naval submarine medical research laboratory · rorschach scores and, to a lesser extent, certain...
TRANSCRIPT
NAVAL SUBMARINE MEDICAL RESEARCH LABORATORY
SUBMARINE BASE, GROTON, CONN.
REPORT NUMBER 1080
PREDICTION OF NUCLEAR SUBMARINER ADAPTABILITY FROM AUTONOMtC INDICES AND RORSCHACH INKBLOT RESPONSES
by
Benjamin B. Weybrew and
H. Barxy Molish
Released by:
C. A. Harvey, CAPT, MC, USN Commanding Officer Naval Submarine Medical Research Laboratory
9 September 1986
A-w-m/w: <Vv- "M-Vii-. •!•«"'<=;»»ft distribution unlimited.
REPORT NUMBER 1080
PREDICTION OF NUCLEAR SUBMARINER ADAPTABILITY FROM AUTONOMIC INDICES AND RORSCHACH INKBLOT RESPONSES
by
Benjamin B. Weybrew and
H. Barry Molish
Approved and Released by:
C. A. Harvey, CAPT, MC, USN Commanding Officer
Naval Submarine Medical Research Laboratory
9 September 1986
SUMMARY PAGE
THE PROBLEM
FINDINGS
Delineated by factor analysis, two Rorschach dimensions were found, one of which correlated significantly with three adaptability criterion axes. Two unique factors were also discovered, one a diffusely-structured ANS factor, and the other a complex, generalized criterion dimension. Taken together, these findings suggest that selected Rorschach indicants, and, to a lesser extent, certain ANS indices may have useful validity as predictors of submariner adaptability.
APPLICATIONS
Most nuclear submariners adapt well to the conditions existing during missions of 60 or more days duration. Others do not adapt adequately. The results of this factor analytical study suggest that a submariner's responses to the Rorschach Inkblot Test as well as his ANS patterns of reactivity to experimentally-induced stress may be significantly predictive of his capacity to adapt to submerged conditions.
ADMINISTRATIVE INFORMATION
This manuscript was submitted for review in April 1986. It was approved for publication on 9 September 1986, and has been designated as NSMRL Report No. 1080.
ii
ABSTRACT
To identify the most valid predictors of submariner adaptability, the authors derived 23 indices from the responses of 170 nuclear submariners to the Rorschach Inkblot Test, 11 measures of Autonomie Nervous System (ANS) reactivity to contrived stress, and five adjustment criteria. Factor analysis of this 39x39 correlation matrix yielded two Rorschach Factors, one of which correlated with three criterion dimensions. Two unique factors were also discovered, one, a structured ANS factor, and the other, a complex criterion scale. Selected Rorschach scores and, to a lesser extent, certain ANS indices emanating from this study, may be usefully-valid predictors of the adaptability of nuclear submariners during long patrols.
in
PREDICTION OF NUCLEAR SUBMARINER ADAPTABILITY FROM AUTONOMIC INDICES AND RORSCHACH INKBLOT RESPONSES
Benjamin B. Ueybrew Ph.D.* and
H. Barry Molish Ph.D.*
Since the Nautilus, the first automic-powered submarine, was
launched in 1954, there has been a focussed effort to identify
valid predictors of the quality of adaptation of nuclear
submariners to long submerged missions often exceeding 80 days.
A review of the literature of submarine psychology in the sixties
indicated that selected biographical, aptitude and objective
personality test scores yielded low but non-chance correlations
with officer's adjustment ratings of enlisted men . However, the
same source indicated that there was a hiatus in the submarine
literature in that the use of projective techniques such as the
Rorschach Inkblot procedure for psychological prediction in this
context has not been thoroughly explored. Moreover, data were
presented to suggest that peripheral indices of autonomic nervous
system (ANS) reactivity and resiliency may be valid predictors of
maladaptive emotional reactions of submariners at sea (op.cit.).
Finally it was argued in the same review that the methodological
possibilities of multivariate analytical techniques such as
factor analysis applied to the submariner assessment problem had
not been given sufficient field testing.
* This study was conducted while both authors were members of the staff of the U.S.N. Submarine Medical Research Laboratory (NSMRL). At present, Dr. Weybrew is Associate Professor of the University of New Haven, Connecticut and Dr. Molish is staff psychologist, Texas Children's Hospital, Houston, Texas. Appreciation is extended to Mr.' James W. Parker, NSMRL staff, who did the computer programming and data processing.
In a sense then the objectives oE this paper were to follow
through on the three recommendations presented in the 196 3 review
mentioned above which, to reiterate, were: to evaluate the
Rorschach technique as a means of identifying predictors of
nuclear submariner adaptability, to determine the validity of
measures of ANS reactivity for the same purpose, and, finally, to
ascertain the methodological possibilities of factor analysis in
this context.
But there is not a complete absence of studies of this kind
in the literature of submarine psychology. For example, one
somewhat antiquated study provided the background for the present
study2. Completed in the early fifties but not reported until
the early sixties, this voluminous factor analytical study
involved 180 variables including 18 Rorschach measures and 16
endocrinological and biochemical indices, some of which were
known correlates of ANS reactivity. The subject sample consisted
of 88 enlisted candidates for the diesel submarine service.
While some additional information regarding the factor structures
identified in this 1950 study will be discussed later, suffice to
say that a Rorschach "Productivity" factor, marked mainly by high
loadings by "R" (the total number of percepts reported on the 10
inkblot cards), and an endocrine function factor, loaded mainly
by measures of urinary 17-ketosteroids, were two of the major
factors isolated in this study.
In brief, the present study parallels certain aspects of the
above submariner investigation. However, instead of a subject
sample drawn from enlisted recruits for the diesel submarine
service, the data for the present study were collected entirely
from nuclear submariners. All ot the data for this study was
collected in the sixties from both enlisted crews of one nuclear-
powered, Fleet Ballistic Missile (FBM) submarine. The tests were
administered during the three months retraining period to the
gold crew while the blue crew was on station and vice versa.
METHODS
Subjects
The subject sample consisted of 85 male enlisted nuclear
submariners from each of the Blue and Gold crews of one FBM
submarine. Their age ranged between 18 and 45 years with a mean
of 27 years. The mean General Classification Test (GCT) score
for the distributions of both crews was approximately 59
{S.D.=8). This mean was approximately equivalent to the 81st
percentile for the U.S. Navy enlisted population as a whole.
Since both the distributions of test scores and biographical data
for the two crews were not significantly different, they were
pooled, the end result being a composite sample of 170 for
inclusion in the factor analysis to follow.
Procedure
The standardized, 10-card form of the Rorschach Inkblot Test
was individually administered and hand scored by the second
3 author, using a modified version of the Beck Scoring system .
Since most of the 23 Rorschach measures involved simple response
frequency counts which characteristically yield Poisson*
'Poisson distribution—a skewed probability distribution.
'distributions, it became necessary to reduce and stabilize the
correlation between the means and variances typically found in
distributions of this kind. Recommended for this purpose was the
Freeman-Tukey square root tranformation which was applied to all
of the Rorschach scores derived from response frequency counts .
Table I contains the means and SDs as well as a brief
statement describing the scoring procedure for each of the
Rorschach measures. These include eight measures of Location
(scores numbered 1 through 6 and 14 and 21 in Table I), five
Determinant indices (Z, C, F+%, Zf, and M—Nos. 8, 10, 13, 22,
and 23, respectively), three measures of Response Style, E.B.,
Afr, and L (Nos. 9, 15, and 16), and two measures of Content, A%
and P (Nos. 11 and 12). Productivity was measured by total
number of percepts (R, No. 7), while the remaining four scores
(Nos. 17, 18, 19, 20) were measures of response rate and latency.
Brief statements describing the 16 ANS and criterion measures
included in the 39-variable factor analysis are contained in
Table II.
The five adjustment criteria consisted of factor scores
obtained from a separate factor analysis of rating data collected
from the same two submarine crews during two successive submerged
missions of 60 days duration. These factor scores for each
subject on each criterion dimension (Nos. 24-28 in Table II) were
derived by summing the Z-scores* weighted by the factor loadings
*Not to be confused with the Rorschach Z-score (No. 8 in Table I), this Z-score is a linearly-transformed, standard score, the distribution for which has a mean of 50 and a SD of 10.
Table I
Scoring Procedures and Means and SDs for the 23
Rorschach Variables {N=170 Nuclear Submariners)
Sy-bol Score Content Computational Procedures Mean S.D.
1 W
2 D
3 Dd
A Ws
Us
Dds
Whole Card Response
Major Detail
Rare Detail
Whole Response with S Component
White Space, Major Detail
White Space, Rare Detail
7 R Total Responses
8 Z Synthesized Percepts
9 EB Experience Balance
10 C sum
Undiluted Color Determinant
il A% Animal Content
12 P Popular Response
13 F+% Superior Form Ratio
14 S White Space
15 Afr Affective Ratio
16 L Lambda Index
17 T/R Response Time
18 T/lst R First Response Lat ency
1? Fin R Fluctuation of R's Between Cards
20 Fin T/ 1st R
Fluctuation of T/lst R
21 SI Relative Number White Space R
22 Zf Form-Determined . Organizational Activity
23 M Movement Responses
Number of W Responses(R)
Number of Responses
Number of Responses
Number of Responses
Number of Responses
Number of Responses
2 of Responses to 10 Cards
Z of Weighted 2 Scores on 10 Cards
M/C sum
Sol all C Values
Percentage of Total R
Total Number Responses
F+/ZF+ and F- (percent)
Number S-Determinant R's
Ratio of R's Cards 8-10/ R's Cards 1-7
Ratio of F Responses to non-F Responses
Average Time (T) per R (seconds)
Average T/lst R over 10 Cards (seconds)
Average Differences Between R's/ Card
Average Difference Between T/lst R per Card
Percentage White Space Responses
2 unweighted Z responses
Number of M Responses
4.9 4.0
16.7 8.2
1.6 1.3
0.6 0.4
1.6 1.0
0.5 0.2
24.8 11.1
22.4 14.7
1.2 1.0
3.1 2.3
50.5 13.7
6.2 2.2
78.4 12.8
1.9 1.3
0.7 0.6
3.4 3.1
34.2 15.3
22.4 11.6
1.2 0.9
17.3 10.4
6.8 6.2
8.1 4.5
1.8 1.3
Table II
Scoring Procedures and Means and SD's for the 11
Autonomie Indices and Five Submariner Adjustment Criteria (N=170)
No. Symbol Score Content Computational Procedures Mean S.D.
24 F-I Favorable Adjustment I
25 F-II Favorable Adjustment II
26 F-III Poor Military Behavior
27 F-IV Good Military Behavior
28 F-V Leadership Potential
29 ARI Autonomie Rebound Index
30 BL EDC Basal Level
31 BL-AD Fluctuation of Basal EDC
** 32 Con GSR to Maximum Dis-
EDC-Max crimination Conflict
33 Con EDC GSR to Minimum Dis- Min crimination (BH)
34 BHRS Breathholding (BH) Recovery Slope
35 Hyp- EDC Displacement Index BH/DI During Hyperventilation
and BH
36 Hyp- Recovery Index, Hyp/BH BH/DI
37 Hyp- Recovery Quotient, BH/RQ Hyp/BH
38 Con-RI Recovery Index to Conflict
39 Con-DI Displacement Index to Conflict
Factor Scores, 12 Ratings
Factor Scores, 12 Ratings
Factor Scores, 3 Ratings
Factor Scores, 3 Ratings
Factor Scores, 3 Ratings
EDC Recovery Rate
Average Resting EDC Level
Average Deviation of Basal EDC
Mean GSR's to Maximum Conflict
Mean GSR's to Minimum Conflict
Slope of EDC Curve During BH (degrees)
% EDC Change During Hyp/BH
% EDC Recovered After Hyp/ BH
R.Q.=RI/DI (to percent)
% of DI (Conflict) Recovered
% Change in EDC to Conflict
225.1 100.2
446.8 329.8
73.3 50.7
66.3 22.1
69.8 55.2
25.3 24.5
83.9 27.6
85.0 63.3
267.9 193.1
239.8 165.6
54.6 52.5
26.8 5.2
20.2 6.3
2.4 1.9
1.2 0.7
58.2 41.9
*See text and reference for a description of the methodology used in deriving those five criterion scores.
**CSR-Calvanic Skin Response; EDC-eloetroderuial conductance, units are micromhos, after Freeman'.
for the specific rating scales defining each of the 5 factors
used as criteria. The rating scale battery consisted of the
following 22 scales; 15 paired-comparison trait rating scales, 5
standard navy semi-annual perfofinance scales, and 2 attitude-
change measures. The content of the trait ratings included
scales for emotional stability, leadership, interpersonal
attitudes, etc. These are described in some detail elsewhere
Numbered 29 through 39 in Table II, the ANS measures were
derived from data that originated in two laboratory-contrived
stress situations the methodology for which was developed and
8 field-testing in another study . The hyperventilation and
breathholding (Hyp-BH) paradigm, designed to measure ANS
reactivity and resiliency, consisted in asking each subject to
take 4 deep breaths and to hold the fourth breath as long as
possible. Electrodermal conductance (EDC) measures were taken
continuously throughout the procedure. Seven scores, briefly
described and numbered 29, 30, 31, 34, 35, 36, and 37 in Table
II, were derived from this experiment (see reference eight for
procedural details).
The second experimental paradigm, the Conflict (Con)
induction procedure, consisted again of continuous EDC
measurement while each subject was time-paced while making the
decision as to the brighter of two lights which incrementally,
trial-by-trial, were made increasingly more similar in terms of
intensity (op.cit.). The "true" brighter light was randomized
right or left during the 16 trial sessions. The four Conflict
scores, numbered 32, 33, 38, and 39 in Table II, were derived
from this procedure.
It is to be noted that the means and SDs for each of the 16
ANS variables are also included in Table II. Relevant aspects of
these distribution statistics for each measure will be referred
to in the discussion section.
Statistical Methodology
The 39x39 matrix of 741 Pearson Product Moment coefficients
was factor analyzed by computer, using the Principal Components
9 Method . With communalities estimated by a reiterative method,
this mathematically precise technique yields as many factors or
principal components as variables; however, most of the variance
is removed with the first few vectors extracted. The axes were
rotated orthogonally by computer using the Varimax procedure, an
analytical technique that mathematically closely achieves what
factor analysts call the simple structure criterion10
RESULTS
The Correlation Matrix*
Of the 741 coefficients in the 39x39 matrix, 156 or 21
percent were significant at the 5% or less confidence level**.
Both the intercorrelations of the Rorschach scores (253
coefficients, r's) and the ANS scores (55 r's) were remarkably
high inasmuch as 66% of the former and 80% of the latter were
*The bulky correlation matrix is not presented here; however, a copy may be obtained by request from the first author.
**"Significant" hereafter in this context means 5% or less confidence level.
significant. However, many ot these high r's are considered
spurious due to the linear dependencies inherent in the scoring
systems for the two classes ot variables. For example, A% and S%
are percentages of the total number of Rorschach responses (R-
score, No. 7 in Table I). Similarly, examples of ANS score
interdependencies are found in variables numbered 35, 36, and 37
in Table II, wherein each score is directly proportionate to all
others. Finally, very high inter r's are seen among the five
adjustment criterion dimensions, (Nos. 24-28 in Table II), but
this time the spuriously high coefficients are probably due in
part to the halo effect, characteristically found in trait
ratings of this kind.
It is important to note that only 6 (5%) of the 115 r's
relating Rorschach scores to the 5 criteria reached significance.
At the same time 12 (22%) of the 55 r's interrelating the ANS
measures to the criteria were significant. Finally, only 2% (5
r's) between the Rorschach indices and the ANS measures reached
significance.
Reverting back to the distribution statistics for the 39
variables in Tables I and II, the question arises as to the
comparability of the score distributions of the nuclear
submariner sample in this study with identical scores obtained
from other population samples. Unfortunately, comparable
normative data existed only for the Rorschach indices obtained
12 from one civilian population sample, N=157 . The mean scores
for only three of the 23 Rorschach measures in Table I differed
significantly from the means of this comparison sample (5% level,
t-test). Thus the FbM group (N=170) yielded lower total response
scores (R), lower movement scores (M), and lower isolated detail
scores (D), variable Nos. 7, 23, and 2, respectively (Table *).
Possible interpretations of these group differences will be
addressed in the discussion section to follow.
The Factor Matrix
In interpreting factors extracted by the Principal Components
Method it is generally considered that to be meaningful the
factors should have eigenvalues ~> unity*. Accordingly, only 12
of the total 39 principal components are presented in Table III.
Furthermore, because of a lack ot" clarity resulting from the
rotated solution, only four of these 12 factors, accounting for
more than half of the total variance, were considered
interpretable. Finally, in order to meaningfully discuss the
structure of these four factors, it was necessary to identify the
marker variables, i.e., the variables with the highest loadings
on each factor. These marker variables are listed and rank-
ordered for each of the four factors in Table IV**.
Consistent with the earlier factor analysis of Rorschach data
obtained from diesel submariners , we labelled the first factor
"Organized Productivity" since it was weighted heavily by R, Z,
and Zf scores, (nos. 7, 8, and 22 in that order in Tables III and
*Eigenvalues are roots of equations which are solved in sets from which the factors or, in geometric terms, vectors are derived. For a comprehensive coverage of factor analytic methodology, see13 .
**Tables III and IV should be used in conjunction when deducing the structure of each of the four factors.
10
Tafale III
Rotated Factor Loadings for 23 Rorschach Indices, 13 Autonomie Measures, and Five Adjustment Criteria (N=170)
_ . ... .
Variables*
w
Fl F2 F3 F4 F5 F6 F7 F8 F9 F10 Fll F12
1 52** -28 46 -19 15 19 32 -19 12 -13 -15 -04 2 D 60 -02 -11 16 -65 06 -10 08 08 05 10 06 3 Dd 33 06 -22 22 -54 09 -26 08 -21 -13 -19 16 4 Ws 46 -12 25 -30 29 -20 -21 -08 -03 12 09 -10 5 Ds 61 06 -13 09 11 -05 -52 06 14 -04 00 -08 6 DdS 32 -04 -15 -01 -11 -25 -25 18 -41 10 10 02 7 R 69 -07 -47 28 -23 02 -17 -06 06 03 -02 15 8 Z 77 -16 -13 06 45 01 19 -11 15 04 -08 -07 9 EB 27 -04 -39 -05 -05 01 31 55 -03 -04 13 -25
10 C sum
50 -09 64 -14 -18 06 02 03 -01 -09 -08 34
11 A% -27 07 -46 24 -04 06 -02 -24 24 18 05 -23 12 P 39 -03 -18 04 -40 11 30 15 14 09 20 -17 13 F+% -26 -02 -57 13 30 -10 33 -05 03 01 09 17 14 S 71 00 08 -20 17 -22 -53 11 -05 08 12 -14
15 Afr 18 07 64 -23 -40 08 14 03 15 09 00 -19 16 L -27 07 -19 04 -53 02 -19 -29 29 03 -01 -31 17 T/R -67 08 -11 -07 38 00 -05 35 10 -04 -05 01 13 T/lst R -51 09 37 -25 06 16 -13 46 17 10 -02 -05
19 Fin R 48 -02 -27 24 -26 02 -18 17 22 05 -22 08 20 Fin T/lst R -52 05 08 -13 21 09 -23 48 31 07 -03 13 21 S% 50 03 06 -24 35 -27 -58 10 -05 07 14 -18
22 Zf 77 -21 -08 04 40 03 19 -15 23 00 -08 -09
23 M 44 -09 -24 05 -04 05 47 55 -11 00 12 -23 24 F-I 06 -19 29 59 -03 39 -01 06 29 00 -08 06 25 F-II -08 -12 08 40 -01 21 -18 25 31 -32 -43 -06 26 F-I 11 -10 -12 42 -77 04 -32 05 05 -11 06 09 -10
27 F-IV -15 -04 -42 73 06 30 06 02 -16 06 10 -12
28 F-V 06 27 -40 69 19 -07 09 00 -16 07 08 -12
29 ARI 13 64 -13 -11 -08 -40 14 05 04 -02 -10 00
30 BL -05 -32 09 15 -02 32 -02 -02 11 48 21 25
31 BL-AD 18 60 -06 07 09 -01 -03 -12 -24 05 05 13
32 Con EDC-Max 04 45 -04 -26 -11 -70 18 00 09 -09 -02 13
33 Con EDC-Min -11 09 -07 -40 -15 -75 14 -05 06 -05 01 09
34 BHRS -04 09 01 -23 -20 -56 31 -07 01 28 -10 -08
35 Hyp-BHDI 18 92 12 03 04 13 04 02 07 00 02 -02
36 Hyp-BHRI 18 95 04 05 04 11 04 -01 07 05 -01 02
37 Byp-BHRQ 16 93 05 09 07 17 02 02 06 -01 -01 00
38 Con-RI 18 95 04 07 08 15 02 00 05 01 -01 02
39 Con-DI 12 89 08 12 09 19 01 -01 05 03 01 00
% Variance 18 16 9 9 7 7 6 5 3 2 2 1
*The symbols identifying the variables are explained in Tables I and II. **Decimals are omitted.
11
X (U JJ X CJ -H CO K U JJ JJ CO X X w
Ö CO O U -H O M JJ 0)
u JJ CO ft fa M
CJ M O XI
■^i ß CO cO X
y 1H -H a e o o > fa ß
M o <u JJ
o> X 3 M 4J <: x>
CO >W •> H O XI
o CO CO CJ X!
-H CJ JJ cn CO S-l
■H O SH Crf CD JJ O cj cn CO >H X (0 x: oi Ü co
t-t dl to x: u JJ 3 w g O O 3 >H >H 4H
c a) JJ
a o o
u o JJ
CJ
fa M-I O
JJ O M In JJ
CO
<
co <u f-t x> cd •H 1H
CO > In 41 Jii M
s
H CO ?~, JJ cd ■H CO <u
1 10 3 x: B XI 0) >. o> 4-1 u 4-1 I-l
MH x: o £ XI •d MS 0) ß 3 (0 J-> -H iH H O COC E-* c x: 4-1 .-1 ~^ 41 C_> •H ^-^ 4> o 4) C x; a)
-O 41 4J f-H H U -O 4-1 - •H x: 00 ß fa O ß d 01
H c -r4 CJ CO 01 o 0) •H XI • CO c c 41 XI a c rH •> o d >H • CO CO co u« x: i-5 D, -H ON o oo in E 1-1 CJ-! r-t >*orf C -H o tO CO CO i-t "s. •HA» u >
E o X) H CO x: SH JJ •H JJ CO >.r-| CJ 0 CO CO .c C ci-i 4J CO CO C )-i CO JJ 01 00 O •H 44
x: a) o o S ß ^4 o O .Ü x* O.X1 41 *H 0) CO 4J CO CJ O ■H ß 60 c >-i <i> E Ti x: 41 U C o OJ o w C JJ CO *H to x:
Do CO CO iH - a» xi u JJ « 3 >> co 01
ß x. x: CO JJ CO O a. <4-i CO 4-1 4-1 4) - ■H ß iH o o iH CO S > o 0) E e 3 CO Crf •W CO 41 > s^s
o 4) = JJ Vl J= T-4 03 CD |j M O Ü 4) JJ JJ r-t
X: 4H 41 O 1-1 3 a. lj ■i-i x: u o XI ?^ 01 14
4J 4J a. O UH > o u-, C 4) 00 U CO ^H o 4-1 O 41 00 00 c ax: co u
u o ß TJ : u C JJ X) 41 d> 4-1 •H T3 •H X 01 ß «4-1 N CO (tufa 41 JJ o i-i 4-4 ■H O CO c C -H CN CO r-t U 4) •» 3
•o 4) O 00 • 41 o * •o x: 4> «JJ oox» > CJ erf ?s ß 4J > 3 0) •H Ü
•-I CO C -H 4» CO JJ JJ CO 44 4J >>4J 3 CO J-t o c CN CO i-l 00 co 3 01 1-
CO CO U Q -^) V) tu c o P>M-1 O CO XI CO O >H X) co a, XI CO CO
•W *4~1 3 41 C 1 • JJ iH « JJ ai > co <-l 4-1 1—
a. c 00 cux: •H W w In HI O0 a 4) kH 4J Q 41 0> CJ -H •H o a) 3 CO 44 X co •o M 00<4-l o 01 CO 41 • 4) O 41 CO
CO o 0.4J ß >-> Ol 41 iH H ß CO CO 41 1-1
X! 4-1 4) 00 00 JJ 3 JJ I-IJH CO c ß ß u CO JJ to O 41 O iH t-i O CJ o -C 4J CO fttJ-O 4J •H 3 4J 00 o oo g CO CO CJ XI 1-1 H c CO 3 O O O CO c 4J S -H t4-| CO CJ H H 1*4 •w co
* f~-r-.i-*.CJ\r-.i-!OCMCN.-(00 Mv.lsvOlOvOiOl/11/llfilOlfl
I I I
4-t Crf CO £-t Crf 6-S NNn«SQ03N4l« e
E-< C co 3 rH ^ CO
EH
OOCN-*r~r~-cn<Ni_|OcO!-IO CN i-l r-J. CN iH CN i-l
U 41 ' >^ 41 > JJ
X> T-t 0) -H E 4J 3 x: -o > 3 UH C7 c; 4) -H a JJ oi •H cfl »NU
ß X) ß x: 1-1 T-l CJ V4 4) CO £> CJ O ß 3 OHJ CO ■w Cfl X) JJ 1 M CJ 00 o CJ X) O Cd J-t t-t CO ß f^ H erf fa O fa fa CO fa
« XI 1 ß e ß oo •H . I-l CO 01 C TH c/> 4) 0) 4J •H CO z co a
XI CO 01 < 00 c o c t-\ C 41 ß co I-l *r4 Q. O 4J CO •H -rH 4J X E -H 3 •H XI iH a iH O JJ XI x: ß co 41 o U CO iH 4J •H >
r-4 x: 3 In 44
9 X- JJ JJ •* 41 JJ • CJ ß 4) u x:
0) CO —^ 3 O .-4 4) JJ OO a) u ■H O O T4
u l-i 0) 4H x: rH 4-4 CO x> XI o s JH O
rH IH CJ CO Cfl -a o O »H CO 41 S-«
>> ß W CO VD x> <0 4J CO ß i-H
CO 4H /-^ < O XI ß x: O CT\ 0. U 41 o JJ CM 3 O N •H 0) * 4H
-H JJ ß 4) XI u X) 14 CO •H H ß o 41 X) 4) iH 00 CO 4-1 CO 4) JJ •H 0\ ai CJ CO JJ O JJ n XI r^- CO x> c CO ß CO 4H 3 iH 4) XI x: ^w' >% o CO > ß 00 CO JJ CJ
X. IH co iH to -H •H CJ CJ ai x: JJ x: r-t CO
OulO to JJ i-l CO >»<o CO XJ X3 <N
iH x: 4-1 CO fa CO XI CO o JJ
Vl o 4) ß CO to o 4-1 H CO 4) 41 « JJ X> M »H i-W CJ to CO ^^ JJ 3 «-I (0 4) •H OO CO JJ C 0)
4H 00 SH cn i CJ o u ß CO JJ 3 •H 3
O CO > - CO IH JJ JJ TH x: •^> f- o JJ O CO e CJ cn a. CO e u o CO 41 4) ß #--N CO -XI 4) JJ o o 41 <x> c x: CO -H JJ o u m co JJ JJ p-l 3 fa JJ >—• CO CO "*-• CO 01 o 4) 4)
X ß JJ 00 ß 41 4) JJ IH iH OO in 3 CJ o 4» iH >% 3 1H a" ß -H > 4) r-t CO (0
-H CO iH O CO JJ E ß JJ 4-1 CJ tO ß C XJ 3 U ß 0) X3 CO IH D
3 o 1H CJ a) co CO XI CJ x: •H JJ
•H ß O JJ 44 JJ 01 x: o o o o -rH co x: H o JJ fa X) ß CX JJ
co JT> co CN o\ sr o o\ o\ o\ o\ CO vO vO
I-l 0*M erf n orf a n Q ä erf sc Ä o H < sa i co « i erf i I ß i t c < J a. o a, n. o co
v£> oo r- iTi cy> CT> --I co co to ro ro CN co
0) >» 3 CJ O O o- iH •H ß
-H e •• E 4» ß o In O iH
^J ß O ß rH r> JJ O --H
1 4J O JJ to 3 CO 3 OJ
r J < fa <: oi fa
4) 3 C
c o o
X
c 0) 4-1 c o o
u o u y CO
•4-1 o
•u y CO J-l 4J
CO
X <
CO CO ÖJ a
.H rH Xi xi
CO- tfl in •r-4. IJ 14 CO Ctl > > U 0J ^ H M <a •2 x)
£
3 01 I o x O r-> u> CD
u-i O) x tw x u <o u ■H St 00 o
d iH
tu J-I x x A-i 3 4J CO 0) to to • 0 C 4J 4J O
CO —I C HI ii) 01 CO J^
CO o> 3 > cn o
u
I to ?-. CO y 41 to C 3
01 0) x -o *J to
d
<y ol t-i > lfl -H
4J 0) -H (30 CO c o
CO
■H O) r-1 > •H -r-t XI 4J
Cfl O CO
CO U C 01
•H 4-1 N
t,w 1) l-l
C 13 o m c
r~4 G3 u x>
U-i CO P> < H AJ
w -H T3 > C £ -H ft) CO
M-l iH eo a 3 C COCO £
Ü C "H -H
WO Kl O cfl d
O -H CO rH >> tO r-1 C 4J C •H d £ "3 co i-l ca cj O -H « Hm u
•H C « C «I > to c
•H -H O 4J en o.
0) O Ol y X *J to c r>> Ol (A J->
CO cu v e -H O XI CO c
o) o
u to
X -H 60 3 /~v O <£> X CM
< O
y •H Pi
e -o •H C
CO
e o cj
10 4-1 O 3 ax
4) <D 3 > X O 1-1 HH u
o u«
00 >4-t e o
1-1 ■Ö 0)
cfl 00 o c
.H -H •3
41 CO > O i4 r-1 4-1 •H CD
CO > O -H C 4-1
c c 0 o
-H -H 4J U a. cu 0) 4-> y -H VJ J-4 01 o a.
ai H > U -r-t O 4J
»44 Cfl 00
•o <u c c to
cu >iX 4J 4J 1-C
> *-» 1-t cfl •u X CJ 4-1 3
T3 4J O O U cfl D. 4-1
C Q> -H X
4_) •a c « a) ^ n o
X 0) CJ > «
•H X 4J O CO to
ai c c ex
CU X •O 4J
CO O U-l
1-4 O
<1) C u o Cfl -H
CO ^ c oo a) rs B
•H -T3 T3 c cfl a) >
CN 4->
a) « <-■
l-i CO <y 4-> CO •H -H VJ X y 4J
CU 4-1 > cfl
■H X
C o G ij CO 3
*o (0
l-l 01 c:
1 X 3 CO
cfl 3 cr CU -a to c
w -a o o o. o
X 4-1 •<-< 3 ^-1 o a»
•H CU ^-1 X 4-1 O
X VM 4-1 o
CO O 4-t > CO
4-> CU 00
T3 00 C 3 cfl CO
l-i O
4-> CJ
•H •V 0) V4
Sf -J N MO * « « O
lit ll
i 4-1 CO t 3 X
"-! 3 -a <n to
cu ai ^-i > X
•r-1 Cfl U-. u
O <u > X to
X o
0) > C «H
•H 4-> U-I CO CU Ü
•o *^ -o
co d
cu oo ^ c 3 -H
4-1 0) CJ X 3 U C 4J S-l
to CU 4J
41 4-> CO Cfl O (X X 3 C
M) l-l <H O CO 4J y cu CO x
cu - 3 tfl C7* -H
1-1 l-l
M 01 o X o 4j a.
TT CN CU
-Ö • 3 O
^-1 5*. y •«*• c
-H C o
CO 1-1 to u C a) l-l 4J *J -H CO >-i VJ y
4-> cu ■H > cfl i-l M 4-1
4-1 CO to
01 CU X C 4-1
a) >4-l X
°+H CO 4) •• i-l CU a. u £ tfl to X « t! O)
U o o to
c o
•T-I
o s CJ tu >
to c O cfl a £
M cu o X iw 4^ U
CU •o o. c cfl CU
iH -X
?s tfl 4J !-• -H O rH > i-l to X M-t cfl 4J «4-1 CO o c
-r< 00 e
r-1 -H CO 4-> C <o O T»
VI 4J CU l-l M X 01 CO AJ O 0) n q i-i 3 o y tt-i y 3
d o d 4.»
•H C4-I O to
-a d a) 4-1
CO CU cu u
3 en tfl .
■H 0) -H «ex 4-1 1 m-au
•H O «H W -I cfl I
> d o
to o
CO U-4 ; i-l o
•O 00 d <N ca
d si- 41
4J d cu u e o
XJ 4J co y 3 tfl
*—) u-i •o CO c o >-l 1-1 cu U C 0)
•H 4J l-l 1-1 CO t-i e u
U sj- CU <M
cfl U i-l (0 M 4.1 o> ■H *J r-1 1-t
e y
X cfl U o > CO
-ctOO ■Old C d i-i fO JJ -a
CO tfl -X o
>. 4-1 i-l 4J •H 0) CJ rH 1-1 > •H t» -H X o> 4-» to X tfl
4J 00 ao Hi
•H O. C X >. to X tu 1- X HI HI U
T3 X tfl 4J • CU o H 41 O
>-i H - o 4-i a. • ■H 3 4J rH CO C i-l 4) X O £ tfl 4-1 4-1 AJ CO CO tJ 3
d «n H m -o tfl 4J Cfl
X o CO 0)
sr tu
d tfl
X CO H
41 •o y ai ß
4-1 tO O -rl d I-I
cfl CO >
-^ cfl
• 4J d o O 44
1-t 4-1 CU •H X CO 4J o a. >H o o M cl.^s
a> co i-i }-i X o
4-1 4J H c O 3 p. o p. y 3 y CO tfl
r- M o\ o\ o O rs i-w >o iO <t <f I I
O
s
6 »I r>S 3 U-i + w <; tu
M > l-l M >
«NSH 1 1 < 1 f*. (K
M > HU r-1 H > M rH O M 1 1 1 1 M | It, fjL4 [i, b |
fe d o
C->
O tO CO r-1 1-4. r-1
lO N 00 «M CM CM
^c r^. oo sj- in ro CM <N CM tN CM f>
t-l CU X 0) > y >> X i-l cfl 4-1 £ 4-1 X 1-t 3 cfl r-1 y > z: 4-i 01 u •H
d X o 4-1 >-l U 41 ftl « y o O t~< -3 01 4-1 4-1 ! •4-1 Ü a T3 •4-1 rt CO C c 1 <ti t-
U, ft) b
x> m l-i ■u )J 01 d o d 0) > •H fi cd IJ 4-1 U. ffl CO
e 3 1 X •r-1
3 •3 ST C/l <
U.
u
•a d CO
to 4)
r-l X
cfl H
T3 OJ d
d o y
o X £ ^ to
s a) LO
r-t X CU to y i-i d u 41 cfl • IJ > TJ 01
0) cw CU 4-1 01 X 4J M 4-1 1-1
£ S •4-1 o o O (-1
01 •4-1 oo M c to •3
•H 01 d CO 4J Cfl r-1 U a> eg CO £ g u
•H 4-1 0) Cl CO X 01 X H o ■f ■rt * + *
IV). The abstract of the content of F in Table IV describes a
unique factor, identified wholly by a pattern of Rorschach
indices. In this context it is important to note that F
contains no elements of either the ANS scores or the adjustment
criteria since none of these variables (nos. 24-39 in Table III)
load F significantly.
Similarly, F2 is a unique factor, defined entirely by
significant factor weights for 7 of the 11 ANS measures, (Nos.
29-39 inclusive in Table III). It is seen in Table IV that F is
referred to as an "Autonomie Resiliency" factor, so labelled
because the ANS indices with the highest weights or loadings on F
are dynamic as opposed to static, that is they measure EDC
changes to, and recover from, stress induction rather than
baseline levels (see Table II and the methods section). Again,
the absence of significant loadings by any of the Rorschach or
criterion measures (Table III) attests to the uniqueness of this
factor.
F appears to be another Rorschach factor but somewhat
different in structure from the first Rorschach factor (F )
extracted in this analysis. In the first place, the total
Rorschach response measure (R, No. 7 in Table III), loads F
positively but is negative on F3 . Thus submariners who give the
most responses tend to obtain a higher factor score on F , the
Productivity Factor, but at the same time a lower score on F .
Secondly as the abstract in Table IV suggests, the major
difference between F and F is that the latter is marked mainly 13 J
by high positive loadings of color-determined percepts (Csum , No.
14
10, and Affective Ratio, No. 15). Together with a negative
weight by M, movement responses, (No. 23 in Table TV), these two
affectivity scores suggest a trait pattern characterized by
emotionality directed outwardly*.
While labelled differently, Rorschach factors marked mainly
by responses to the chromatic aspects of the Rorschach cards have
been identified in other factor analytic studies. For example,
one investigation which identified a "Productivity" factor, also
17 found a factor called Unique Low Form Dominance . This
dimension was marked by high positive loadings by color responses
and by negative loadings in torm-determined responses, the latter
marker variable being consistent with the loading on F3 of -.57
for variable No. 13 (F+%) in Table IV.
In sum, since a Rorschach factor, identified largely by
color-determined responses, was identified in at least four
studies involving quite different population samples, some
assurance can be gained that a "Rorschach Color Factor" does
indeed exist. Moreover, when found in a given population, this
factor presumably suggests the presence of some complex emotional
trait dimension associated with significant extrovertive and
impulsive tendencies. Also alluded to in the abstract of the
content of F in Table IV are the interrelated findings that
criterion variable No. 26, shown to be indicative of unfavorable
military performance , loaded positively on F . At the same
time, the two favorable performance criteria {Nos. 27 and 28)
*Factor-analytically derived patterns of Rorschach measures have also been used to describe classes of psychopathology
15
received negative weights on the same factor. Taken together,
these findings suggest that the Rorschach Affectivity Factor (F )
may be predictive of inadequate submariner performance underway.
The proposition contained in Table IV that submariners who obtain
a high score in F also tend to possess strong extrovertive and
impulsive traits appears to further support this predictive
relationship since these traits tend to characterize a
significant segment of those submariners who develop debilitative
18 psychiatric symptoms at sea
The configuration of factor loadings in Table IV presents the
clear outline of F as another unique factor defined almost
entirely by the five criterion variables numbered 24 through 28.
Thus a nuclear submariner will earn a high score on F if his
score on the negative criterion F-III (No. 26) is low and/or his
scores on one or more of the positive dimensions {Nos. 24, 25,
27, 28) is high. That is, the higher the score on F the more
adequate the submariner's adjustment is likely to be.
DISCUSSION
Beginning with the establishment of the U.S. Naval Submarine
Medical Research Laboratory at New London, Connecticut in 1946,
the history of Submarine Psychology has shown periodic attempts
to sift through masses of data in order to identify
psychobiological dimensions usefully predictive of individual
differences in submariner adaptive capacity . The present study
is one such exploratory, multivariate investigation designed to
sort through several classes of potential predictors of nuclear
16
submariner adaptability. As stated earlier, these data were
collected in the sixties, and, as a result, the findings may not
be applicable to the modern nuclear submariner population.
20 However, there are data suggesting that the distributions of
submariner aptitude and personality test scores have remained
relatively constant over the past 20 years or more. While
largely speculation, these findings argue that were this study
replicated at this time, essentially the same conclusions could
be drawn from the data.
The interrelated findings that the nuclear submariners in the
present sample obtained significantly lower Rorschach movement
(M), major detail (D), and productivity(R) scores (Measure Nos.
23, 2 and 7, respectively in Table I) than did a comparable sample
from the civilian sector should be mentioned. The lower
productivity (R) of the submariners may have been the result of
their interpretation of the Rorschach procedure as having low
face validity for nuclear submariners, the end result being
reduced motivation and hence low response scores. While low "R"
scores are sometimes indicative of a lack of originality most
16 often found in groups of average or below average ability , this
would not seem to be the case in a superior ability group such as
the present sample of nuclear submariners.
The lower average number of movement responses (M, No. 23 in
Table I) found in the submariner group, particularly in the
context of normal or slightly elevated C (color-determined
responses. No. 10 in Table I), suggests traits associated with
less suppression of acting-out behavior, or, as described in the
17
'abstract for F in Table IV may indicate trends toward impulsive
and extrovertive behavior-much more at least than are found in
the civilian sample used for comparison. Too, the negative
factor loading (-.39) of Experience Balance (No. 9 in Table III)
on F would lend additional support for the presence of
personality traits such as outgoingness and unrestraint as more-
or-less characteristic of those submariners who earn a high score
on this factor.
In terms of the major objective of this investigation,
namely, to identify by factor analysis, clusters of valid
predictors of submariner adaptability, what do the results
indicate? Overall, the findings would seem more to suggest the
classes of variables that are not likely to be useful predictors
of submariner adjustment than to pinpoint those that have a high
probability to be useful for that purpose. Thus F , a unique
Rorschach factor, "Organized Productivity" (Table IV), as well as
F , also a unique ANS factor labelled "Autonomie Resiliency", are
uncorrelated with each other and neither is correlated with any
of the adjustment criteria. These criteria in turn tend to
cluster in still another unique factor, F , "Favorable Submariner
adjustment" (Table IV).
There is considerable evidence in the literature of Submarine
Psychology that measures of ANS reactivity and resiliency are
generally correlated with submariner adaptability under most
conditions1'' . Yet, F whose main components were defined by ANS 2
indices, was not related to any of the criteria (Tables III and
IV). Why was this? One possible explanation is low reliability
18
of the criterion dimensions themselves. Another reason has to do
with the possible invalidity of ANS measures taken in shore-
based, laboratory-contrived stress situations for predicting the
adaptability of submariners during long submerged missions,
during which time quite different environmental conditions would
likely prevail.
How is the fact that the Rorschach factor (Fj ), a dimension
marked largely by Rorschach indices of functional intelligence
and perceptual organizational ability (W, R, and Z, Nos. 1, 7,
and 3 in that order in Table IV) does not correlate with any of
the criteria, but that the other Rorschach factor (F3 ), an
affectivity dimension, does correlate with three of the criteria
(Table IV)? The answer to this question suggests an hypothesis
regarding the direction submariner assessment researchers might
be advised to take. Because nuclear submariners are thoroughly
screened in terms of a variety of abilities, the range of
individual differences with this population i.e., the variance,
is greatly reduced with regard to this particular trait area. As
a result, the Rorschach indices VI, R, and Z, which also "tap"
certain aspects of the abilities domain, tend to be uncorrelated
with any of the criteria largely because of this low variation.
In contrast, the fact that F3 whose major components are
affective or emotional in nature, (at least as indicated by the
Rorschach score pattern) does correlate negatively with three of
the adjustment criteria would seem to argue that measures of
temperament and emotionality may be usefully predictive of
individual differences in the adaptive potential of nuclear
19
submariners during protracted submerged missions. This finding
may represent a meaningful cue as to the focus research in
submariner assessment should take in the future.
20
REFERENCES
1. Weybrew, B.B.: Psychological problems of prolonged periods of marine submergence. In Burns, N., Chambers, R. and Hendler, E. (Eds) Unusual Environments and Human Behavior Free Press, Glencoe, III., 1963.
2. Cook, E.B.: A factor analysis of personnel selection data: Intra- and inter-area relationships of biochemical, physiological, psychological and anthropometric measures. NavFieldMedRschLab XI (No. 26), 1961.
3. Beck, S.J.: Rorschach's Test I: Basic Processes. Grune and Straton, New York, 1944.
4. Freeman, M.F., and J.W. Tukey: Tranformations related to the angular and the square root. Ann. Math. Statist., 1950: 21: 607-611.
5. Weybrew, B.B.: Prediction of adjustment to long submergence aboard an FBM submarine I. Interrelationship of the adjustment criteria. Nav. Sub. Med. Res Lab. Rept. No. 383, 1962. ~'"""'"' ~
6. Weybrew, B.B.: Prediction of adjustment to long submergence aboard an FBM submarine IV: Psychophysiological indices. Nav. Sub. Med. Res. Lab. Rpt. No. 416, 1964.
7. Freeman, G.L.: The energetics of human behavior. Cornell University Press, 1948.
8. Weybrew, B.B.: Behavioral energetics I. A factor analytical study of individual differences in modes of energy discharge resulting from experimentally-induced frustration. Nav. Sub. Med. Res. Lab. Rpt. No. 378, 1962. -
9. Hotelling, H.: Analysis of a complex of statistical variables into principal components. J. educ. Psychol. 1933: 24: 417-441.
10. Kaiser, H.F.: The varimax criterion for analytic rotation in factor analysis. J. educ. Psychol. 1956:23:187-200.
11. Fructer, B.: Introduction to factor analysis. D. Van Nostrand and Co., Inc., New York, 1954.
12. Beck, S.J., A.I. Rabin, W.G. Thiesen, H.B. Molish, and W.N. Thetford: The normal personality as projected in the Rorschach test. J. Psychol. 1950:30:241-298.
13. Harman, H.H.: Modern Factor Analysis (2nd. ed.) University of Chicago Press, 1967.
21
14. Weybrew, B.B.: Selection of men for hazardous duty from indices of autonomic nervous system reactivity. Nav. Sub. Med. Res. Lab. Memorandum Rpt. No. 6 5-1, 1965.
15. Molish, H.B.: Schizophrenic reaction types in a naval hospital population .as evaluted by the Rorschach test. 1J.S. N. Bureau Med. and Surgery Project No. NM. 007-089.31, Washington, D.C, 1957.
16. Rickers-Ovsiankina, M.: Rorschach Psychology. Wiley, New York, 1960.
17. Coan, R.: A factor analysis of Rorschach determinants. J. Proj. Tech. 1956;20:280-287.
18. Weybrew, B.B. and E.M. Noddin, Psychiatric aspects of the problem of adaptation to long submarine missions. Aviat. space and env. med. 1979;50:575-580.
19. Weybrew, B.B.: History of military psychology at the U.S. Naval Submarine Medical Research Laboratory. Nav. Sub. Med. Res. Lab. Rpt. No. 917, 1979.
20. Weybrew, B.B. and E.M. Noddin. The Longitudinal Health Study: Changes in the psychological characteristics of enlisted submariner volunteers over the past two decades. Nav. Sub. Med. Res. Lab. Rpt. No. 817, 1978.
22
Unclassified SECURITY CLASSIFICATION OP THIS PAGE (When Data En(errd)
REPORT DOCUMENTATION PAGE READ INSTRUCTIONS BEFORE COMPLETING FORM
1. REPORT NUMBER
NSMRL Report 1080 Z. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER
4. TITLE (and Subtitle)
"Prediction of Nuclear Submariner Adaptability from Autonomie Indices and Rorschach Inkblot Responses"
5. TYPE OF REPORT 4 PERIOD COVERED
Interim report 6. PERFORMING ORG. REPORT NUMBER NSMRL Report 1080
7. AUTHOR;»
Benjamin B. Weybrew and H. Barry Molish a. CONTRACT OR GRANT NUM8ER<»
9. PERFORMING ORGANIZATION NAME AND ADORESS
Naval Submarine Medical Research Laboratory Box 900 Naval Subase Nlon Groton CT 06349 - 5900
«0. PROGRAM ELEMENT. PROJECT, TASK AREA & WORK UNIT NUMBERS
11. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE
9 September 1986 13. NUMBER OF PAGES
22 1«. MONITORING AGENCY NAME 4 ADORESSfU dlllerent from Controlling Office)
Naval Medical Rsch and Dev Command Naval Medical Command, National Capital Region Bethesda Maryland 20814
15. SECURITY CLASS, (of thie report)
Unclassified IS«. DECLASSIFICATION/OOWNGRADING
SCHEDULE
16. DISTRIBUTION STATEMENT (ol this Report)
Approved for public release; distribution is unlimited.
17. DISTRIBUTION STATEMENT (of the »betratet entered In Block 20, It different from Report)
IS. SUPPLEMENTARY NOTES
19. KEY WORDS (Continue on revere» tide It neceaeary end identity by block number)
submariner selection; Rorschach Test; factor analysis; autonomic indices; submariner adaptability;
20. ABSTRACT (Continue on revetee elde II neceaeary and identify by block number)
To identify the most valid predictors of submariner adaptability, the authors derived 23 indices from the responses of 170 nuclear submariners to the Rorschach Inkblot Test, 11 measures of Autonomic Nervous System (ANS) reactivity to contrived stress, and five adjustment criteria. Factor analysis of this 39 x 39 correlation matrix yielded two Rorschach Factors, one of which correlated with three criterion dimensions. Two unique factors were also discovered, one, a structured ANS factor, and the other, a complex criterion scale. Selected Rorschach scores and, to a lesser
DD )JW 73 1473 EDITION OF 1 NOV «5 IS OBSOLETE S/N 0102-014-6601 Unclassified
SECURITY CLASSIFICATION OF THIS PAGE (When Date Bntered)
Unclassified .'-IJ^ITY CLASSIFICATION OF THIS PAGEfWhen Dot» Entered)
item 20 -- continued
extent, certain ANS indices emanating from this study, may be usefully-valid predictors of the adaptability of nuclear submariners during long patrols.
Unclassified SECURITY CLASSIFICATION OF THIS PAGEfWhen Data Enlatad)