DEAR PAUL,

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B05171A000600060069-3
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 3
• Document Creation Date: December 28, 2016
• Document Release Date: February 7, 2003
• Sequence Number: 69
• Content Type: LETTER

Appro d For Release 2003/03/04: CIA-RDP7$W5171A000600060069-3
STAT
College of the Liberal Arts
Department of Psychology
First of all, I feel somewhat frustrated in trying to put comments on
paper, When you write something, it acquires a kind of permanence which
* spoken comment does not imply. I really need more information regarding
a number of the subtaske both from the contractor and the sponsor. If
possible, I could certainly be of more help in a face-to-face situation
as I am sure that mate of ray comments or the thoughts on which they are
based, could easily be answered. With these reservations, here are my obser-
vations ordered according to subtaske:
1. Performance measures: This research would seem to me to deserve a high
prior y. ica none of the so-called research in image evaluation
has h a d the benefit of objective performance measures. W i t h o u t such q u a n -
t i t a t i v e indicators, one can s i m p l y n o t be scientific. The contractors pro-
posal a p p e a r s to be well t h o u g h t o u t and. well written. I only su estion is
that you might want to get a reading from someone lik not so
much in evaluation, but for suggestions to the contractor as to how the re-
search mey be carried out.
2. TI,,,,CQT seems like a splendid concept. It is often the case that ideas
and, concepts which seem reasonable in the abstract run into difficulties
which one attempts to apply them. This is especialSy true in operational
situations such as yours. TICOF would provide a check of t: a ideas, and
also serve a teaching and demonstration function.
3. Intea r ter machine namics: This proposal was not as impressive as
the first two. The or is relying heavily on the work carried out
STAT by which I have not read. What is the experience
of the contractor in this area?
4. Image-characteristics: This is an area in which APRO has had extensive
experience and., ifat a possible, they should be consulted. I cannot de-
termine, without additional information, whether such an analysis would be
of value to you. Do you not evaluate all imagery independent of its quality?
Or the other hand, I am certain that the kind of data which one may obtain
from this type of research would be invaluable to engineers designing sen.s-
STAT
STAT
Declass Re e kreU1 'i 1 e9(6se 20043/03/04: CIA-RDP78BO5171A000600060069-3
 Approved For Release 2003/03/04: CIA-RDP78EW171A000600060069-3
tion.
terns. At this point, I should not comment without further informa-
5. Ir e-utilization..stereo: Asst everyone I have interacted with during
my survey cvxemen to e, , us with emotion, regarding the value of stereo.
My own observation based, on the available data and our knowledge of the vis-
ual system is that stereo should be unquestionably useful under some conditions,
Would appear, to be helpful under others, and of no value in some cases. I
STAT do not quarrel with al's results, but feel that their conclusstons
cannot be generalis > Clearly, what is needed is a definitive study using
ground truth- Mwlaaa th
-.1 - _~_ __ _. _
._.._
e
- -
le
d
-
and compreh
en
lively, the data will, be of li ted~usefu nee s and will add to rather tha
n
contractor's Proposal does not strike _me~as being adequate to spr
The roduc finie
itive conclusions. I think they could do a real good job on this problem
d ees
If supported and encouraged.
I will not comment on the apparatus problems.
v. v7.suai mecan.sm .' This is the area concerning which we have had muc
ga d ng
he s
f
_ _
,
Y%"v o
t~i:G:vli1~l o i7 i.3 on S I J IS
the measure of accozm odation of value in predicting or monitoring interpreter
performance? This is tacitly assumed in the proposal, but I have not seen
the supporting evidence. Accommodation in a critically important visual
snesrohia>! 6MR b
t 4... -a _ _, _
u
leva
..._
n % v your situation ra w
quires justification.- or ox alple, why is accommodation more promising than
blink rate, ."ttsional range, or even pupil diameter, all of which are linked
in the literature to visual performance? What is the value of studying blood
flow by infrared techniques (any personal bias)? Unfortunately, accommodation
is the most difficult of all these measures to quantify. If I were in a po-
sition to make a decision, I would insist on more Justification, particularly
in view of the relative difficulty Of measuring acct dation in the opera.
tional context. 2) We discussed the precision of the ultrasonic technique
STAT'af ore and. I pointed out tha_t~ questioned both the validity and pre-
cision of the method. You indicated that your re ant trip to
was reassuring with respect to the precision aspect. However, we a- oratory
types are eternal optimists and you should bear in mind that the justification
IS
nr~t v~a#~.-
f
L.ZCir.I. eSZ.lzoiiVal
~y,~,~} y,, checked y_ ~n y F review,, and has not been -_ _.__??' ~w?vwv.r Y+.r.LG wzow o
by other investigators. Also, it might be
wise to make a mental note that laboratory scientists are notoriously in-
accurate regarding when their research will be completed. This latter com-
ment would be particularly applicable in relation to your operational situ-
ation.... Overall. I am pessimistic about this subtask. could
25X1
STAT
Approved For Release 200?/03/04 : CIA-RDP78BO5171A000600060069-3
 Approved For Release 2003/03/04: CIA-RDP78 O5171A000600060069-3
give you as more educated guess regarding the measurement problems.
7. Not qualified to comment without additional information.
I1iumi.nation; I am most familiar with this subtask by virtue of pre-
STAT vious discussions with you and with I My feeling is that research
on illumination at this stage is of limited value. There is already h vast
literature on this problem which should permit specification of illumination
levels and color temperature values which would account for the majority of
the variance attributable to these variables. Additional research will be
useful, but its value in the total context would seem to me to be
As you know, we discussed, these issues when you and.l
and also during our conference phone conversation. The main justification
given for this research was based on un ur_pu'liehed master's thesis at Ohio
State on size perception. I feel that size is only marginally related to the
kinds of visual tasks with which your organization is concerned. I would
prefer to discuss this face-to-face. However, the evidence available to me
does not support a strong recommendation for this line of investigation.
9. Mensuration: Not qualified to comment
10. AU
propos
teex l; Could not be supported on the basis of the present
12. State!-of-the art reviews These can be valuable but only if properly
carri. - -57u. or exam* e, 17 the reviewer can digest and evaluate the ma-
terial so that the written material is relevant to your operation, than this
conilation would be valuable. If, on the other hand, the review simply lists
uncritically the studies falling under the various headings in the outline,
the only advantage is the saving of some time in the library. Take for ex-
aple the value of stereo viewing. There are at least a thousand papers on
this topic concerned.with range finders, image evaluation, pathology, per-
ception, eta. The information of value to your people would be in the form of
statements similar to what I indicated above i.e., that stereo is of value
under some but not all conditions, and that these have to be identified. A
listing of all the studies might simply be confusing. One could do better by
making a few telephone calls. I cannot help but be reminded of the review
STATyou gave me, by I think 0 which listed studies uncritical y, some of which
had no relevance at all to your problems. Most reviews I have seen are not
very helpful. This contractor could do an excellent job if so instructed by
the sponsor.
Please call if there are any questions.
Sincerely
I
o essor of Psychology
STAT
STAT
Approved For Release 200?/03/04 : CIA-RDP78B05171A000600060069-3