ALTERATION OF CODIB MINUTES (CODIB-M-100)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP79B00314A000300070011-0
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 6
• Document Creation Date: December 16, 2016
• Document Release Date: December 21, 2004
• Sequence Number: 11
• Publication Date: April 15, 1968
• Content Type: MF

Approved For Re se 2005/02/14: CIA-RDP79BOO314A?300070011-0
Salk-C-R-3-T
C~
15 Apr fill JL~i68
HZ~MUMLYA POL 0 Cha flr' - nr Ci
LSO ICI -H-Ra l~t.
'l%i &CHMH` IL
25X1 &TT 5X? Y
~l1Da J ~ o 81-tcesot ROM of WORD l ftart~eo ? QCWII>?I3=W-10?-
X would app-Tc fiatce '~he following @homaeo An the rreporzi ng of
my oowMmiTo fair T h(2 aeetfivag of C?I X) an 13 N@Tehn 1968.
2.&,. O O O O LS epl ied
th t eotievty ogeflvnot filled z9oo 3 illeblle In The present oyotem
but Thant the pr^oggtmfiang for the t 3i toh probably had no
pr eoeant c oy of fldeuntkgf yI ng the ir, eott of ririidrievfiducA fiLeo
on oymt m pePfFor n oeo Thna@a t h d%,4 would be no pTement
ntzmno of d flot l nganflohiivng the ocelleot;ced CORMS M em ffroam other
ff fi lleo An ter m. of oy t eo pier ? r~c~ce o Mari would there
be e ny =ono oellcecctfivelly to reduce Tl: a f>'fille Loud on the CORNS
network if6 the evolluunt orgy adlditiooo of fl ILeo ohoaalld oever7Lood
the oyc team. The problem .... swfiteh o
2.1. 0 0 0 o He oooeTted that otmndQTdAsmTIon of file
elleruaevatt;o fair the pm pose of the CORNS enperlmenT would
Introduce m ooanofider?mlblle @ddifltioaall dleley in the experimevag9
would eigvnfifleQn Ly inoTemoe the cost of the euper?i g t9
ond Imo not neceoomrry to the f nd aent nl objectivem of the
extpercirmevnt, there ullreodly ?being mixed oodfiango mixed query
lmvngnogeo cet c ootsntoi neck in the oyote
3.b. loot sentence Mr. o, o, doca ento 9 CCnA
participation woo to provide content oonts` rol code on documents
as they were diotTlbanted which provided not only for a test
of the mlbillitsy to offffin the code but @ loo for a test of the code
as an coolot to ddflooemin@tion tshnoh he anndesatood to be the
objective of the teat.
CH& Member
CDDIt
25
25
Approved For Release 2005&i1iili?=KDP79B00314A000300070011-0
 C?l`50 I M21lcm pSL
c34 =110 .
Cho , into on ReM 2% ofd t ce CMEM ?o nob om M (PO goo UO(D 2m C OM
5Q ~~1 ~ I eyollc {~io~a ItnQll o Coked by WDD RIB 'RM TeeyfLVS tda(R
onl~G@z &GpoOt md to imlbwtce UhDt DaceT -R:DacR plan Zoos cg
ol MA
nL ySl% dloTO on Q o noQ Mndl(.W Of COME C -=M be
~~c is tc~t c
1 ? the oafs ?f teen Mao' Tco n(Dedl by Rte =)Ilb bell Boa, fiMllmfiow
ini,~flalllly in t Mla ?P p~~f ent~9? Q aA aonn of is fib t lllL olloo
be oeQfi bd in t oyoTc o Me I eyab o ion I ?b G &,G36 Vno a oILy
o8 the fro It bovf 'mo o m note 4o M93:
L. an ? 1 r ft reeyallaaaf ce the Pey'ffo uacce ofd the COEMS I1aQW&MPCe/
oofi more Otaboyotpa, OLI the a ivitsy fits the oyotam maotba.
,.:.
emofidlc dl o at ho not toeehnf 6a1lly fceo ilbllce to m2he o that i@t i@elL
Lyoceo t3ndl eot teo of pce c~ nc~~ aBamnctcs-fiotico an t nQ
Ibooio o:Z f IMI Ds=Lc .lsce ofd the inEM2: toe andl ts'Es'mEgac
aitDain9 a osyoteB@o
00t o on My a~ceotoi?tc l 3eq: of fiil eo ern IbQ PTac punt (Ed
oopaPat TL if deogx~edlo Ent data on the otDac Mt0 In tDace
oyofta~ sillll alloo Ibe mail oblT
Approved For Re,-
5?Dae pllonncd ingdlcept h i?at ce en~ic of dly of oyot em aatfillinotion
mod ll anoceT iaotion ooaabdl be 11ruafite?Il To meePO of the
a~ t iots`~c
QQT of M em" but the 7 valLmT10n Panel Ibelievea that dlcecioflosao
on the emapooition of the int~ce e~i ?. O 2Lceo ohoan)Ldl be dlcefZtewcedl
until ooc dlatsa on Moll mw ?f the Mee in e, ffiyotw we
adafillob Lcee At pweoerat9 we know rebmgfLerely little ahoant a1rnat
counTits''ant eo a aaocef m L fillce 9 we wcanlLdl' trnoft s t9 TIlacemf?zce, to
S geTT Lot he field of epic too ooono
With Pceop=t to the ?M Itn&L ? o uPit mia 2 ma 3s its uIldt
icppceay to be mm0t ImpaT ent in selecting a Me f?P inalLenoi?n
In COUNS to a efflIme the pacemcent intceaaascucey coaa apniCatiorrn
TeLOTIVG to the dote In ganceot tom and to develop ceotI=tceo
Off tllace,nrbc so of zaceaiceo t&net oaoIgM be en tced1 ft
inteo eot ed nZ envies
IFOR DIFIFECM USE ONLY
Approved For Release 2005/02/14: CIA-RDP79B00314A000300070011-0
- ,79B00314A 300070011-0
 Approved For Re a 050, /1t,IA , P779~B~00314A,300070011-0
15 APOL . 196 0
sEU.2.Kcm P W223
C XB M-101
MikCHNEW . 3
l1, t i~sl~t its Il ? itrs of o or ~r e cal ofs as l dzu an
Util zo -lan ud Uza D g@ bmn Q i Q e: ? e h 2'V
0oft OyIIgem wi 1 b(2 290ml?ad y~t~ii~ally f?P i ye@P b3gihiing
1?n Only (Df 968 In ~;Wu y0 an by the oyota~B itcegf :am o lcg smfingminned
by nt h ?Q1 L t 'f a@ thsounsn tie M120fin omd by z9ano of anSc
2as fo lamed by InGM MR: rm Me c i i log 10
@l> woft f~ f SEE r02Mt fiMiltrng G rem A of tRbe pPe e t ILJ
) @aT1Wityd The mow 110f fam WILL sere to gat her &t onn the
idezt Sty off is he wisp and b2.c office Q the tAma@ s R~ which qmePies are
p cd and Yeo w P ilved at the is ermine lL, the
azceP ? o
omtiof otitona or ds~ s~ f f ~i f~c~~ with the res se9 the grii ilgr l
oea (s~ fog d~a$s~et ii ft~r~t it?nn, and an Inndfiot tiion w to whether the
Qsia2 2mfor tjona could have be obte ed ilm cow other waiy Tina 'time to
serve the user's lol cez a The data from the log form Will be
coded end together with the dMt from the switch log willll be
lysed at the end of e&t t =rr n period t armug ioaat the year,
In order to ob i i statietiio5 on nze of eon file,, response i s
Preens flow delays, non-men onuses ad Peep for them, &And
a ser Pematiors , The uzse log fora will be tried out on so preeent
uncer o before e final verslan of the form iq' reoo anded for uze
t out COINS.
2. Dare Intensive ano vie of tali t ion and user Pena lon o An
ilantaq,v.ew guide is be:bng developed for i ra-de h Iiiiri vic of selected
uaers in order 'W'Learn more t at the ?iroemstances of uze, with
Mftemtion to both favorable and unfavorable ex erienoeo with the
3y stt , to obtain user?t3 views an defioienales, needed i rover nU,
tanI i ng , &Dt iti l use Of Duch a cyst , mx~d zlLte tiwe3 to ti;?X
Belk-s'tilann of the sample of users to be -Interviewed willl be based
Largely on the bimonthly statistical date an use and user reaction.
310 DemaP .t tiion of errndii a aiL oc flitaonns affect ~?n eo In
order that the data on use and user reaction e~y be ~?o~aerly understood s
l rtioi ti ng agencies will be asked to describe the environmental
ctondItio In each agency that may effect use of CUi NS9 such as
training, essistence provided, eteis taken to encourage or urge
potsatiel users to try system, booetion a of termi & is in relmtio n
to pateetiel users,, and previous experience with on-Line querying
of ooeuterrbee files
FOR 0}ffgcCicL USE ONLY
Approved For Release 2005/02/14: CIA-RDP79B00314A000300070011-0
 Approved For Rel. 2005/02/14: CIA-RDP79B00314A0o300070011-0
FOR OFFICIAL USE ONLY
25X1
I Apr 1l 1968
2VIA ATZ?? OF COINS 9
Tho puPposce of this brf paper iis_'to fur~ihcev The prroc~cess of
ollarrriifyiing the uceennimrg and the purpose of 9Ie'r-:u@ttlorn" with rrespcectt
t? C02 H9 ~. I veP since my - m~Dpointtmennt his cehaariirim sr of the CORMS
I9v611aabtf.onn Parnns 9 II have beam ]bothered by- the sceu [inTiic ambiguif fig
the tteluaaas "experiment" a rrnd Y9cevtaiuationnly as applied to COK !S9 and K.
beliidve It would be helpful to all concerned With C OM If we cvugd
agree on what Is to be understood when we us6 these teams.
It has seemed clear to me from the CORNS doom ?ntts that
'experrimennt9l has been used t U. along iinn the. broader sense of a
?'trrial9' and not In the narrower sense of a scienntiifiic or controlled
experiment. The community has developed and will be trying out a
system that allows analysts to access iunfbmatio n in selected fIles
in other agencies, .1 suspect that there would be little disagreement
with this understanding of the nature of the CDXNS "experiment.
The term "evaluatlo, M46"' however,, causes 're ddiiffficulties, II
would argue that "evaluation " like "experint nnta'" should be under-
.Stood In the broader sense of "learning as much as we can from-the
COINS experiment 9 or trja.19 4? 1n order to determine whether we are an
the right track with this approach to a c runiity system and to decide
on the most promising next steps; It should 'not be understood In the
.n . avrower sense of measuring the performance brr usie of the system against
some standard, or base- llfneq or set of eValluative crrfiteria. Ii' shall
trry to explain why Z believe former interpretation of "evaluation" is
the more realistic one.
Since ?ZNS was .not plShnned as a controlled experri nntt9 the
groundwork has not bet n laid .for comparison in any quantitative -sense
of the eeffectivenness of-COINS with whatever info al arrangements for exchange of Information constituted the comaunity "system" before COINS.
In other words, comoarrisonn cif the "before"amand the "after" Is not .
feasible because we have no data on the "before," no base line
against Which to measure the changes brought about by the Introduction
of the new system. We cannot "evaluate" performance and use data on
C?;-Lm against criteria that must be at, because no such ewiiterita have
been established; nor do we have any statement of precise requirements
or design objectives that would constitute a sound basis fOr dewelop.n '
such criteria now. And we cannot campare the effectiveness of C?'XNS
FOR OFFICIAL USE ONLY
Approved For Release 2005102/14: CIA-RDP79B00314A000300070011 r0
 Approved For Rese 2005/02/14: CIA-RDP79B00314A, 300070011-0
FOR ?WXC1 L USE OMW
2 4
with any altemartlvice rrncee3 Enefle ois acccmpUeh n - th(e Z obJeCTivc s,
because only one of the arltemotfives was ftEpllQu rats do X am not
ieaplyrZng that things should have been donee that wGwa not donee, faT
?C Think that in many instances a Coed case cairn be ode fors LatMduofth g
a nndw system in order to give the users expePfiennce with .it and To
learn how they usd and how At might be in tdQ to serve them better.
lllnat we can do in the present circumstance Is to learn as
much as possible about the community?s c-xperienc(e in atticeanpting
to wise CONS and about the tcecnical perfoPmonnee of the system; and
this is the objective of the plan the CORNS )Evaluation Panel wus
askied to develop narrrrelly, a plan for acsquf.Pirrag, awlyz ng and
presenting data on system performance, system utilization, and userr,
reaction.
The data on the performance on the CORNS hardware and software
will be quantitative In nature and should be useful In indicating
what needs to be done to retake the system perform better and to over-
c user dissatisfaction caused by inedd? quacies in technical
pcerrforimweeo I would say that such data will give us performance
ambsurcements, but not evaluation in the strict sense of meeting
certain standards or criteria.
The more significant: question of potential value or utility
of such a system to the community and decisions as to where we go
from CORNS 1t will depennd more on the response of the users o II
believe the main goal of our analysis of utilization and user reaction
should be understanding of the factors that account for differing levels
of use 9 users ? reactions 9 and their views as to the potential utility
of such a system. I would expect that conclusions as to what has been
learned from the CORN'S experiment and decisions as to tie next steps
will have to be products of informed Judgement a4 inaforrmed9 I would hope,
by the data we plan to present. In this connection, it would be
helpful to the COINS Evaluation Panel to understand more clearly what
kinds of decisions must be made9 and when, on the basis of the data to
be presented o
In the last analysis, use is the best indiicattor of the value9
or utility, of an information system, but as of now there Is no
basis for saying how much use must occur before the CORNS system ors
any file within it has demonstrated its utility. Statistics on use
must be treated with considerable -carrea Even If a new kind of
information system has reached the operational stage, and -CORNS 1:=; not9
it still may take quite a'. long while for the e? unity of users it
Approved For Release 2005/02/14: CIA-RDP79B00314A000300070011-0
 Approved For Rese 2005/02/14 :CIA-RDP79B00314A000300070011-0
FOR OFFICIAL USE ?AII,Y
serves to learn to make effective use off it and to adjust their
work habits to it. Ile sharer}. be pr-oposfig9 to oe o to gyothc T and
@nal~ze not only quernag$tat gve dtatta an n but also detailed innffor at9an
f rain selected uaess co the trim. for Tbelw aaisfims @and views sand
descriptive f ffotion an the differing coQQAs An t p~rr~ticiipaf~fir~g
agencies that may help to expl@ien difffcent pa of' use.
We shall propose to gather and an&A_yze drat or both performance
and use for a full year, because we und)ratand thrt ihae. ajst will
continue to evolve. Performance data ifty ehahge so what as equip nt
changes a; made and as .oft Warr is improved ii a ore to Qehieve
better performance.. Use patterns should ehonge &s users learn more.
about the system, Chan s in both perffor n e rand use data shoutld
occur if new files are added, or if files are roved from the system.
COINS has to be vi d0 therefore, as -a dynamic system,, and changes ftn
it should be inflected In the data presented o
The COINS Evaluation Panel was also asked to recd nd other
data gathering and analysis activities that would be desirable An
connection with system ev aluaationo We have not yet Turned our
attention to this task0 but we may well give .further thought to they
fourth point made in our response to the C?tN'S files questio 0 L3a
the case of any file recomnded for Inclusion in COXNS o It may be
desirable to require that C=art. he presented on the extent of present
file activity,, an interagency co nication with respect To such
data, and on .net demands for such data throughout-the c aunfitya
There would then be a base line for comparison with unity use of
the file after it ?becoms available through C020So
Approved For Release 0R! &IA D 0314A000300070011-0