Nat Tesr Analogies Pdf Free Download

A factorial design was used to relate MAT scores to the age, sex, and academic field of 192 applicants to graduate school in psychology and education. The results showed significant differences ( P < .01) in test performance as a function of age and field, but not of sex. The education students scored significantly lower than the psychology students (means of 48.2 and 58.4, respectively), and the youngest and oldest groups (median ages of 23 and 41) performed equally well, whereas the middle group (median age of 30) scored significantly lower than the other two. It was suggested that the surprisingly good performance of the oldest group may be due to selectivity in terms of intelligence and general competence, i.e., that only especially effective and self-confident individuals are willing to embark on a graduate career near or after the age of 40.

Figures - uploaded by Fred J Thumin

Author content

All figure content in this area was uploaded by Fred J Thumin

Content may be subject to copyright.

Discover the world's research

20+ million members
135+ million publications
700k+ research projects

Join for free

Psychological

Reports,

1966, 19,15I-714. @ Southern

Universities

Press 1966

MILLER ANALOGIES

TEST PERFORMANCE

AS RELATED TO AGE,

SEX,

AND ACADEMIC FIELD

FRED T. THUMIN AND CAROL BOERNKE

lY/ as hi ngt

n U n i uers

Summary.-A factorial design was used to relate MAT scores to the age,

sex, and academic

field of l)2 applicants

ro graduare

school in psychology

and

education. The results

showed

significant differences

(P <.01) in tesr

per-

formance as a function of age and field, but not of sex. The education students

scored significantly lower than the psychology srudents (means of 48.2 and

)8.4, respectively),

and the youngest

and oldest groups (median ages

of 23

and 41) performed equally well, whereas

the middle group (median age of 30)

scored significantly lower than the other two. It was suggesred that the sur-

prisingly good performance of the oldest group may be due to selectivity in

terms of iotelligence and general comperence,

i.e., that only especially

effective

and self-confident

individuals are

willing ro embark on a qraduate

career near

after the age of 40.

The Miller Analogies

Test (MAT) continues to be used widely in select-

ing students

for graduate training, and irs populariry has sparked a considerable

amotlnt of research

which has added appreciably to the over-all understanding of

the instrument-its assets as well as its limitations. Most of the validity studies

relating MAT scores to the grade point average of graduate students in educa-

tion and psychology have yielded correlarion coefficients r^nging between .20

and .50 (e.g.,

McNemar, 1948; Plarz,

McClintock, & Katz, 1959; Stafford,1952;

Strowig, I9t2) . On the other hand, a few investigators have obtained values in

excess

of .65 (Cureton, Cureton,

& Bishop, t949; Robinson, 19t8).

In some studies, rather marked differences have occrlrred

berween the MAT

performance of graduate students in psychology and educarion, the latter scoring

relatively

low (Miller,1960, p. 6; Robinson, 1958, Travers, 1948). With the

exception of Cook (1942), who found that str-rdents over 30 yr. of age scored

relatively iow on the MAT, and of Abeles ( 19tB ) , who apparenriy found no sig-

nificant difference between males and females, it appears that virtually nothing

has been done to investigate MAT performance as a funcrion of the age and sex

of .ls. The major objective of the present study was to examine the relationship

of these

two relatively neglected

variables to MAT score, whiie a secondary pllr-

pose was to obtain additional information as to the alleged discrepancy

between

the scores

of education and psychology students.

Mrrsoo

Subje cts a:ed. Materi.als

Between 1958 and 1965 the MAT was administered to 173 gradr,rate

school

applicants at \Tashington Universiry's Adult Counseling Service. Most of these

individuals (79%) were applying to graduate school in education, with the re-

-7

<'' F. J. THUMIN & C. BOERNKE

maining 2l% applying in psychology. Four forms of the test were used (H,

J, K, and L) all of which are equivalent for practical

purposes,

according

the test author and

publisher

(Miller, 1960,

p. 5; Miller Analogies Test, 7)64).

Experimental Desi.gn

A factorial design

was employed

to study the relationship of ,Ss' test scores

to their age, sex, and field of academic endeavor

(uiz.,edtrcation

or psychology).

Due to the relatively

small number

of older

psychology applicants in the sample,

it was necessary

to use rhree rather broad age categories-undet 26 yt.,26

through 35, and 36 and above. Thus, the final design consisted of 12 cells-

three

age

grorlps

by two fields

of study

by sex. As the complrter

program

to be

used in the analysis

required that each

cell contain

an equal number of Ss,

the

cell with the smallest

number

of Ss

determined

the size

of all cells. The limit-

ing cells (male and female psychology

applicants,

36 years

and over) each con-

tained 16 ,Ss. A random

sampling

procedure

was used to select 16 Ss

for each

of the remaining 10 cells. Thus, the final sample consisted of 192 Ss. The

median

ages for the three

age

groups

were 23, 30,

and 47 yt.

The various treatment groups

were almost perfectly matched in terms

of in-

stitutional choice,

i.e., in all groups, approximately

35% of Ss applied only to

Vashington University,

63%0 to rVashington

University

plus one or more other

schools, and 27o to institutions

other than \Tashington Universiry.

Rnsurts

The results

of the analysis of variance are

presented in Table 11 and indicate

that significant differences in MAT score occurred among

the three chronologi-

cal age

groups

and between the psychology

and education

groups. Scores,

how-

TABLE 1

Rrsurrs or 3 X 2 x 2 ANervsrs oF VARTANCE

Source MSdf

Academic fieid (F)

Sex

(S)

Age

(A)

FXS

FXA

SXA

FXSXA

\(ithin replicates

Total

L 497 L.5

I 178.3

2 1809.3

r 584.t

2 L44.r

2 727.6

2 100.1

180 289.6

19r

r7.t7

0.62

6.21

2.02

0.10

t <1

o a5

<.01

1A computer printout has been filed with the American Documentation Service, Auxiliary

Publications Project, Photoduplication Service, Library of Congress,

\(ashington, D. C.

20540. Remit $1.25 for 35-mm. microfilm or $1.25 for photocopies of Document No.

9060.

MILLER ANALOGIES TEST SCORES. AGE. SEX AND FIELD 753

ever,

did not vary significantly

as a function of sex, nor were any

of the interac-

tion effects significant.

In order to determine

the directions of the significant effects,

the means

the various treatment groups were calculated and summarized

in TabIe 2. As

there

were no significant

differences

between

males

and

females, the data for the

sexes

were combined

in making these

calculations.

TABLE 2

MnaN SconEs

ron TnratMENT GRoups

Academic

field Under

26 Over-all

26-3t 36, up

Education

Psychology

Over-all M

5).4

60.8

57 I

40.t

53.9

l-j',

t0.7

60.t

) ).b

48.2

58.4

53.3

Note.-N = 32 |or each of six treatment

groups.

Thus, the psychology applicants scored significantly higher than did the

education

applicanrs

(means

of 58.4

and 4B.2,respecrively),

and the direction of

the difference was consistenr for each age group independently. Regarding the

age

variable, Table 2 shows that the significant effect is attributable to a marked

decrement in MAT score among the 26-35-yr. group. Although the decrement

occurred both among psychology and education applicants, it was especially

pro-

nounced among the latter. Additional analysis

by I test indicated that the mean

score

for the 26-35-yr.

group was significanrly lower (P < .01) than those

the youngest

and oldest

groups,

whereas no significant difference existed

between

the means of the latter two groups.

DrscussroN

In certain respects the results

of the present study were as expected,

whereas

in other respects

they were rather surprising. For example, ir was anticipated

that male and female Ss would perform equally well on the tesr, as in fact they

did. It was also anticipated correctly that the present data would substantiate

earlier reports on the relative superiority of psychology to education srudents

the MAT (Miller, 1960, p. 6; Robinson, 1958; Travers, 1948) .

It is perhaps noteworthy that the median score for psychology students in

the present study was 58, some nine points lower than the corresponding value

appearing in the normative data of the MAT manual. This discrepancy is un-

doubtedly explained by the fact that all Ss in the normative study had completed

their undergraduate training ar the rime of testing and most of them were al-

ready enrolled in their first year of graduate study. Ss in the present study had

not yet been accepted to graduate

school ( many were, in fact, turned down) , and

at the time of testing about one-third of them were srill in the senior year.

The mosr unexpected

finding was the marked decrement in MAT score be-

754 F. J. THUMIN & C. BOERNKE

tween 23 and 30 yr., followed by an improvement between the ages

of 30 and

41. Although the writers would not wish at this point to preclude the possi-

bility of an adventitious sampling peculiarity, it is interesting to speculate on

at least one alternative. The decrement berween 23 and 30 yr. offers no par-

ticular problem, for one might expect that the latter group, comprised largely of

employed individuals who had been absent from the academic

scene for a num-

ber of years,

wouid score

relatively low.

The question as to why the oldest group (4I yr.) perforrned as well as the

youngest group ( 23 yr.) is more difficult to reconcile, but the writers would

like to suggest

that considerably more colrrage and self-confidence are required

to embark on a graduate

career

at the age

of 4l than at the age of 30, and that, in

the main, only the brighter, more competent and effective individuals are will-

ing to accept

the challenge. This hypothesis

receives additional support from two

recent

investigations

(Thumin & .Wittenbery,1961; Thr.rmin

& Boernke, 1966)

which indicated that femaie secretarial applicants in their 40s and 50s were either

eqttal or sr.rperior

to their yollnger counterparts in mental ability, verbal compre-

hension,

typing speed

and typing accuracy.

REFERENCES

ABELES, N. Profile patterns

of counselor

trainees. Amer. Psycholoeist,

l9)8, 13, 35I-

312. (Abstract)

Coor, \7. $/. Predicting

success of graduate

students in a college

of cducation. Sch.

€t

oc., 1942,

6, 192-r9t.

CunrroN, E. E.,

CUREToN,

L. \/., & BrsHop, R. Prediction

of success

in graduate

study

at the University

of Tennessee.

Ant.er. Psychologi:t,

1949, 4, 361-362.

McNrrtan, Q. Unpublished

manuscript,

1!48. Cited

by W. S. Miller, Miller Analogies

Tett manaal. New York: The PsychologicalCorp.,

lg6o. P. 11.

fuIiller

Analogies

,Test

sapplementary

manaal: Form L. New York: The Psychological

Corp.,1964.

Mtrrrn, (. S. Miller Analosies Test manual. New York: The Psychological

Corp.,

1960.

PLATZ,

A., McCrrNrocr, C., & KA^rz,

gies

Test

on prediction

of graduate

289.

Undergraduate

grades

and the Miller Analo-

success.

Amer, Psycbologi:t,

1919, 14, 285-

RontNsoN, D. \tr. A .comparison of two batteries of tests as predictors of first year

achievement

in the graduate school of Bradley University.- In Fit'teentb Yearbk,

Nat. Coancil Illeasmt

Used. in Edac., 1958. Pp. 118-127.

STAFFoRD,

J. rW. Unpublished manuscript, 19)2. Cited by $f. S. Miller, Miller Analo-

gies Test manual, New York: The Psychological

Corp., 1960. p. 12.

SrnowIc, R. \W. Predictive values of the Miller Analogies Test for master's

degree stu-

dents in education. Unpublished doctoral dissertation,

Stanford lJniver., {952.

THUMIN, F. J., & BoERNKE, c._ Ability scores as related to age among female job ap-

plicants. J

. Geront.,

1966, 21, 36947 | .

THUMIN, p. J,, .& \wrttnNnnnc, A. Perso.nality

as related to age and mental abfiity in

female

job applicants. J. Geront.,796t,20, l}t-l}j.

Tnavrns, R. M. unpublished_manuscript,

1!48. cited by \7. s. Millcr, Miller Analogies

Tett ntanual. New York: Psychological

Corp., 1960. p.14.

Accepted

Ausust 19, 1966.

Ruth E. Murray

ment of the test's predictive function. The problem may be compounded by the considerable variability in candidates' qualifications across the institution. This lamer proposition questions such studies as those of Jansen and Johnson (1967) and Thumin and Boernke (1966); the former study involved three broad groups of applicants to graduate education programs in administration, school services, and teacher improvement while the latter studied applicants for educational psychology and education in general. While differences were found, neither study adequately explored the possibility of divergency of qualifications. The present study addressed directly this question by comparing Miller Analogies Test scores for applicants to six areas in education. Subjects were prospective graduate students in six departments within the education division of a midwestern university, all had taken the Miller Analogies Test within the last three years. A sample of 100 examinees was drawn for each area: special education, educational psychology, elementary and secondary education, educational administration, and instructional technology. A one-way analysis of variance was computed across these areas.

Edward E. Cureton
Louise W. Cureton
Ruth Bishop

Compared the Miller Analogies Test, Form G, and the I. E. R. Intelligence Scale CAVD, Levels M, N, O, P, and Q, Form 3 as predictors of success in psychology graduate study. Ss were graduate students of psychology enrolled at the University of Tennessee. Two criterion measures were used to check the validities of these tests. The first was the grade-point average in all courses taken while enrolled as a graduate student at the University of Tennessee. The second criterion measure was an overall rating completed by each member of the department who was teaching graduate courses in psychology. Findings suggest that for the population of which this group may be considered a sample, the 50-minute Miller Analogies Test is at least as valid as the CAVD, which requires about four hours to complete. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Arthur Platz
Charles McClintock
Daniel Katz

Ss were "124 students who entered the graduate program at the University of Michigan during the years 1951 through 1955… . Information available on each applicant at the time of entrance included the following measures": total undergraduate grade point average (GPA), undergraduate GPA in science and mathematics, undergraduate GPA in psychology courses, MAT score, objective comprehensive examination in psychology. Measures of "success" included: GPA in graduate courses, marks on doctoral preliminary examinations, faculty ratings of potential professional contribution and potential scientific contribution. 6 tables of results are presented. The best predictor of marks in graduate courses was undergraduate GPA in science courses; it also best predicted preliminary examination grades. "The MAT significantly predicted graduate course grades and was the best predictor of potential scientific contribution of the student." (PsycINFO Database Record (c) 2012 APA, all rights reserved)

unpublished_manuscript, 1!48. cited by \7. s. Millcr, Miller Analogies Tett ntanual

R M Tnavrns

Tnavrns, R. M. unpublished_manuscript, 1!48. cited by \7. s. Millcr, Miller Analogies Tett ntanual. New York: Psychological Corp., 1960. p.14.

Unpublished manuscript, 1!48

Q Mcnrrtan

McNrrtan, Q. Unpublished manuscript, 1!48. Cited by W. S. Miller, Miller Analogies Tett manaal. New York: The PsychologicalCorp., lg6o. P. 11. fuIiller Analogies,Test sapplementary manaal: Form L. New York: The Psychological Corp.,1964.

Miller Analosies Test manual The Psychological Corp gies Test on prediction of graduate 289. Undergraduate grades and the Miller Analosuccess

Mtrrrn

Mtrrrn, (. S. Miller Analosies Test manual. New York: The Psychological Corp., 1960. PLATZ, A., McCrrNrocr, C., & KA^rz, D. gies Test on prediction of graduate 289. Undergraduate grades and the Miller Analosuccess. Amer, Psycbologi:t, 1919, 14,

A .comparison of two batteries of tests as predictors of first year achievement in the graduate school of Bradley University.-In Fit'teentb Yearbk

D Rontnson

RontNsoN, D. \tr. A.comparison of two batteries of tests as predictors of first year achievement in the graduate school of Bradley University.-In Fit'teentb Yearbk, Nat. Coancil Illeasmt Used. in Edac., 1958. Pp. 118-127.

Profile patterns of counselor trainees