underestimate extreme

scores for these scales. The difference for scales Hs

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and Pa reflect a tendency for the Mini-Mult to overesti­

mate extreme scores for these scales. Similar results

were found for the male group which showed significant

mean differences for scales F and Hs (p<*01) and D (p^05)*

As in the total group, the difference for scale F was

attributed to a tendency for the Mini-Mult to underesti­

mate the extreme scores for that scale; and the differ­

ence for scale He was attributed to a tendency for ex­

treme scores to be overestimated. The difference for

scale D, which was not found for the total group, was

attributed to a tendency for the MiM-Mult to overesti­

mate extremes for these scales. Besults for the female

group were somewhat similar to the male group and quite

similar to the total group. With p-^01, significant mean

differences were found for scales K, Hs, D, Pt, and Ma

for the female group. Further, differences for scales

P and Pd were significant with p<«05 for this group.

The differences for scales F,. K, Pd, and Pt were at­

tributed to the tendency for the Hini-Mult to underesti­

mate extreme scores, for this group as well as for the

total group. In addition to the difference for scale Hs,

which was attributed to the tendency for the Mini-Mult

to overestimate the extreme scores for the total and male

groups, the mean difference for scale Ma was found to be

significant for the female group and this difference was

also attributed to a tendency for overestimation of ex­

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
treme scores. A significant mean difference was found

for scale D in both male and female groups but not in the

total group* However, unlike the reason for the differ­

ence reflected by the male group, the difference for

scale D for the female group was attributed to the tend­

ency for the Mini-Mult to overestimate the extreme scores.

In view of these results, it is apparent that the origi­

nal hypothesis, that there are no statistically signi­

ficant differences between the corresponding scale scores

of the MMPI and the Mini-Mult extracted from the MMPI

answer sheets, must be rejected.

One method of studying the degree of correspondence

between validity scales of the MMPI and the Mini-Mult

was presented by Armentrout and Houzer (1970). The re­

sults they obtained by evaluating profiles as invalid

when either scale L, F, or K was equal to or above a T

score of 70, were slightly higher than the results ob­

tained in this study. Using this method of classifica­

tion, their study found that, for the total group, the

Mini-Mult misclassified 26% of the profiles as to validity.

Using the same method of classification for this study,

the Mini-Mult misclassified 21% of the profiles for the

total group as to validity. Further analysis indicated

that the Mini-Mult misclassified 2h% as to validity for

the male group and 18% for the female group. Hevaluat-

ing profiles as invalid when either scale L was equal to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
or above a T score of 60, or scale P was equal to or

greater than a T score of 80, and with K unrestricted,

Armentrout and Houzer reported that the Mini-Mult mis­

classified a similarly high proportion of profiles as

to validity. Applying this second method of classifi­

cation to the present study, the Mini-Mult again mis­

classified 21# of the profiles for the total group, 24#

of the profiles for the male group and 18# of the pro­

files for the female group as to validity.

When the three point code is considered, there is

evidence of appreciable instability. The evidence ob­

tained in this study is similar to the evidence obtained

by Lichtenstein and Bryan in their study of test-retest

stability of the MMPI. Xn their study, the probability

that the top ranked score would remain among the top

three on retesting was 87#. This probability was slightly

lower for the present study, being 75# for the total

group, 78 # for the male group, and 72 # for the female

group. However, the probability that second and third

ranked scores would fall outside the three point code

was 37# and 39# respectively, as evidenced by their

study (1 9 6 6 ). The slightly higher results of this study

showed a probability that the second and third ranked

score would fall outside the three point code 45# and

46# respectively, for the total group. Equivalent pro­

babilities occurred in both male and female groups.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27

This evidence indicates that clinical Judgment based on

information obtained from the Mini-Mult is likely to

result in a high rate of classification errors.

Another method of evaluating profile correspondence

was presented by Armentrout and Houzer in 1970 •• They

studied similarity in elevation by comparing the occur­

rence of clinical scales above or below a T score of

70 for profile pairs. The results they obtained were

consistently lower than the results obtained in this

study. Their results indicated that the Mini-Mult mis­

classified 8 % of the cases for the total group, lk% for

the male group, and k% for the female group. Further,

only 10% of the profile pairs for the total group, 8%

for the males, and 20% for the females, had the same

scales above a T score of ?0 for both tests. The results

of this study indicated that the Mini-Mult misclassified

15% of the cases for the total group, lk% for the male
group, and 16^ for the female group. Also, only 12%

of the profile pairs for the total group, 18% for the

males, and 8% for the females, had the same clinical

scales above 70 on both tests.

Although the amount of error acceptable for prac­

tical purposes is a matter of Judgment, errors should be

kept to a minimum. The results of this study showed

statistical significance for correlations between com­

parable scales on the MMPI and the Mini-Mult. Signi-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
flcant differences between the means of comparable scales

were found in $ of the 11 scales. Also, a variable

and generally low degree of profile correspondence was

obtained by three different measures in this study.

Therefore, the correlational results of this study,

which support Kincannon's results, suggest that, when­

ever use of the MMPI is not possible, the Mini-Mult

may be used. However, other results of this study sug­

gest that, whenever the Mini-Mult is used, considerable

caution should be exercised in extrapolating the results.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A

Product-moment correlations between K Corrected


Saw Scores from the MMPT and the extracted
Hini-Mult: A Comparison of Three Studies

29

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A

Product-moment Correlations between K Corrected


Haw Scores from the MMPI and the Extracted
Mini-Mult: A Comparison of Three Studies

Scale Present Kincannon Hartman

L .76 .82 .74

P .80 .87 .83

K .84 .93 .88

Hs .26 .93 .87


D .78 .91 .89

Hy .48 .82 .82

Pd .74 .90 .82

Pa .75 .84 .85


Pt .82 .96 .87
Sc .67 .90 .79

Ma .55 .80 .66

Mdn. = .75 Mdn. = . 8 7 Mdn. = .83

30

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES

Armentrout, James A. and Rouzer, David L.. Utility of


the Mini-Mult with Delinquents, Journal of Consulting
and Clinical Psychology- 1970, 34, 450,.

Borgatta, E. F. A short test of personality: The


Behavioral Self-Bating form. Psychological
Record*. 1964, l4, 275-284..

Foulds, G r * A,, Caine, T. K,-, and Creasy, M. A. Aspects


of extra and intro-punitive expression in mental
illness. Journal of Mental Science, I960, 106,
599-610.

Hiartman, Gerald B.,- M.A,. "A Comnarison of the Standard


MMPI with a Short Form:- The Mini-Mult." Unpublished
Mastery's thesis, Western Michigan University, Kalamazoo
Michigan, December 1971* Pp»- H i + 37

Jorgenson, C. A. Short form of the MMPI, Australian


Journal of Psychology. 1959, 10, 341-350.

Kincannon, J. C. Prediction of the standard MMPI scale


scores from 71 items: The Mini-Mult. Journal of
Consulting and Clinical Psychology,. 1968 , 32, 319- 325.

Lichtenstein, Edward and Bryan, James H. Short-term


stability of MMPI profiles. Journal of Consulting
Psychology, 1968, 30, 172-1757

Lacks, Fatrlcia B. Further investigation of the Mini-


Mult,. Journal of Consulting and Clinical Psychology.
1970, 35T12T,- 157.-------- ---------------- *-----

Lacks, Patricia B. and Powell, Barbara J. The Mini-


Mult as a personnel screening technique: a prelimi­
nary report. Psychological Reports. 1970, 27, 909-910

31

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

You might also like