h
“an...
.0 1:
Au... ' x.
‘ a O.“ Ii ..
R e T.
O I... g.‘
.3}; I.
fit“ 2. V. 5.3.. s
1 . .. :5. v n ‘
a? u .. :.§.-$ {’30
.. .. 3. a .4.
n- . . AN 5. 5 .0... 1?.

MW 3M m: an. 5.. 3......
Ma...» .1...” ‘3 .. : "I... 4&6

s... .3 1.. «a... IT...
I 05;. n...‘ .l 11’- a:
. .v, . .
\ .. I. L. E)...
0;: o 9‘ 1. O.‘I‘ *
’3‘ a ”.3. I. i b 3...
.O .L ...m._ {1‘ ‘. - I‘fa
. .o f. .. In. .
‘3. ‘1. I «‘0.
c 0.} x .x
H.“ x“ 1 Wt. 63 m... V quid
. ..
an. I“ . ‘ .o I'L «Eh-r
w... .n— JR“ 1.9.».-
irq .- .
”.8 i... W... m
a '1 u l.
C ”‘HI Wu.'” 4““
3a.. ... .84

__£333::___:_:____:___ mmm

 

MillllllllHllllillllHllllillllilllllllllllllllllllllllllllfiL

31293 10381 5621

 

   
   

LWRY
._ 41 113

ncrslty

 

MSU

 

 

 

RETURNING MATERIALS:
P1ace in book drop to
remove this checkout from

 

LIBRARIES
“ your ‘record. FINES M“
be charged if book is
returned after the date
stamped below.
W 0/

 

 

 

 

A SENTENCE COMPLETION TEST TO MEASURE
THE NEED FOR AFFILIATION

By

Paul S.D. Berg_

A THESIS

Submitted to the School of Graduate Studies of Michigan
State University of Agriculture and Applied Science
in partial fulfillment of the requirements

for the degree of

MASTER OF ARTS
Department of Psychology

1959

ABSTRACT

The present investigation was concerned with the
construction of a sentence completion test to measure
the need for affiliation. The purpose of the new
instrument was to avoid the content analysis required
by the TAT method of measuring affiliation which is
used by McClelland et a1. It was expected that the
obviation of content analysis would result in a more
economical, self-administering, independent research
measure of affiliation.

The two tests were administered to a sample of
100 neuro-psychiatric aides at Battle Creek Veteran's
Administration Hospital. The correlation between the
two tests was computed; items of the sentence comple-
tion test which elicited the least amount of affiliation
imagery were dropped from the test and.this revised form
of the sentence completion was correlated.with the TAT.
Finally, multiple correlational methods were used to
predict the maximal weighting of all items on the sen-
tence completion in order to increase the correspondence

of this test with the TAT.

The final form of the Sentence Completion Test
correlated .23 with the TAT, which is significant at
the .05 level. However, the overall results of the
study suggest a positive, but low relationship between
the two instruments. The multiple correlation weights
determined from the first half of the sample did not
hold up as predictors of the performance of the second
half of the sample. The lack of stability of the weights
were viewed in terms of a) the small sample size, b) the
large number of variables, and c) a basic and inherent
difference between the two instruments.

A theoretical discussion of projective tests sug-
gested possible explanations of the results. The theory
of the drive reducing effect of fantasy behavior involved
in taking a test and the experimental concept of asso-
ciative refractory phase were considered. It was proposed
that the method of “validating" or measuring correspondence
between two tests by administering them one after the
other is nd: a sound psydhological procedure. Limitations

of the present sample were considered.

ACKNOWLEDGEMENT

Grateful acknowledgement is made to Dr. M. Ray
Denny and Dr. C.F. Wrigley for their sincere and
active interest in this research. Their helpful
suggestions contributed imeasurably to the comple-
tion of this thesis.

The cooperation of Dr- Stewart G. Armitage in
providing the facilities and time for the author to
collect the data for this research is also gratefully
appreciated.

DEDICATION

This thesis is dedicated to my parents and
grandparents who share more responsibility for my
being in a position to continue my education and

interests than anyone could know.

TABLE OF CONTENTS

Page
IntrOductionOOOOOOOO0.000000000000000000000000 l
MethOdOOOOOOOOOO00O0.00.00.00.0000000000000000 15

Results......o............o.....o........ooooo 23
Discu331onooooooooooooooo0.0000000000000000... 30
Summary....................................... AZ

Bibliographyoooooooooooo00000000000000.0000... uh

LIST OF TABLES

Table I... Pearsonian Coefficients of
Correlation Between TAT Test
and Three Forms of the Sentence
completion Test.................... 2“

Table II.. Pearsonian Coefficients of
Correlation Between the Sentence
Completion Test (Minus the Four
Lowest g_aff Eliciting Items)
and the TAT for the First and
Last 25 Subjects in Sample 1 and
in SW16 20.00.00.000...000.000.... 27

INTRODUCTION

The Thematic Apperception Test1 procedure has en-
Joyed an extensive usage in researches and in the practice
of clinical psychology during the past twenty years. In
this test, human motivation is assessed through.content
analysis of stories written in response to pictures shown
to the subject. Its usage is steeped in clinical-theore-
tical points of view. The theoretical basis for the
measurement of motivation has been in part provided by
Atkinson and Reitmen (3) :

Particular motives, (Achievement, Affil-
iation, Power, etc.) are conceived of as latent
dispositions to strive for certahn goal states
(aims) that are mgaged in performance of an
act when the cues of the situation arouse the
expectancy that performance of an act is instru-
mental to attainment of the goal of the motive.

The total motivation to perform the act is con-

ceived of as a summation of strengths of all

the motives that have been aroused by appro-

priate expectancies of goal-attainment cued

off by the situation.

The TAT has provided a means for the assessment of
these latent dispositions such as need for achievement,

for affiliation, for power, etc. McClelland (10, p. 107)

 

1 Hereafter referred to as TAT.

and his co-workers have written:

The choice of the thematic apperception

method for collecting data followed from our

acceptance of the Freudian hypothesis that a

good place to look for the effects of motiva-

tion is in fantasy. The Thematic Apperception

Test has been developed by Morgan and Murray

(1935) for Just this purpose, viz., for bring-

ing to light the strivings of the individual

in imaginative stories suggested by stimulating

pictures. Its popularity as a tool in the

clinic may certainly be offered as some evidence

of its success in achieving this purpose.

One might question at least two features of this
statement. First, hypotheses are generally regarded as
theoretical prepositions which are acceptable or not
according to whether the empirical evidence accords with the
deductions which are to be drawn from them. McClelland
and his co-workers offer no such empirical evidence in sup-
port of the Freudian hypothesis. Their acceptance of it
seems to be based upon nothing more than an act of faith,
or perhaps an intuitive agreement with it on the grounds
of their clinical eXperience. Secondly, popularity is no
satisfying evidence of the value of any theory or of any
testing instrument. However, the passage is quoted here to
illustrate the high regard some well-known psychologists have

for the TAT as a measure of human motivation.

There are, however, some serious practical and
theoretical disadvantages in the TAT. A major one
has been the difficulty in validating the test, i.e.,
in showing that it is really measuring need for achieve-
ment, etc. as it purports to do. Another has been the
time-consuming nature of the content analysis which is
needed to assess the motivational structure of the sub-
ject. Both deCharms (S) and McClelland (2, pp. 21;. 25)
have stressed the complication of the present system
of content analysis and the need for some simpler method
of assessment.

Before proceeding to outline the present study,
the history of need affiliation2 measurement will be
outlined, and this will be followed by a review of re-
search upon the relations between TAT and sentence com-

pletion measures of motivation.

T_h_9_ History if; M Affiliation Measurement

After the initial work on the fantasy measure of
hunger and of need achievement, research was initiated
which dealt with g aff. The first attempts to measure
2 eff arose out of Birney's work on the need for secur-

ity, which was regarded as not general enough to be a

 

2:I-Iereafter referred to as g eff.

A

valid measure of human motivation (13). Shipley and
Veroff (13), in 1952, published the results of their
studies "validating" n eff. They used two designs: 1)
An experimental group was affiliation-aroused before
the TAT testing by the administration of a sociometric
test; the control group, for their initial activity,
received a food preference test. They hypothesized that
the affiliation-aroused group would have higher affilia-
tion scores on the TAT. 2) The stories of groups of
students that had been rejected from fraternities to
which they had applied were analyzed. It was hypothe-
sized that there would be more n eff scores in this
group 's stories than in a group who had been accepted
by the fraternities of their choice.

In both designs 1 and 2, the results were in the
expected directions; in Design 1 the experimental group
scored higher on affiliation on 5 out of the 7 categories
that were employed and in Design 2 the rejected group
obtained higher 3 eff scores than the accepted group,
confirming the predictions. A further bit of evidence
for the validity of the measure was the negative corre-
lation they obtained between popularity as measured by

the sociometric test and n eff scores. The theory behind

5

this view of n eff is primarily one of rejection and
in fact, Shipley and Veroff (13) defined n eff in terms
of separation anxiety resulting from the separation of a
person from objects of affiliation. This conceptuali-
zation has been revised by Atkinson, Heyns and Veroff
(2) in 19514. who came out with a broader definition of
3 off:

Any evidence of concern over establish-

ing or maintaining a positive affective re-

lationship with another person(s) (in addi-

tion to concern over restoring broken rela-

tionships) is viewed as symptomatic of moti-

vation to affiliation according to the present
results.

Atkinson, Heyns, and Veroff (2) conducted a similar
study to Design 1 of the Shipley and Veroff (13) study.
Similarly, they had their experimental group partici-
pate in the sociometric ratings in the fraternity, but
the control group had their non-affiliative related
task administered in a classroom setting rather than
in a fraternity dining room as the Shipley-Veroff control
group had. Furthermore, the pictures that were used
were assumed here to be more every-day like, as opposed

to the Shipley-Veroff pictures which appeared to elicit

stories of basic interpersonal problems. The combination

of the more clearly differentiating testing conditions
and the more "natural" pictures, combined with the
broader scoring system, yielded a significant difference
between the two groups, thus producing a broader scoring

conceptualization of g aff.

Evaluation of; the Characteristics 2f the TAT
5nd t_h_e_ Sentence Completion _l_‘e_s_t_§

A Sentence Completion Test has been suggested to be
similar in function to the TAT. Better (12, p. 296) for
example, has stressed the apparent similarity of the two

methods :

When used for clinical purposes, the
sentence completion method is very similar
to the Thematic Apperception Test. That
is, it tends to give information of person-
ality content rather than personality struc-
ture. It deals with feelings, attitudes,
specific reactions to people and things,
rather than providing a so-called underlying
personality structure.

Rohde (11) has similarly regarded the Sentence Com-
pletion Test as a projective technique:

The sentence completion device in which
the subject is asked to read to himself the
forepart of a sentence and to write anything
he wishes to complete the sentence is essen-
tially a projective technique utilizing free
association.

The possibility would therefore seem to exist of
developing a sentence completion test for n aff” measur-
ing much the same thing as the need affiliation score
on the TAT. The new test would be self-administering,
so that a skilled administrator would not be needed;
and less time would be required for the test. Such a
test, if secured, would therefore be of greater economy,
utility and research value than the TAT.

The purpose of this study was to try to develop
such a sentence completion test of n eff. If this second
measure of need affiliation can be developed, and the TAT
measure of g aff is found to correlate highly with it,
we thereby get some assurance at least about the general-
ity, if not the validity, of whatever psychological
function is measured by the TAT score. If, on the other
hand, little or no correlation is found between the two
intended measures of n eff, it will serve as a reminder
of the specificity which may possibly be characteristic

of certain projective tests.

Regegch Upon TAT m Sentence Connletion Measures
2;: Motivation
Many investigations have already been concerned with

finding the relations between the TAT measures of motivation

and those provided by other standard tests such as
Sentence Completion and the Rorschach tests. In gen-
eral, the correspondence of the various instruments
has not so far been demonstrated. we shall now review
some of these studies.

Although McClelland (l, p. 31) says that the only
response system.which.has been extensively explored to
date for unique motivational effects is organized
fantasy in stories, there have been many attempts not
only independently to develop new'instruments, but
efforts directed towards "validating” popular standard
tests like the Sentence Completion and the Rorschach
with the TAT measures of motivation (9, p. hll; 10, pp.
25k. 255) .

Other alternative methods, seeking to avoid the
content analysis of stories and/or the verbal factor
which might be limiting the validity of this method have
been reported, i.e., Graphic Expression Test by Aronson
(1, Ch. 17), French's Test of Insight (1, Ch. 16),
deCharms' Self-scored Projective Measure of Achievement
and Affiliation Motivation (5). and others (6, 8).

we shall new review some of the studies which have

employed sentence completion tests in relation to TAT,

either for direct comparison of the two instruments
or for a variety of complementary purposes.

As far back as 19113. Sanford (11, 12, p. 299)
utilized a 30 item sentence completion test in order
to measure need action patterns among children. The
data were analyzed according to the Murray need-press
system of personality description and when combined
with data from a ”Completions of Pictures Test,‘ and
an "Interpretations Test," it appeared to Sanford that
they revealed consistent patterns of responses in many
areas. Unfortunately, Sanford did not submit his data
to any quantitative analysis.

In 1914.6, Rohde (12, PP. 299. 300) published results
of a study using a 61.]. item sentence completion test.
She also analyzed the data in terms of the Murray sche-
ma of needs, press, inner states, and cathectic objects.

Wheeler (12, p. 298) used a variation of the sen-
tence completion, via, a “Similes Test" where subjects
were presented with a list of adjectives and responded
with similes of a similar nature. Again Murray's con-
ceptual framework was used.

McClelland (10, pp. 2511-257). in describing a study
by Holzberg using a 50 item sentence completion test,

10

says:

Another rather commonly used instru-

ment for getting at an individual's per-

ception of the world is the Sentence Com-

pletion Test. For us it has the added

advantage of being another so-called 'pro-

jective technique,’ the results of which

could be conquered with those obtained with

the picture-story or TAT projective method.
However, results of the two tests showed very little
agreement, and McClelland proceeds:

The fact of real importance that emer-

ges from this study is that the Sentence

Completion Test gives results which differ

from those given by the thematic appercep-

tion method and which are more in line with

conscious self-judgments.

Bron and Auld (6), in a 1951.1 study to investigate
the effects of confinement aboard a submarine for 30
days, measured personality variables both by the TAT
and a Sentence Completion Test. On the qualitative
level, they professed to find similar personality trends
reflected in both measuring instruments, and accordingly
suggested that similar variables were measured through
the use of either test. Indeed, perhaps their most val-
uable contribution may have been some of the critical

insights about the nature of the TAT and its limitations

11

which were revealed in the process of this research.
(See P.38). However, as in Sanford's study (11), there
'was no quantitative evidence supplied in support of
their position.

Lindsey and Heinemann (8), compared the TAT results
with those of a 100 item.Sentence Completion Test (an
adaptation of the OSS scale), with questionnaire answers,
and with scores on the MMPI, as part of their work on
developing group-admhnistration techniques of the TAT.
The correlations were mostly positive but low.

Lindzey and Heinemann.tried to explain the results
of their study on the basis of imperfect scorer relia-
bility and the theoretical notion of levels of assess-
ment.Reliability coefficients on the TAT ranged from..60
to .93. and on the sentence completion from .62 to .85.
The authors argued that with the lower reliability coeffi-
cients, quite a high percentage of the variance was not
accounted for and that this factor might minimize corre-
lations with the other measures. They suggested further-
more that the measures might be assessing motivation at
different levels. Introduction of this concept of assess-
ment levels does not, however, provide any conclusive

explanation. At the most, it offers suggestions for

12

research to explore...

French (1, Ch. 16) developed a lO-item "Test of
Insight" which.was hypothesized to offer subjects an
Opportunity to make responses relevant to their parti-
cular needs. A sample statement is, "Bill always lets
the 'other fellow' win" (1). The subjects were asked
to respond to each statement in terms of an explanation
of why each.man behaves as he does. The subjects' scores
‘were correlated with observational judgments of motiva-
tion and goal attainment for both 'affiliation' and 'ach-
ievement' and also with.their performance on a question-
naire. The correlation between the Test of Insight and
the observational judgments ranged from.-.05 to .05. The
correlations between the Test of Insight and the ques-
tionnaire were .18 and .19 (significant at the 5% level)
for the need affiliation variable. The content of her
test seems to be similar to a sentence completion form,
except that there is a greater latitude of response pos-
sible with her items.

Feshbach (1, Ch. 11), investigating the drive-reducing
function of fantasy behavior, used a.slightly modified
version.of the RotterJWillerman Sentence Completion Test

as another measure of aggression to be compared with the

13

TAT measure. He reported that the sentence completion
test was sensitive to the arousal of aggression in the
experimental condition, and.yielded significant differ-
ences between experimental and.control groups. One of
the possible implications here is that the sentence come
pletion test can be ”validated" through the use of ex-
perimental arousal, which is the primary method that has
been used in validating the TAT-type method of motive
measurement. (For a further discussion of this, see p.357.
McClelland (1, p. 3h), after reviewing much of the

evidence in this area, draws the following conclusions:

0n the other'hand, if the verbal one
is a.portion of a sentence which the sub-
ject must complete, the content of his come
pletions is not related to the content of
his stories to the pictures. Perhaps this
reality component is larger in the sentence-
completion than in the picture-story or
"motive-insight" methods. Although.really
systematic work in this area.has not been
completed, the safest practical generaliza-
tion appears to be that cues will tend to be
equivalent so long as they suggest vaguely
the content area of the motive being measured
but do not elicit strong or specific associations
based on.particular cultural or personal past
associations.

This generalization, in and of itself, does not seem
to warrant predicting that sentence completion tests will
not be tapping similar sorts of information as the TAT. I:

1h

the fact that there is more conscious control on the former
is not a major detriment to correspondence of the two tests,
then we feel that there is enough similarity in their appar-
ent nature to warrant further exploration, i.e., both tests
deal with personality content (see p. 6 ). Secondly, both re-
quire the subject to offer the linguistic responses that he
has available. With.reference to the latter, Solomon and
Howes (15) have said:

...it follows that in perceptual experiments

in which linguistic responses are elicited from.hu-

man subjects, 'any variable that is a general pro-

perty of linguistic responses must also be a property

of any perceptual concept that is based.upon these

responses'...

Also McClelland reports (9, p. All):
we have found that n Achievement score is sig-
nificantly related to the kinds of linguistic cate-
gories that a person uses in attempting to express
himself at least in an achievement related situae

tion, and so forth.

Certainly there is not enough conclusive research in the
entire motivation assessment area which allows for the formu-
lation of precise statements of relationships between test:
and/or variables.

In the course of examining the correspondence between the
newly designed sentence completion test and the TAT, statisti-
cal procedures will be used to eliminate those sentence comple-
tion items which are not effective in eliciting affiliation
imagery. A multiple correlational analysis based upon the pre-
dictive value of the individual itemm will indicate the items
to be retained or dropped. This is likely to be more sound than

the "hunches" involved in intuitively dropping items.

15

METHOD

Apparatus

TAT- In this study data for need achievement,
need power, and need.affiliation were collected al-
though only the §_aff data are relevant to the present
design. A set of six TAT pictures used in a national
survey study conducted by the Survey Research Center,
University of Michigan, and developed by Joseph Ver-
off (1, p. 836) was used. It consisted of pictures
2, 101, 102, 28, 83, and 103 which are described in
Appendix III of Atkinson's book, ”Motives in Fantasy,
Action and Society"(l)."l A study is currently being
conducted on a nation-wide representative sample which
should provide normative data, but which.will not be
available for a few months. (Personal communication
from Dr. Veroff).

Sentence Completion- The second test used was a
sentence completion test consisting of 61 items: 1
blind 'neutral' item.for practice and demonstration,
20 affiliation items, 20 achievement items and 20 power

items. The compilation of the 20 g'aff items was as

 

h Reitman and Atkinson (l, p. 36) recommend multimotive
scoring of the same pictures because it has been shown
that there is no correlation between achievement and
affiliation scoring.

16

follows: Murray Melbin, an advanced trainee at the
Veteran's Administration Hospital, Battle Creek,
devised 20 affiliation items with face validity. (Items
h, 19. N5, and 57 came from a study by Dorris, Levin-
son and Hanfmann, Authoritarian personality studied by
a new variation of the sentence completion technique,
is .9220 29.20 321322.310: 9-2. 1951+. PP. 99-108, and

the remaining 16 items were intuitively constructed).
The 20 items consisted of 10 pairs, i.e., items of
similar content, either in first or third person, 6.3.,
"I like people who," or, ”He likes peOple who,‘t etc.
The items were administered to 23 psychiatric aides

at the Veteran's Administration Hospital, Battle Creek,
with.instructions to complete the fragments. The data
was scored for presence or absence of affiliation imagery
by 2 scorers with experience in g;aff scoring (1, Ch. 13).
Six of the pairs,5 when scored by the two scorers, re-
sulted in significant agreement between the two scorers
at the .02 level or better, using the Sign Test for
Paired Observations. All of these items also elicited
aff imagery in at least 25% of the aide sample. To
these six pairs, four pairs of affiliation items with

face validity, constructed by the writer, were added,

5 These 12 items are numbers a, 13, 19, 2h, 31, 38,
AS, 53, 56, 57, 60 and 61, on the final sentence
completion form.

 

17

yielding the final 20 affiliation items. The serial
order of presentation of the affiliation items in the
final 60 item.test was based on the following consider-
ations: 1) Throughout the 60 items, the first and third
person items were in ABBA order. 2) A judgment was
made by Melbin and the writer as to what each of the
six TAT pictures seemed to depict best, i.e., achieve-
ment, power or affiliation situations. The ordering
of the sentence completion items was then made to
correspond roughly to the ordering of the TAT, accord-
ing to the judgment of the writer. For instance, the
first TAT picture of the six shows two men working
around a machine and was assumed to be a good achieve-
ment picture. Therefore, the first sixth of the 60
sentence completion items (10 items) consists of a
heavier representation of achievement items (6 achieve-
ment, 2 affiliation and 2 power items). The same prin-
ciple was followed throughout the remainder of the item
ordering. 3) An attempt was also made to keep similar
first and third person items apart from.each other in
order to minimize copying one to the other.

Other Materials- In addition to the Sentence Come
pletion Form, the TAT cards and blank for recording tie

the responses (see Appendix), each tester received a

18

guide for the interview, which.included introductory
remarks and an explanation to the subject of the pur-
pose for testing him, instructions for administering
the two tests, standardized responses to various ques-
tions that might arise during testing sessions, and
concluding remarks. A copy of this interviewer's guide

also appears in the Appendix.

Sample
The sample is composed of 100 randomly selected

male psychiatric aides employed in the day and evening
shifts at the Veteran's Administration Hospital, Battle
Creek. (Female aides were not used because there is
some evidence that females behave differently on motive
measurement than men (10, pp. 172-181). Only one of
the aides in the sample refused to be tested and was
therefore replaced in the study. The sample was appro-

ximately equally divided between Negro and.White aides.

Procedure

Pro-Study Preparation- Approximately two-thirds
of the tests were administered by the writer. The re-
mainder of the testing was completed by three psychology
trainees and a pharmacist at the hospital. Prior to

the actual testing, conferences were conducted with these

19

persons, at which each tester had an Opportunity to
interview and administer the TAT to a "practice" sub-
ject. Following these practice sessions, discussions
as to how the interviewer handled the situations that
arose were talked over. It was hoped that these con-
ferences would result in a more or less standard ap-
proach of the various examiners to the subjects -- a
problem which is non-existent in group testing situa-
tions. From these conferences and from.the interview-
er's guide (see p.19) it was possible to approximate
the "relaxed," and "normal," conditions that are re-

commended by McClelland (1, p. 36).

scenes

The TAT protocols were scored by Melbin and the
writer. In order to become expert scorers, each of us
trained himself with the affiliation manual (1, Ch. 13)
and the affiliation practice materials (1, pp. 735-776).
In order to compare our proficiency with the criteria
recommended by Fold and Smith, (1, Ch. 15), we computed
a) percentage agreement between each of us and ex-
6

pert scoring on the presence of motive-related imagery,

and b) Pearsonian coefficient of correlation

 

6 This method is commonly used for reporting scoring re-
liabilities for content analysis. The formula is:

2(number of agreements between S and expert on the pre-
sence gf_imagery)
number of times S scored imagery plus number of times
expert scored imagery present. (1 p 238)
, O

20

between each of us and expert scoring on the total
score assigned each subject. These statistics were
computed for Sets F and G, the final sets of practice
stories which are used as the criteria of expert scor-
ing. (see Appendix). On Set F, Melbin % agreement
was .93 and the coefficient of correlation was .97.
The writer, on Set F, obtained percentage agreement

of .95 and Pearsonian r of .93. On Set G, Melbin's
results were .89 and .97, respectively. The writer,
on Set G, obtained .9A and .9h. In all cases, these
results compare most favorably with.the results reported
by Feld and Smith (1, Ch. 15) for research.purposes.

The entire scoring of the sentence completion items
was done by Melbin. The same conceptual framework and
scoring system as applied to the TAT (1, Ch. 13) was
utilized for scoring the completions. Hewever, in the
case of the sentences, the fragments that were presented
to the aides could in some cases themselves be scorable
as affiliation imagery. Thus, for a completion to have
been scored as containing affiliation imagery, the 93m-
pletion itself must have contributed to the sentence's
imagery of affiliation. For example, if the fragment,
"I like people who,” were completed by, I'eat lobster,"

affiliation was not scored, whereas it would have been

21

if the completion read, "like me," On this basis,
each completion was scored either 1 for presence of

aff imagery or 0 for absence of aff imagery.

Experimental Design

For purposes of statistical analysis, all aides
were administered both tests (one half took the sen-
tence completion first and the other half took the
TAT first) and then randomly divided into two samples
of 50 each, hereafter referred to as Sample 1 and
Sample 2. The tests were given one after the other in
order to measure any one individual at approximately
the same time, since, in this case, motivation condi—
tions may be assumed to be as nearly the same for both
tests than if they were given days apart. (McClelland,
(l, p. 33) reports how difficult it is to re-establish
the same set in a subject).

For Sample 1, the subjects' scores on the sentence
Gompletion test were correlated with their scores on
the TAT, using the Pearsonian coefficient of correla-
tion. It was hypothesized that if the four sentence
completion items (of the 20) which elicited aff imagery
the fewest number of times be dropped from.the test,

and a correlation with the TAT score be recomputed on

22

Sample 1 using the remaining 16 items, the resulting
correlation would be higher. The next step was the
determination of the maximal multiple correlational
weights for the 20 items with the TAT.

For Sample 2, the same initial statistical pro-
cedure was employed; however, this time, the corre-
lation.weights determined from Sample 1 were applied
to Sample 2. Once again, it was expected that the 1h
item sentence completion would correlate more highly
with the TAT than the original 20 item test. Further-
more, it was predicted that when the correlation weights
from Sample 1 were applied to the 20 items on Sample 2,
a higher correlation would be obtained. The final
step was to compute the correlation between the 1h item
sentence completion test and the TAT for the full sample
of 100.

23

RESULTS

The major correlations between the TAT test and
different forms of the sentence completion test used
in this study are presented in Table 1. They range
from..ll to .2h. The final correlation for the full
sample of 100 with the four least effective items drop-
ped, is .23, significant at p.<§05. For Sample 1, 2
and the total sample, it can be seen that the correla-
tions between the TAT and all 60 sentence completion
items, which includes items designated as achievement
and power as well as affiliation, are lower than the
correlations using only those items designated as
measuring g aff. This may be taken as an indication
for the efficacy of those items which were a priori
called §.aff items, since they correlate more highly
with the TAT measure than do all the items taken tog-
ether. Furthermore, as predicted, when the four items
which elicited affiliation imagery the fewest number
of times were dropped and the remaining items were
correlated with the TAT, still higher correlations

were obtained. Although these higher correlations

211
TABLE 1

PERSONIAN COEFFICIENTS OF CORRELATION
BETWEEN TAT TEST AND THREE FORMS OF
THE SENTENCE COMPLETION TEST

 

SAMPLES FORMS OF SENTENCE COMPLETION TEST

Aff items minus
All 20 n eff L1 lowest

 

 

All 60 Items items eliciting items
Sample 1
N = 50 ell 013 e20
Sample 2
N = 50 .17 .211 .211
Total Sample
N ‘ 100 e15 e19 023*

 

* Significant {.05

25

were not in any of the samples significantly higher,
they might be interpreted.as indicating a trend toward
increasing the correspondence between the two tests.

In light of the above results, more items were
dropped from the sentence completion test and further
correlations were computed with Sample 1. The six
lowest glaff imagery items were dropped and this test
was correlated with the TAT. The correlation of .21
is not significantly different from the previously re-
ported correlations for this sample. Then the 12 low-
est 2 eff items were dropped and the correlation was
.20. Further low-eliciting item elimination.was dis-
continued. In order to check on the possibility of a
reverse trend, the four items which elicited the greatest
number of affiliation responses were dropped and for
Sample 1 the correlation of this test with the TAT was
.08 which is low but does not differ significantly from
even the highest correlation fer Sample 1, (.20).

About halfway through the study, some of the aides
complained that this study was being conducted in order
to weed out those aides who were not emotionally capable
of working in an institutional setting with mental
patients. Even though only one aide refused to be

tested, it was decided to issue an official memorandum

26

to the effect that these studies were not job-related
and results would be held in confidence (See Appendix).
Circulation of this memorandum.is likely to have intro-
duced another variable which then makes those aides
tested before the memorandum and those tested after the
memorandum subject to different experimental conditions.
Inspection of the test results revealed that the date
of issuance of the memorandum could be used as a divi-
sion of the 50 subjects in Sample 1 and the 50 in Sample
2 as to the first 25 tested and the last 25 tested in
each sample.

Using the sentence completion test with the four
lowest g aff eliciting items eliminated, correlations
with the TAT were computed for the first and last 25
of Sample 1 and 2 and are presented in Table 2. The
correlation for the first 25 of Sample 1 was .57. Which
is significant at less than the .01 level. (For the
latter 25 of the first sample the correlation was .16
which is not highly significant. For Sample 2 the cor-
relation was an insignificant .09 for the earlier 25
but for the latter 25 was .112 which is significant at
less than the .01 level. Thus, in Sample 2 the trend

noted in Sample 1 as far as testing dates is concerned

27
TABLE II

PEARSONIAN COEFFICIENTS OF CORRELATION BETWEEN
THE SENTENCE COMPLETION TEST (MINUS THE A
LONEST n aff ELICITING ITEMS) AND THE TAT

FOR THE FIRST AND LAST 25 SUBJECTS IN
SAMPLE 1 AND IN SAMPLE 2

 

 

Sagle First 25 Last 25
Sample 1 .57* .16
Sample 2 e09 eli-Z‘u'

 

5 p <I.01

28

was reversed. However, in both cases the difference
of the correlations between earlier and later testing
was not significant at the .05 level.

Some further exploratory work was conducted on
the data of Sample 1. The 16 item.sentence completion
was correlated with cards A, S and 6 of the TAT, those
three cards of the original six which elicited g_aff
responses the greatest number of times. The correla-
tion was .25 and thus very similar to the previously
reported correlations in this study.

The final step was the computation of the multiple
correlation. The maximal multiple regression weights
were obtained from the scores of Sample 1. 7 The mul-
tiple correlation was .73, which was highly significant
at (.001. However, the multiple correlation may some-
times overestimate the relationship between variables
by maximizing the chance variance. In order to test
the stability of the beta weights obtained from.the
scores of Sample 1, these beta weights were converted
into B weights. The conversion into B weights allows
application of the beta weights obtained frogxsample

onto another, since beta.weights must assume a normal

7 Only 18 variables were used in the computation of the
multiple correlation. Items 5 and 39 were drOpped
from.the computation because they had no variance in
Sample 1; none of the subjects in Sample 1 received
any score on these two items.

 

29

distribution of scores and the B weights are scaled
to the specific scores of the sample upon which they
are to be tested as predictors.

When the B weights were applied to Sample 2, the
multiple correlation was .16, which was not significant
even at the .10 level. In other words, the very high
correlation which Sample 1 suggested was an overesti-
mation and was inflated by capitalizing on chance dev-
iations. It did not hold on the subsequent data.

30
DIS CUSSION

Review 9; Results

Although the correlation between the TAT and the
Sentence Completion Test (with the four lowest affil-
iation eliciting items dropped) for the full sample
of 100 was .23, which is significant at the .05 level,
the relationship between these two instruments is very
weak. All of the different correlations, using both
samples and different item and picture combinations,
yielded low positive correlations of much the same
magnitude, i.e., around .20. The multiple correlation
of Sample 1 was .73; when the B weights were applied
to Sample 2, the multiple correlation was .16. The
decrease in the multiple correlation may be evaluated
in terms of the small sanple size,8 large number of var-
iables, and most likely- an inherent and basic lack
of correspondence between the two tests. In fact, the
multiple correlation of .16 of Sample 2 is in accord

with the previously reported Pearsonian correlations.

 

3 This limitation of a small sample size attempting
a mltiple-predictive function was first pointed
out to the writer in a personal communication from
Dr. Paul Meehl.

31

These results should not be interpreted as measures
of the validity of the sentence completion test. The pre-
sent study has been concerned with measuring the corres-
pondence (and if possible, increasing it) between the TAT
and the Sentence Completion Test. In no case would show-
ing that the Sentence Completion Test yields similar re-
sults to the TAT attest to either one's validity.

However, we have obtained very dissimilar results
between both instruments in the present study which.must
be accounted for. We shall consider three theoretical
positions: 1) The concept of fantasy levels, 2) Mc-
Clelland's "relational fertility", and 3) a methodolo-
gical suggestion.

Levels 2; Fantasy

The suggestion has been made in the literature that
the low correlations obtained in Sentence Completion-TAT
studies may mean that different levels of the unconscious
are being tapped by each test. Specifically, it has been
argued that the TAT elicits material from.the subject which
is more part of his fantasy life and is less under conscious
control (1, pp. 25. 3h: 8, 10, pp. ash-257). McClelland
(l, p. 3h), in.reviswing French's Test of Insight, places
it on a comparable level with the TAT technique and says,

32

"Perhaps the reality component is larger in the sentence-
completion than in the picture-story or 'motive-insight'
methods".

Lindzey and Heinemann (8) summarize this view:

The relatively slight association between

the TAT and independent measures should not be

a matter of great surprise in view of the comp

plexities introduced by the different levels of

behavior measured by the various instruments...

Yet Rotter (12, p. 295) points out that the Sentence
Completion and the TAT have great similarities along the
fantasy-consciousness continuum. Neither one exerts any
great pressure for immediate association (as in a word
association test) and the responses are not completely
determined by the subject's own interpretation of the
standard stimulus but are both about equally dependent
upon what the subject is able and willing to write under
the standard test conditions.

Rohde (11) too implies that the sentence completion
technique is not very different from the 'more projective'
tests. She describes the sentence completion as,"...essen-
tially a.projection technique utilizing free association,"
(ll). Cromwell and Lundy (h) and Symonds (16) also per-
ceive the sentence completion test as a.projective instru-
ment.

To add to the uncertainty of McClelland's rationale,

33

it seems to us that the Test of Insight (mentioned above)
appears more like the sentence completion method than the
highly projective TAT.
McClelland's Relational Fertility

The studies which.have compared the results of the TAT
measure of motivation with.other instruments, ranging from
clinical judgments (l, p. 38) to a variety of psychological
tests (10, p. 27h): raise some very serious issues in.regard
to the entire question of the validity of psychological tests,
especially projective tests. As we have indicated before,
mere popularity or extensive usage may not be regarded as
scientific validation of usefulness or even importance. we
do not presently possess the definitive research findings
which.would enable us to understand with any clarity the basic
processes and characteristics of projective tests. The great
majority of our speculations have to be admitted to be that ~-
speculative.

McClelland (1, Ch. 1) has introduced the concept of "re-
lational fertility":

If one thinks of validity as meaning some-

thing measures ‘what it is supposed to measure',

it is incorrect to make such a decision in terms

of the correlation with some other measure which

is somehow more apt to be a 'true' measure than

the one you are 'validating', but it is correct

to assess the validity of the measure in terms of

the 'number and extent' of its connections to other
theoretically-related variables, 'among which.may

be' the 'truer' measure. Properly speaking, it

is 'the extent or richness of the network of asso-

ciations within which a construct occurs' that de—

termines its validity, not that it happens to cor-

relate with something which may be regarded as a

truer measure.

The TAT rationale does not itself appear to fulfill these
assumptions. The ”positive-but-low" correlations between the
TAT and other standardized measurements of motivation makes
the relational fertility of the TAT suspect.

The supportive evidence for the validity of the TAT is
cited in its relationship towards certain behavioral variables,
i.e., various tasks and experimentally-induced motivation drives
(1, p. 39; 9. P. hll), and its lack of relationshdp with other
"logically”-constructed devices to measure the very same psycho-
logical conditions are excused on this basis. Such a rationale
then assumes that test-behavioral relationships are final indi-
cators of validity. Although this has been historically consi-
dered as scientifically sound, it offers absolutely no ggplan-
gtiggLOf the inconsistency of different devices to produce
comparable results.

In the present study, we too have found nothing but a low-
positive relationship between the Sentence Completion and TAT.
These were the findings despite the fact that the Sentence Cam-
pletion Test appears to have face validity and employs a scoring
system.which.is directly derived from.the Atkinson-MCClelland
behavioral sequence.

‘He have considered the "fantasy-level" and "relational-

fertility" eXplanations and theories. Neither seems sufficient.

35

The literature does not contain any agreement as to the
"projectivity" of either test. Re-envoking speculations and
untested hypotheses (regardless of their popularity) about fan-
tasy and the unconscious cannot suffice as substitutions for the
understanding which.may directly be derived from explicit re-
searches of the fundamental tenets of projective tests. Only
when we have defined fantasy and its properties, and have evidence
of its effect upon the testing process, will we be able to account
for our results, or lack of them, as the case may be.

In regard to the present research, the use of the TAT as a
validator, or the explanation using the fantasy hypothesis, are
not in any sense answers to these more fundamental problems.
However, certain researches are suggested:

It would be most interesting to experimentally validate the
sentence completion measure of g_aff in the same manner as the
"validation" work on the TAT (2, 13). A study by Feshbach (1, Ch.
11) suggests that this is possible. He found that a sentence
completion test was sensitive to the arousal of aggression and
could be used to measure differences in this motivation. However,
once again we have a suggestion for research and not the establish-
ment of a definite relationShip. Another interesting possibility
*would be investigating the correspondence of the Sentence Comple-
tion motivation scores with other tests measuring motivation, e.g.,

Test of Insight (1, Ch. 16) and Graphic Expression (1, Ch. 17) in
order to arrive at some rough indication of the "relational fer-
tility" of the sentence completion.

In the broader sense, however, the main issue still remains,
viz., investigation and critical analysis of our present "beliefs"

about the nature of projective techniques.

36
A Problem 9; Methodology

The results of this study, and the suggested interpre-
tations, are of course dependent upon the soundness of the
experimental design. The particular design which was util-
ized in this study raises certain particular problems.

The technique of "validating" (which is really 221.1
demonstrating the correspondence of the two tests) one instru-
ment by administering it to a sample who also receives another
instrument immediately after, seems methodologically ques-
tionable. A design of this nature must of course make allow-
ance for alternate presentation of the two instruments to
each half of the sample, as we have done. However, does this
technique which eliminates a serial-order error, cancel out
all other presentation-effect errors?

Feshbach (1, Ch. 11) has shown that fantasy will reduce
the strength of a motive by means of symbolic behavior. Does
the subject who "fantasies" aff imagery, when he is taking
the TAT test for aff imagery, decrease the likelihood that he
will have subsequent aff imagery when he is taking the sentence
completion test? Does this effect of the drive-reducing func-
tion of fantasy behavior carry over when the sentence comple-
tion test is given first and the TAT second? This would seem
to be reduced to a question of reliability; for if we assume
that the two tests are entirely similar, then the method which
we have used, i.e., presenting one after the other in order to
measure their correspondence, would in effect be a test-retest
reliability check. Taken from this frame of reference, we may

consider the concept of "associative refractory phase," "...

once a given associative response has been made, there is less

37
likelihood that it will appear again immediately," (1,
p. 33). This has the effect of reducing the internal
consistency and test-retest reliability. McClelland (l,
p. 20) concludes that the practical effect of this inter-
pretation, as far as the projective measures of motivation
are concernedgithat at this point we cannot demand high
test-retest reliability from.our measures. In summary,
then, even if the sentence completion and the TAT were iden-
tical measures of n aff, we would not expect them.to cor-
relate very highly, since we are running a reliability
check (which already assumes validity) and not validating
either instrument. Once again, we must conclude that the
usefulness of‘g.aff as measured by the sentence completion
test can better be determined by independent experimental-

validation of the sentence completion.

chgg Experimental Difficulties

Inspection of the data sheets of TAT scores (see
Appendix), shows very clearly that Pictures 3, h and 5
elicit many more aff imagery scores than do Pictures 1,
2 and 3 which elicit very few aff scorable responses.
Is this set of six pictures appropriate for obtaining
meaningful data on ELaff? This set of pictures is re-
commended in Motives in Fantasy, Action, and Society (1),

but not particularly for n_aff measurement. Inspection

38

of the picture content and the scores which they

elicit would introduce doubts as to their appropriate-
ness. Although multi-motive scoring of pictures have
been used with success, (1, p. 36), Eron and Auld (l)
in their study of confinement aboard a submarine felt
that pictures should be utilized which are likely to
tap a single underlying motivation, and should evoke

a considerable number of stories with the appropriate
theme. In consideration of this criticism, we corre-
lated the sentence completion scores with the scores

on Pictures 3, h and 5. The correlation of .25 is very
much in line with the other reported correlations of
the present study and thus it would seem.that the first
three pictures were not responsible for the lack of
correlation between the two instruments.

We mentioned on p.15 that the attitude of some of
the aides in the sample may have had confounding effects
on the results. At one point, it was even deemed impor-
tant to issue a memorandum to counteract the rumor that
the tests were being used as a job evaluation device.
Although the'results reported on p.26 suggest that the
issuance of the memorandum had no appreciable effect on

n eff scores, it is quite likely that because such an

39

attitude did exist among some aides, perhaps the re-
sults of the study were affected by the feedback, i.e.,
discussions among aides with their co-workers who had
already been tested. French (1, Ch. 29) has reported
that "feeling" feedback can affect results of §_aff
measurement. The difficulty is that we cannot know
the exact effects of the feedback suspected.

Another set of variables which we must consider
are related more specifically to the particular subjects
who comprise our sample.

Approximately 50% of the sample were Negroes and
the socioeconomic class of the hospital aides was at
the lower end of the scale. Birney, (1, Ch. uh), in
his review of thematic content and its relation to the
cue characteristics of pictures,has reported the findings
of a study by Thompson which showed that changing the
TAT figures to Negroes resulted in different performances
by Negroes on this test. Rosen and Douvan (l, p. A93)
and Milstein (l, p. #93), have reported differences in
strength of motivation to achieve as related to differ-
ences in social class background. Singer (1h) found
differences in the patterns of stories given to the TAT
by subjects of middle and lower class socio-economic

status. Thus, we find our sample biased in the direction

no

of social class and race, differentiating it from.the
usual White college sophomore who finds his way into
the data. Furthermore, we may assume that the verbal
ability of the aides in our sample is much lower than
that of the typical undergraduate. Aronson (1, Ch. 17)
says:

Furthermore, although the imaginative

measure of n Achievement appears to be

a valid instrument, there is some evi-

dence which.indicates that the validity

of the measure is limited by its verbal

nature. It seems very possible that

many individuals who in fact have a high

degree of achievement nwtivation may score

low when tested because they experience

anxiety in relation to achievement and

failure, which.inhibits the expression of

achievement imagery.

Although this quotation is more directly applicable
to achievement imagery, there may also be concomitant
effects upon 2 aff scores. As for the verbal level, it
is also possible that a vocabulary deficiency affects
the two tests differentially. McClelland (10, p. 220)
speculates that the TAT measure is applicable over a
wide range of verbal ability Since a category can only
be scored once per story, and because a significant

relationship has not been found between number of words

in imaginative stories and n achievement scores. Our

hl

question remains: How does the verbal factor affect
performance on the sentence completion?

All of these considerations, 1.6., the effects
of race, socio-economic status, verbal ability are
only hypothesized rationales for the results of this
study. We do not have precise measures of these var-
iables, such as I.Q. scores, vocabulary scores, etc.
However, these considerations still seem intuitively
pertinent and one must reserve.final conclusions until
they have been measured and until research findings
have explored more fully the effect of these variables
on the various tasks.

In conclusion, the present study has failed to
demonstrate with any certainty a relationship between
the TAT measure of g’aff and the present Sentence
Completion test of n_aff for the sample which.we have

described.

uZ

SUMMARY

The present investigation was concerned with the
construction of a sentence completion test to measure
the need for affiliation. The purpose of the new
instrument was to avoid the content analysis required
by the TAT method of measuring affiliation which is
used by McClelland et al. It was expected that the
obviation of content analysis would result in a more
economical, self-administering, independent research
measure of affiliation.

The two tests were administered to a sample of
100 neuro-psychiatric aides at Battle Creek Veteran's
Administration Hospital. The correlation between the
two tests was computed; items of the sentence comple-
tion test which elicited the least amount of affiliation
imagery were dropped from the test and this revised form
of the sentence completion was correlated with the TAT.
Finally, multiple correlational methods were used to
predict the mazimal weighting of all items on the sen-
tence completion in order to increase the correspondence

of this test with the TAT.

#3

The final form of the Sentence Completion Test
correlated .23 with the TAT, which is significant at
the .05 level. However, the overall results of the
study suggest a positive, but low relationship between
the two instruments. The multiple correlation weights
determined from the first half of the sample did.not
hold up as predictors of the performance of the second
half of the sample. The lack of stability of the weights
were viewed in terms of a) the small sample size, b) the
large number of variables, and c) a basic and inherent
difference between the two instruments.

A theoretical discussion of projective tests sug-
gested possible explanations of the results. The theory
of the drive reducing effect of fantasy behavior involved
in taking a test and.the experimental concept of asso-
ciative refractory phase were considered. It was proposed
that the method of "validating” or measuring correspondence
between two tests by administering them one after the
other is not a sound.psychological procedure. Limitations

of the present sample were considered.

2.

3.

u.

S.

7.

9.

BIBLIOGRAPHY

Atkinson, J.W. (ed.) Motives ig_Fantas , Action,
and Society. B. Van Hostrand Co. Inc.,
NeJe, C

Atkinson, J.W., R.W. Heyns, and J. Veroff. The
effect of experimental arousal of the
affiliation motive on thematic appercep-
tion. J. of Abn. gnqugg, Psych., 195h,

£2. h0535157'__’

Atkinson, J.W. and N.R. Reitmmmt Performance as
a function of motive strength and expec-
tancy of oal attainment. g;_g£ 522! 22g
mo: 195 a 5}.) 361’3660

Cromwell, R.L. and R.M. Lundy. Productivity of
clinical hypotheses on a sentence co le-
tion test. J. Consulting Psych., 195 ,

m. ual-ueu.“

deCharms, R. A self-scored projective measure of
achievement and affiliation motivation.
J. of Consulting Psychl, 1958, gg, p. 172.

Eron, L.D. and F. Auld,Jr. A study of the thema-
tic apperception test stories and sentence
completions of subjects in Operation Hide-
out. Medical Research Lab. Report No. 2M3,
Vol. XIII, No.'h,theb, I95h.

Guilford, J.P. Fundamental Statistics in Ps cho-
logy and Education. McCraw-HIIITN.Y.:I956.

 

Lindzey, G. and S.H. Heinemann. Thematic Appar-
ception Test and group administration. 1;

Personality, 19559 ELL. PP. 3’4'550

McClelland, D.C. "Measuring Motivation in Phantasy:
The Achievement Motive." Studies in Moti-
vation. Appleton-Century-Crofts, Inc. N.Y.,

1955. Chapter 37.

10.

11.

12.

13.

1’40

15.

16.

1L5

McClelland, D.C., J.W. Atkinson, R.A. Clark and
E.L. Lowell. The Achievement Motive.
Appleton-Century-Crofts. N.Y., 1953.

Rohde, A.R. Explorations in personality by the
sentence completion method. g;.p£ gpplied
P826119, 1914.6, 19, 169-1810

Rotter, Julian B. "Word Association and Sentence
Completion Methods." In: Anderson, H.H. and

G.L. Anderson. §p_lntroduction pp_Projective
Techniques. Prentice-Hall, Inc. N.Y., 9 1.
Chapter 9.

 

Shipley, T.E. Jr. and J. Veroff. A projective
measure of need for affiliation. g&_p£ Exp.

Psych., 1952, Q3, 3h9-356.

Singer, J.L. Projected Familial Attitudes as a
function of socio-economic status and ps -
chopathology. J; Consultipg Psych., 195E,
lg: gg'lou-e

Solomon, R.L. and Howes, D.H. Word frequency,
personal values, and visual duration thres-

 

Symonds, P.M. The Sentence Completion Test as a
projective technique. J. of Abner. and Soc.

P820110, 191...], E3 320“32—9_0

APPENDIX

 

 

 

 

 

 

 

 

 

 

>V ‘ ‘ Ila,
.
I . . .
. _ ‘ r .
’ ‘1: 7' ~ 5.1.
‘4 >I } a- _ V _‘
.
. A
2 . . ‘. ’l\
‘ ‘ z ‘4‘”
a... I. - -.__ 1.. K
‘ 'u'&-J a- r ‘ ‘
. . ‘ .
M ‘ ‘ A. ‘
" ‘ "v‘ are... .7 ‘
.“ V ’ , - V
. r A K
A
.1 ‘ SA )
V '1 r r. I
l
,
.
y .1 I . V‘
— ‘ ( I. _A
'1

 

 

 

 

 

° “uh-Hm

._., , 4. ‘
c ,. , .
k' Ah .5 .. _x..
w. .
|
‘ ~.
7 A‘ "‘ r ..
-—~..r.., -, _
j.
u o ,
" v.
V r—fi-
..-o V.-. ., ~.. r
‘7 v “I
' I
-7“- 2. .
~ 7 " ..Vu-k.‘—_
‘
‘- ‘U -. ._
' "‘-'.. . 7’

he

 

‘ ' v— .
‘3..o-,‘_" n,
YL'...’-_
’ uh..-

 

   

 

   

     

   

     

 

 

 

 

   

   

. - - , .- -- . - ,
.- - ... . - - ~ . . . . .
1 A _ -- .- \ _. J 1. -. - m... . . .-... . ,
s. - .
l ‘ . ‘ t “ , ‘ . 3
f. _. ‘ -. u t .
‘1 . . ..
- . - . - . . r- - 1 . .
. w ' ‘ 9 a. .
Tu .. r I I - t
I . . -. ..fl - _- \ J .7‘ ~
. ..
.x_..:- e... .- .~. »....-. ...r. -.-r .7 . .. .. .g - ...-
l
,p‘,“ .. . _, . - 5...- .. .—,-: .. i,- » .._.. 1h. .. “v... .»
'I"p~‘lp »- '"v -:-.q -,_.,_.,-.._. . v ‘_. ,‘
. ,v . ..h > H x. v . a v -< 4 . .._.....—.n- are .7: .
.-‘ - A‘ ‘ , 4.. 9‘ '4 . - ,
.'- - - '17 A ..“-"..
. . . .‘7 1 I . .. A -
"". ‘ a “u
. -. .. _ , . .
) " ' , - f“ x . f .L? I
. . -
.. . f . .. _ -
.«i_ A.) _.
In-- 'A‘CA‘---¢. mm-~o-..ew.n. ‘-A. ..._ 4. .- p .1
- - .a—e. .. ,._— -_-_ . . _..,_-_____-.«.e-.—_.-. --
.1- ~ .. - .uukr- .‘_~—. ..- ~75- I.- .a .. . ‘ ‘ . a ' , -.
._. ____, .1,.....-.___......--... 1.... -.ruL-.. :..., w...——._._....—..n--—-o.w.
, . ‘ . d . ‘l; ‘ A
f i ‘ 1i,< 1 .
... . .. ... . . - .
,. . - .
' ‘ 1 I - e
_ , _.. ,‘ A . f. .. . . ,. . ”A;
’ A A c ‘ .7 ‘ ‘ 0(3 ‘ ’f l' [71‘ ' ‘
>-.. ..a..- ~ .) xfiv'u ., . I..- —~ ~
. . .L
It _a..’ r--. A I'An ‘ «O‘- -- 7' .' 5 7 —.~-' In Ari--
... .. Ax— . . .. .n v.1-..._.w_-_—.-_.- ...i.. .r..,... ..._.. --.-.r.--.. -
I u e I. In. - I e ‘ ‘ ‘1‘. Mn. .‘ .~ I ,v- v a»- A - - 1 -. s
.71. _, -v,, ..--_, .—~ V . ..— .uem-"F... . —.a . ...v....:_ v- . ..r
-.' ' , _ ' . "
. 4 ~
. . ,:| .o. J
J ...l .. s E L- e- I
, ‘ A » 1 ., .
. A . - . , , .-.
.. _‘) . I \' .... ..~*f. !. 3:1.
, . _ . .r .11 . . - . L . ... .1‘
s . 'd
-, -...__...-_...,.. --- ..- - —-. ..-. .... . «(..-A..-
_. . 12.2. J,.,__ n , , . ;.w .-1._......... -.,~ 1 . . A“... . -e ~WA. IJ.;.,-,
. . II c-.. ”'7". rn ‘ _A. . -‘r . g... ,.. ...-. -
A, . ,. - m .- -. x. -- .. -—,.‘--. -.._. . .- ..__..._-_..... .-
V ‘ . I
\. . v , . . .
n ‘ ' 1 \
7..t._. _‘ )v
f. u; l
x‘ ‘ r -<‘ - .1 .. ~
' 4 ’ ' I
r I ‘. \ ,' . I c . g l '- 1 A
.774- "
- .. . _ l .y . .~ . .- c
.--.. -. ... --....- --...r-..-cu . -
4~.v-~‘--..<.....x.\v.. ;._ -. , .._.'
war J u \.' ru-y ...r. .4 .-. --l n ‘-

 

V~ .r ..-.3 gm vw. v.1. w- -. *--».-o«-..-.. “4‘ A

T19 teacher I liked best was one

M110

 

 

a)- She meant more to him because

 

 

’- f

_:x:. Hhen.his father decides for him

he

27' If I get control I

 

~33. I apply myself when

 

 

His greatest ambition is

 

30.

The quality he likes best in a

person

_fl310 I can3t get along without

 

 

' 32 I feel strongest when

 

 

__330 If he gains power he

...3

...3

II
I-.. o

1'“

De

 

 

Henry tries hard when

 

 

On a difficult task I

 

 

 

   

 

 

T ) '
.
,
‘ I
. J ..
" ' 4 . . L
' . a ’
a ' s
4 . o
. -
.4 -
. ..
. ‘ A I.
. . . .-
e - ' ' ‘
‘J H
u a7 I - " ' 7 ‘ '
n n,_ . ...‘3-L
..
. ..1 j .-
. i I ! -
n ' .y‘
. o
‘ l1 ‘1 u
v . .- .-.-
_ o ’. ._ ' . h. - . . .
. -:
' ,
y 9
, . . a
K\ .'. 5'1 . '. ,
I -'
‘ . ..
.. ’- ‘9 - ‘ “" - '
. . .. . ,. _
IIL.\. . - ‘l
\ e ' 1
. J . “ n
- « ) . A»
A\
[.. r‘ ' ','
, , . I -'~
.
.A. _‘ - _~ ., ., . T:A ~. I. 1r ' .
‘e -.,,“a"'.“:
. --. ~_ . - v . .vzt . .. '1
h ' ,P‘s.‘ ..k‘
- -- * ' ~ ‘wo
. .. . ,. .
7.. . ‘ . . 1, . ‘ ...
,.
.-
. . - I v
.4-
.I
..-v.
, .,,,... .._‘. ... .‘J; 7 ~
.1 C
..t I _. - 9.: ' ~ | — ...1:.-.. ‘e ’ ~

 

 

..
.\ ,
¢ ‘ '
, . ...
b ‘ m
‘ . .
. ‘1 ..
' ~’
. .

. a
‘ .. ....
u v ‘ '
I o

.5 ‘

 

 

 

 

. . 1 n
‘ ,
.
. . - _. . .. T w .
. — . , - _ .. . n A " .- . . -1
.4
- - Au
. . . . L - . I! n '- l ‘7
I] II , I A \‘ I. ‘1 .39.‘ '5 l I— I v . .
.s . *1 ‘ . > -. ‘ e ,
., . > . . a, . x.}
)< b '7 - I. V l 15*5' l-‘. .u b- . ,- . ‘I
‘ - I' AC. V Y , - - A ~ ~.~-.‘I“_’-’~p I ..
.
. i . . . I
. . J . _ a _k ‘
- ~
...- ,. V ,_ . s . n .- ..r _. a» x n .
.- n , -. u " 'Uis— a... .~~¢ n. ,_ -5 - _.
l r .
.4“ v.’ x 7. .h. u. . _.. . a... . v.._ _‘
1. -.. ’v .4 . - ~ ~ -v§- d '. '~ - ’ '0.
L .
' _ I
. .4 -
. —--.‘i - .1. -r.-~u. v.ve ...g ._. .. .
, -.o -_ a...“ “.4“, ...n l.-

 

 

 

 

‘ ‘.- | I h .v I - r I
I, 16" ..i . ‘1 1,". l 1. ‘3» ‘ ' LA; I
‘a ,- w 5‘ . 1 ‘
“- 1-)" 21" (.'(1;'1‘..".(.‘4 t’zliz'J'
'.1 1 . ‘ T . ‘ - r . I
I“) ‘71 ‘r 3 é; I .l -J “V ‘!~?|‘.r\ytty
r. -:.‘ gnaw»... .- . .— -A§§h~nb|.—~r~l—--‘IO§I-fl-L‘iliu-kul Hue-1....IA a...» A-h-t-—A. ". .....
“1‘=-A“:-ox:‘r._r.Jx.A-ia..;‘uLana.- a. Haw.u.au 1.13;”m “gamut-v" °
. .--r su..-.._.5 i..n._~_-._v,-~..... -d-_-‘~mmr~ uw.‘ tau-word” ~- xiv-r-Adanvux
1pm.: w '3‘» m|n~emnnz Axu-dw-uv zeta-e.“ nu can.“ gnaw-5e... ..3'7‘) .‘u- ...- .
, l ' ' ' ‘ l I v 'p ' ‘3. .- ‘ ‘
.l .henmr. l sax-5.». Lo my bass 1
pr" v‘." r» ‘ . _ ‘_ "’ __._,' . . .
H”- Myer he. Has new: Teal he
”am Q
-51. HE‘S—“‘3‘ -_‘-I" m* 1‘7““ —r.— H*-W-‘~M ‘w—W.‘ O. J- -' -’-.
‘Ill» Inn‘h'l’ flax-E’x—ulc MA._ “‘Wu-eeAcaL-u._;_sdmnl.g_fl -nu’u. a... . ._
g. .- .-¢~=«m .w-.- m ...wm- Mad-n» ..~- . ...i-
uu; “‘n"’.’_l.-‘1'AZ‘ ..- nL- -4_—« ..- u‘. .-x r“ ‘IJ‘- mun-slum» Iain" n1: 4
v. (‘ ,..,. - U, .A 1" ‘i‘,.(."
1 as me: .01: EA),- .. :30 mm Damn-pa
172:: U 15:19 ---.1,‘:ri?.7‘=(;‘llt «more 1n his two"
ins-K .4
‘v l‘"“v§lr“ -1‘--“nxa"u~ - — ‘- uses—ult-
113
.mu - -t-".
.»r—»;:' ‘1 n4. -- harmtmcm- ..ut-lh tub-r-Jznv. r—rxu- -..«n- ‘5' ac. Mr.
—x — —-— "J“'_I’_’ .‘ v n .4-
” f’ '7 . v‘) ~ 's:-" (3 N I) .‘m 5 ~ -1 ~ 'E\ -.
,j: I l'. ’ 1' .4! AL‘C: r,(‘«'.’...1'¢f No.5 Rule ‘1)? 9-‘
‘ v - ~ - b-.._e\-w- .v' m », A. — P .-
:C Li My ..f en’s? 01‘ 533 we: 1.1’1' 9’1?! ‘-
.31.-
",u’a'r;u i'-lolI('-1, ~‘— "Barnum-1.2..
T
d. .. w‘ '. nanr..=--
w"4:-N-?'.aA—-aq..1er .. .- ..-... -—— — x-v—h-l.
—-.— M‘.‘ v _' f, I ~u-cx».
T ' ‘- ' J ' “1'. . s, ' t" .
Lima 3.1-5 mot-Ter 1.1.1.155»: as
:20 met .1. vent in a 215-190 is
...y....
i.- welt—1.211.?) I. ‘vnn..;-nur;r—.-4—.’- __, V V '1 C(”Jeun-‘-
1H.“ u- w- — —g.»g...,_
- nerd-- 31!” m' Ain't-M ans—u ‘ -= W. ‘ V‘fi '-i“'--"-‘l'=-“-‘wvl-
“maxim—IV: ' ““ n v. ‘1' --' 5 my .
A- 1 '1 V); i y, ,.__s ., . .x h, ‘ A“ .J
_. / If», A. :13 171.] r;.(..- "Nut
.1 I. - . A ve- ‘3 3. uz‘ e -\ a w -. --F‘ e. .
b--- 1w m3.“ 1. at! ~10 10 an» n n
In... 7".
nu -.:.ma BAT-‘4 rm— ..- Ann-s-Hlu‘d-Nw 4‘...er .u- m- mun—...- “-mu---;Lup-—
m . rum--..fl jun-a... ta‘w wan-.-mi ,
~ «saJ-rc Inca-Ax .- “—L—mn- w uw
~‘IJJ'w h— .5 V ‘W fi . l
7 ”f '
-.~ J '.2
~21 ’ rm 1. ~ ,, N .~' on. ‘7 r, '1“ q a» , .5 1», «as = r‘ .‘ ’*~~'+r"‘~fi . A...“ -'. we — v
-... 1 . it" 34!-.. 1.73.370 .1.-L]...L.1-1;Ck._‘ (‘3 k. \ chug. (humid, ..L-!_ Levi} d‘iileqiitb-‘A 9-111)..- 9.; Lil r-" )2. ‘.
. . - ' 47'
”v, .v)’ .8 '..1.
r--' : .. .. ' ,. 4... A . -:‘.-- .. ,6. _, ‘°. .1 1‘ ‘ ... .-. a “ ‘
-s L- a; 2,311 as UK; the way you. .4. 1:481. am- as er er ~-:-.~u,..<.,.. .1 eel. are ac ’t, , -o a mus . )
.1" 3 1 . .. ,-
L (A. '. ()1, WISE. encompa-
“ ‘ . v,.. , -‘ ‘V .8. .-. 1‘. -v.. .- .—° . . 6“ ° .. . . ..
If The rtERTffll-Jau does not 1 afar m7: 3 .lu. ya... a manna { ' . in f mm Is: the 301-. 22:? .cv:
aun»-r— ‘
’1 . . .1 .1. . r... . <1. .. .‘r‘?,..‘,,.‘
Ln“. 0' a * (7:11.31 .ss.n,.>nz*e (11.1.3.2 Ely .. f- :1 1.6.7338 n42 monks,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/F*L ‘77 M _. _ T” In S[/77 //

One tling we want to fild out is what people think of situations
GEPH may some up in life. I‘m going to Show you some pictures of these

:Ltsaeions and ask you to thins of son; ies to go with them. ”he situations
7')

rI—J

 

 

J -

u
I
.J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ y '1‘ , . .
won‘t be 010 r17 one Chi“: or annenern—so fee] Tree to think of any
9 ~.
. ,_ r3! ”'rx ~t w. *1‘ a.) ‘ A ‘ I
Story you want to. (QLUJ AILIUAJ ) LOP enamels, here 3 the
.':'- -A ."_ .,.q , 1 1 r 1 -' ‘ t ' ' v ' 4L .
IIySL picture. I‘d iike yoA ‘o sgcnd 3 fps mc:manS thinking of a story
a- v " _I_1’ ‘0 ‘l_ r11 . - _ .3 - . y 1. - ‘_ ‘ p “A. y" ‘ r . '1 c 1"
To go Ties It. 1‘ get 9L tn- Seocj you-Te than.1ag 01 I ll 33x you
Q a ~v'V N ‘ - ’ fi ‘.' ‘ L. ‘ -. .a~.
awzcsrtiorua liens: .310 ance *fllCSK).?3CHlie£ ‘-.: L, 00 inccy'xzcix :' arti no (Hi.
I _ ' - ._ . ..A _- '— n-A‘ ‘ ~‘- \ 0 ‘-\ ’1' m‘ 1
Just answer with schhi«E Lm'g comes L LuL. The 6 S10 no right or
mecca answers.
Q h ‘1 . . - -
L min are anso people? ”hit {.2 ans" vim"?
~ _. W - -' I -' "' "'- ‘u—nd
I
.7 :g—g ,. - ‘1 . _~_-: .v Y , :~ '~ - ‘ - >'
:8. what has lee up to nism~wnae pace on Defore?
__ .. A— -~ ‘h ‘—'¢
- 7 1: ._‘ I- J- 7..- 1 . _ -. 1 "V. -. ~- g
in. Jhac do they acute-nos co ‘hey feel?
'1 ‘7'“; ,1. ,, .9 ‘1 1, . .- . a .. ".9 ' v . , (‘1
Ac. huab Jili neoncn~~nom will It ends
--

 

 

 

. ‘1 “
. . - r
. . v '~
’ ’. f .‘ ‘- J " V '3 f' -‘ ' ‘ .. ‘
' .‘ J . . ,«9 ~. 1. L ‘\ 1.;
t -..H. .. 7k»,-,._ ...h .._. .I.., - ,.
- __ 4, . .-._ A- -__- .w -. , V V .i. ,_. ... . .. - .
”I.-- _- ..r. -_ u-.. ..._- -FHI- .-. . .-. _ ”hm- ,- . .-. .-._ . __ --
'v , ' ‘u . . qu ‘~ .'\ i- \ '~.’o‘ ,‘ .-‘ I '1 . r
‘_ _ -1 \ _ , . ‘ : - M ~‘t1 ‘ ‘ Lul "_ w
. . .. ---“a. .. .._- . -.I , ..
, _ A I .--,_I V , _ _ ..... _,,,, . .
| f -. v . ,
‘ ‘ _ ‘ ‘- -.». w. ~. , ' j .- . .,
. ‘ . I . I H a
L ., .i .J ,(‘_” a.\A,i J‘. . _.
..-.. - .. . . - w- ,- i . A » \ .. -

 

 

.v V ’v
O .
‘ '. 3 .
.1 . v
u .

v

n

_. 4
v,

1 ’\
u I > ‘
" u ‘L.‘..
. I"
v, --- ‘ 5 I
-.i < . I .
-_.\_ _ 7 7 _ ._
i. i . H . - . ...---..‘
, 5- _. ' ,7 . _ ‘
A ‘ . n; l ‘i -. ‘
- V--- ,7 ..<\ ...v . - ...ia
.. .--w ._._ -.. ....-
— -..- .u .. .—-. .V... -. ._
w...,. ,-_.- ..-.._~.... ...i
.. . -- . - _ A i -

—---

 

v
‘v
A

 

 

5

 

 

_ ‘
{
g A _ q -- ” ‘ ' ' ' -E\
_ (
Y ' ‘ . .
I ‘ I ' .4,
x
I. > i“ .I I ‘
J "-
, ) _ _ h. I _ ‘ l A l .
4 9 ‘ H 5 ' K.
‘ , V ‘ n ' , l V o A.
, . ‘ t ‘7' 1 . ‘ L7 ‘ h .
‘ ‘ "
..‘ ‘ y I I
A . ~ A. Nu ‘ w " 'T ‘ ~ A
‘,olw* 7 _
.
. ' _ ' -‘ I 7
' -, .._ , .
| g ‘H _

 

 

     

 

 

, 7..-.- 777_7.7-.7... .77”...
..
'7
. I . v, . 7. WP ’ '
t ‘ 7
I . «.1 I. I i ,J :7:
A
_ 7 ..7 7....
» 1 ‘ .. - ° 5‘
' . . . : ‘7‘ . r‘ > ,’ " " ' , 7 ' A ' V 0 y
_' , .' .7" -\J g . x 4 2 . J ’ 7 - 7 '
o -. . . _ -. . _ . _ 7 7 .
, 7 7 . 7 77 7- 7 7—7-7 7 I 7 7 7‘ -..7 7 7 - . -7 . 7 7 -7 . 7 7 7 7. 7- .7.
7 . i .7 77 -._ 7 77 7 . 7- 7777 7. 77- 7 7 - 77 7. 7- _ .7 7 7 . 7 7..
- . 7 7 7 7 . 7.77.7 7-7.7- 7 -77777, . .7. 7 ..7. 7 7 7. 7 7 7. 7 7 7. 7 7 7 . .. .7 7 -. 77 7 7 ..77 . .77;
' 7 1‘ 7 »- 7 7' 7 7 .7 ’ 7 7. ‘1 '
. . - f 7- . . . \ , I . ' 7... . '- ‘ '. " '
. ‘ _I _:_ ‘14 7 I\ _ .. _ - 7 ‘ 7' . \‘ . *7 t - '
A . 7 77.- 7 7 77.77- 7.
_, _,..__. -77747 __ . 777 7 77-7 .- 7.-. .7... 7. - .7 7 . 7. 7. 7-7-... .,, .7 .—7. -7 7 . 7. 7....77-7 .7.
7 7 . . 7 7 77- 7 ...77. _. 7. .. . 7 77 7. 7.- 77. .7 .-7 7 .7. 7 . . 7... . . 7 . . . I 7 . 7. 7 .. 7 7.7 .7 .7 7 7 ._- 77-7
7 7 7.7.7 7 7.7 7. ‘7. 7.77.77.77-- 7.777777 - 7...-.77 7. 7 7 7- - .77...- _ 77. .77 7 .. . 7 7 ,. 7i 7 7 7 7 77. 777.7 7. 7.7—... 77-777. .7 .7777. 7 . ,
k ‘7 u 7 - 7.7, ., . ‘ . .' ' ‘.
_L..;:! (J ‘1'?! I ‘,7 7,. 'k‘. . K . .
\7 x x A 7" . a. « ~ ~. A‘ 4 7 7
~ .7 ‘77- -7- 7 77-7.-- 7.- .7-- — 7 .77 . 7.
, -7. .7 7. - 7 ...77 7 77 _-...7. 77.. ._,_.-....-..__.._--7- ...7 7-77... 7 7.. 7— .7 -777 777 7 77. 77 7 7 7 .7 . 77 7--. ‘7 — 7 7.7 . 7 .7 7 -- 77 777 7 -
_ . 7 7- 7 7 7777 7 7.- . .77.... 77—77777 7 77.7 77.-7_77 77.777.777.7- 7 .7... _, .777. __77__ _ 7 _ .77_ .7 --.7 --.77 . 77 .7 . _..- _ . -7 ...7. .- . 77. 7 7.... .
> .. 7 . , . 7 7 . . 1
. . - I t. 7 . 7 . ' ' ‘ . 7
‘ v .’\ - ‘ ‘ ,. _‘ ~ ! 7‘ _ ‘ 7 'If Vt _ I + . . \ ,.
. ' u | ._ , 4‘ ‘7 J. . . ‘7 '. . '_ , , ,1 v 7 .
' 7 - --7-7, 77 7,.,.,, .--.—_777r.. 7 ,7
7 7 7 7 _ . -774. _ ._ -._ __ _ 7 77 7 7... .. 7 ._ . - 7 .7 7 . V , ,. . 7 7 7. . , 7 7 ,7 7 7 7 7
. 7 v7 7- .-—_~-_.. .7 7. 7 . . 777777. .7 7 7.- _ ._77-.7.7 . 7 7... .77 7 -77 7 . .7 _. 7 77-777. - .7 7 77 7 7 7 7 77 7

;. 7 ..7 fly... m.—.—.~77oow—

/ /

4... ““7— -.—.-v- .-

 

 

‘ .

E)
’7‘?
L4
3)
('2
57773
(7)
C’)
0
”j
(3'
-’. )
{7.7!
(D
x‘)
("f
'“ti
(D

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W‘Ww-crflu‘u 0...A‘. v
7.7
’J V —'— -. --.
7,7 .7 .77 7.... .7 ...— ._7 7
_ .- ....7 "77......“ .77 7_-77...77 7......7. . 77 7 7-7... 7 7 7-.....777777. 7.7. ........,._......7.. 74,...
7.7—. 777.77..- 77.7.77 .7..-...7......77.7..7..77_7__._.7 .7.777-..7.7.._._7.._..7. -777-7.777.7.77-.. 777.777.7777--. .7 77.
.y‘ u ‘ - 1 'n r n A I‘ ,_ ‘_ A q." - I?
a 7« w- firs I 7.77 rr~-~ 77 77 7, ”<\;‘ n 1:; .Ono
.7. . It! _~,'.'._ ,J _, g_\,v ~; _7- b_;,,_i, 7 J 47 .,_k- m _. 7.7 77v.l7.- .I . ~77. .77 77 I-
L w W —_-.~n-
7_L 7- .77. ...7 _7 7 -7... 7.... 7.777. 7.77.. 7.7. -7777 7.. 7.77.7777. -.7. ...7 .7. . 7..
-‘ “v“ “‘(w‘7‘-n_.—-—W‘s--~m—p— ._-—.——nom-—--—-- —« w" 7—-nv~ Ffl-J A—hv‘ b-Q .-.‘DU. 4
.-..—..-.,~ . —...77. .-.7. -7- ...77 7...... 77-7.7.7”.“77—7—777--- 7.7 77.777.77—7..7— 77 777.77.--. 7...- .- .7... Y. ,
7 ...__.._. - 77 .7777.-.
- -v‘~v . ‘ ,1 [s -v n
7 ‘ ._ , 7 7 7,7. . 7‘ ‘
h “up ...7 r~ 7n. 7 ..\7\ r
7.7 7 ..7‘.;7_. ._,r 77L\ :7. ~72 - 7 7... v 0
V M-“-v 747-..— w‘
7 7777.77.77 .77 77.. .~ -.. . ‘w
. - -0“..- -
.7 r 7.77-- 77.... 7777-...-.7777777..-_..7.7 - .77 7 7 7.7777 7.... .7... 777...... r———----- m"
~~g7 -* ..— fimbvww‘mn o-I ‘-t-‘."‘"m.-Wmfl’ '7. “7.....wwnu. -"Hh-‘A‘d~‘--w 7 pal-II- w- . m
— 7...... . 777...-.7.-.7.__77.7..7....7.-7_ 7777.77 ..7777 77.... 7777.77.77... 77...... 77-77.... I _
MW '. 3 "Y ‘ . - 0 “1 g “u
“ '“WL Vh7« ‘““”‘Gfi ~- ’cvr-n r Haw»
. . . . 7 . . . ~ .
V; -v...( 7.-..7.-- J7-t-.v7 ..z 7 7.7 v7. 1 .-J--7 77. L: 77:. \777
1' 7' Ha!“ -- _ 7——
_.- 7 7-7 777..~..7--.7..7 7.7.7-7777..- .77.... 7. 7 7.7- 77.7. 7.7.7-7777..._.._7._.77.77.7._77..77..7..,. -.-777.-....77.....7 M7-.. -..7..-
.77 4 N.‘ u h " -‘vfl--M-V"~_ -‘-..h"~‘-O -a'u—‘~- - --—4 7. 4-- -sn—“w-‘A—u_ .—i‘nw1‘..M—~v—‘fl—— ...O-'s-'A'7~" :- ¢'--- .—
,,7._ 77.7777... 7 77.77.. .7... 7- n,“

 

- 774-»7».r...-.7

 

 

 

 

 

hA.

MJGJ“

V“..

 

umuwuumunmm"mumummuuum1mm

 

3103815621