PROPOSAL FOR PERSONAL SERVICES CONTRACT A PRELIMINARY STUDY OF SELECTION OF PERSONNEL FOR FOREIGN LANGUAGE STUDY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP61-00017A000100050002-5
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 4
• Document Creation Date: December 9, 2016
• Document Release Date: March 19, 2001
• Sequence Number: 2
• Publication Date: August 10, 1953
• Content Type: STUDY

.COPY
Approved For Releass 2001/04/04: CIA-RDP61-000174000100050002-5
PROPOSAL FOR PERSONAL SERVICES CONTRACT
A PRELIMINARY STUDY OF SELECTION OF PERSONNEL
FOR FOREIGN LANGUAGE STUDY
August 10, 19 3
I. 'OBJECTIVES
The background for the proposed study is the fact that relatively
large numbers of personnel in military and governmental agencies are
currently being selected and sent for training in certain modern foreign
languages of strategic importance. Experience with these programs has
indicated that many otherwise promising trainees fail to learn a foreign
language as well and as rapidly as is desirable. In view of the length
and expense of the training programs, the consequent wastage of effort
is of serious proportions.
One factor in the failure of many trainees in learning a foreign
language'is a lack of certain abilities and aptitudes for this kind of
training. By seeking to measure these abilities and aptitudes, it is
believed that it would be possible to develop procedures for selecting
trainees which would markedly reduce the number of failures. This is
the chief objective of the research proposed herein.
It might be thought that suitable procedures for selecting language
trainees would depend chiefly upon tests of intelligence. Previous re-
search with the foreign language aptitude problem, however, has indica-
ted that this is not the case. To be sure, intelligence tests have
proven their worth in predicting success in certain types of language
courses, namely, those of the traditional sort which emphasize only
the acquisition of a reading knowledge of a language. The types of
skills emphasized in current goverment-sponsored language courses
are, however, those having to do with the speaking and the understanding
of the spoken language. The evidence points to the conclusion that
intelligence tests of the usual variety fall short of desired standards
of validity in predicting success with these aspects of language
learning. It should also be remarked that trainees in governmental
and military language courses are in general already highly selected
with respect to intelligence.
In the last few years, research on various sorts of verbal and
linguistic tests has revealed the existence of a number of traits of
individual differences which are largely independent of general in-
telligence. These tests are regarded as promising prognosticators of
foreign language learning success, particularly of success in oral-
aural skills, but they have never been tried out in this connection..
The focus of the present study will be to determine the validity of
such tests in measuring foreign language aptitude. If some of these
prove to be good predictors, they can provide the basis for an improved
selection program.
Approved For Release 2001/04/04: CIA-RDP61-00017A000100050002-5
 Approved For Release 2001/04/04: CIA-RDP61-000174000100050002-5
(2)
II. PROPOSED RESEARCH PROCEDURES
In accordance with the usual plan for validity studies, the
following steps will be taken:
1. Assembly of a battery of promising foreign language aptitude
tests, including certain already constructed tests as well as certain
tests of novel design.
2? Administration of these tests to various groups of trainees
in goverment-sponsored and/or university language courses, preferably
at the start of training.
3. Collection of criterion data relating to the success of
these trainees in learning various foreign language skills, including
oral fluency and aural comprehension.
it. Statistical analysis of the relations between predictors and
criterion measurements to determine predictive validity of the tests
and to establish a suitable selection procedure.
The tests to be investigated fall into a number of classes:
1. Tests of known mental abilit factors. It seems likely that
tests of V (Verbal Knowledge of the Naive Language), R (Reasoning),
W (Word Fluency) in its various forms, and M (Rote Memory) will make
at least some contribution to the prediction of success.
2. Perce teal tests. The ability to discriminate complex speech
sounds and patterns thereof is regarded as of fundamental importance in
learning to speak and understand a foreign language. Most of the tests
in this classification will have to be constructed "from scratch," but
it is believed that the preparation of such tests will be relatively
easy if suitable electronic equipment is at hand.
3. Speech performance tests. The ability to imitate the sounds
of a fore gn language and to per orm other tasks involved in speaking
a foreign language is obviously important. As in the case of the per-
ceptual tests, most of the tests in this classification will have to
be developed from the ground up. Some tests of ability to speak the
native language already exist and will be tried out, on the theory
Mat-native-language ability is predictive of foreign-language success.
it. Work-sample tests. Greatest promise resides in a group of
tests which are essen ?a ly work-sample tests, that is, tests in which
the subject learns a small sample of a foreign language (or an artificial-
ly constructed language). Several tests of this character are already
available.
Approved For Release 2001/04/04: CIA-RDP61-00017AO00100050002-5
 Approved For Release 2001/04/04: CIA-RDP61-000170001 00050002-5
(3)
It was stated above that tests will preferably be administered
to trainees before they begin training. This is because otherwise one
has no evidence as to the extent to which training in foreign languages
affects performance on the tests. However, this implied that a certain
time must elapse between test administration and collection of criterion
data. In view of the short period of the proposed contract and the
difficulty of finding suitable groups of starting trainees, there will
be a temptation to administer tests, in many cases, after the start of
training. As far as possible this temptation will be resisted, since
the evidence at hand indicates that performance in the first few weeks
of foreign language programs is fairly predictive of success after a
longer period. Therefore, the investigator would prefer to collect
relatively short-term criteria after administering tests at the initia-
tion of training than to use long-term criteria in conjunction with
tests administered relatively late in training.
The contractor will submit a final report on the investigation.
In case the investigation is successful ir$3isclosing useful predictors
of foreign language learning ability., the final report will include
details of the tests and procedures recommended for use in a selection
program. If a need for further research is indicated by the results,
the report will so state.
The contract is proposed for a period of six months, to begin not
earlier than 14 September, 1953.
The cost of the above research is estimated at $7000.00. A de-
tailed budget is presented on the next page.
Submitted by
Approved For Release 2001/04/04: CIA-RDP61-00017AO00100050002-5
 Approved For Release 2001/04/04: CIA-RDP61-0001 M000100050002-5
(4)
Salaries (including, in each case, leave and Social Security)
Principal Investigator ..........< ................ $1000.00
Research Assistants and Field Representatives ?.?. 1200.00
Consultants ...................................... 500.00
Secretarial and Clerical Assistance ???...+??.?..? 600900
$3300.00
$3300.00
Equipment, supplies, and services
Stationery and related supplies .................. 100.00
Statistical analyses .............................. 1000.00
Printing and reproduction .? ...............