YOUR REQUEST FOR ANALYSIS OF SOURCES OF INTELLIGENCE ITEMS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80-00948A000100010014-5
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 2
• Document Creation Date: December 20, 2016
• Document Release Date: March 31, 2006
• Sequence Number: 14
• Publication Date: January 31, 1955
• Content Type: MF

Approved For Rele a 2006/04/13:: GfXLF~
'I;i4iWks
31 January 1955
SUBJECT . Your Request for Analysis of Sources of Intelligence
Its
1. I attach hereto two tables giving percentage breakdowns of
the sources on which the 1954 production of finished intelligence by
my offices has been based.
w
i; y re
gable reports that ive veracity to a whole mass of basic
2. The first table gives all offices engaged in published intel-
ligence research, and the second gives a further breakdown with respect
to the Office of Current Intelligence. The reason for this is that
only the latter is broken down essentially by geographic subdivisions,
whereas the other offices, as you well know, are functionally subdivided.
I hope that this OCI analysis will be adequate for the purposes of
enabling you to evaluate the relative contribution of different sources
to intelligence on each of the major areas of the world.
3. You should note that the first, table is based on the total
of primary intelligence sources used. Thus, citations to other finished
intE-71 igence are excluded, as are references to unclassified authorities
and literature. The first seems justified because it is safe to presume
that the cited finished intelligence in turn depends on approximately
the same proportion of various primary sources, as do the 1954 reports
I have had analyzed.clusion of non-intelligence items is justified
because in large measure they are cited as historical backgrounds or as
substantiating the logic of the analysis or to provide the reader with
,guidance for general background reading on the topic.
4. Finally, I hope you will keep in mind the serious dangers of
being misled by such statistics. For example, one-line items are given
the same weight as a document or dispatch many pages in length.iore-
over, qualitatively such a table ignores the fact that a reliable report
that a cabinet may be about to fall may outweigh a hundred items of
limited significance. The structure of our knowledge depends on all the
information we get, and its strength depends particularly on a relatively
fe
h' Ohl '
1
Approved For Release 2006/04/13 IA;=RD
Eq a - .5-IP
P80-00948A00W010014-5
25X1
25X1
 e,r e- S" 9.3c
Approved For Rel a 2006/04/13: CIA-RbPP.r 948A00 0010014-5
data and focue attention on the portions thereof that are relevant
to policy making. Thus, one clandestine-service-obtained leak from
a highly placed government official may be the all-important item that
confirms dozens of otherwise inconclusive indications and enables the
rejection of an equal number of otherwise plausible bits of evidence.
5. if 1 can be of any further use to you in this or any other
matter pertaining to this review of our work, please let me know,
25X1
AUBERT RYa R. I 'I
Deputy Dire ctor/intelligen e
Approved For Release 2006/04/1x3:CIA-RDP80-00948A000100010014-5