CONTACTS 

SUBJECTS
Analysis Poor Formal Writing Reports Professional Business Writing P

3203 -
3203 -    ..
3204 - Summary/Objective
3205 -
320501 - Follow up ref SDS 33 0000. ref SDS 28 0000.
320502 -
320503 - Asked Gary to comment on challenges preparing written analysis for
320504 - decision support, reported recently in testimony before Congress
320505 - explaining erroneous intellingence on national security. ref SDS 0
320506 - SY3O  Earlier in February Gary did a similar review on problems that
320507 - arise in daily communications.  In a recent telephone call he
320508 - indicated some of the same forces impact written reports even when
320509 - prepared with teams collaborating on major projects. ref SDS 0 LX5F
320510 - This review helps illustrate requirements for tools and practices of
320511 - Communication Metrics that reduce the time and effort for finding
320512 - critical details when needed to prepare accurate analysis.
320513 -
320514 -
320515 -
320516 -
320518 -  ..
3206 -
3207 -
3208 - Progress
3209 -
320901 - Analysis Writing Accuracy Reports Poor Quality Control
320902 - Documentation Accuracy Poor Reports Writing Analysis Intelligence
320903 - Intelligence Poor Accuracy Reports Writing Analysis Documentation
320904 - Writing Accuracy Poor Intelligence Reports Analysis Documentation
320905 -
320906 - Follow up ref SDS 33 6O6O.
320907 -
320908 - On 060408 review of Gary's record on 060222 shows mistakes in daily
320909 - communication from relying on personal memory that gets confused by
320910 - the complexity of constant information.  People giving up on good
320911 - management paralyzes productivity. ref SDS 33 424O
320913 -  ..
320914 - A study on 060408 shows examples of communication failure due to
320915 - failure to perform due diligence checking the record to authenticate
320916 - understandings. ref SDS 33 PV4O  Recent testimony before the Senate of
320917 - the US Congress reported that analysis for intelligence and other
320918 - reporting that strives to make sense from complexity of daily working
320919 - information is often incorrect because people do not have time to
320920 - verify representations, and so rely on personal memory from
320921 - conversation.  Since conversations are rarely authenticated, this
320922 - makes written analysis largely cursory, stream-of-conscious work
320923 - product that is unsupported and so often erroneous and incomplete.
320924 - The requirement for "sensemaking," described by Jack Park in a letter
320925 - on 040622, ref SDS 24 FM5W, largely fails.
320927 -  ..
320928 - Yesterday, in a telephone discussion, Gary commented on challenges
320929 - people face preparing written reports with analysis of engineering and
320930 - management issues, which includes health care.  He said that testimony
320931 - before Congress on difficulties preparing intelligence analysis for
320932 - national security occurs in business and other areas, e.g., health
320933 - care, education, science, etc., due to limited time, and difficulty
320934 - finding critical details in time to meet deadlines for deliverables.
320935 - People call collagues and collaborate in meetings, and through
320936 - correspondence, primarily using email to gather background.  Since
320937 - these procedures are inherently error prone, due to limited span of
320938 - attention, as reviewed on 060408, ref SDS 33 6O6O, work product for
320939 - formal reports and analysis is sometimes largely unsupported, belying
320940 - the appearance of well researched and documented analysis.
320942 -  ..
320943 - Gary said another problem arises from bureaucracy, reviewed on 040203.
320944 - ref SDS 18 E44K
320946 -  ..
320947 - Analysts get instructions to prepare reports from managers, who in
320948 - turn report to upper level people.  Requirements on scope and sources
320949 - received from the buyer, consumer, stake holder, originating
320950 - authority, etc., of the work product often are unclear, conflicting,
320951 - and incomplete.  The analyst is told not to contact the originating
320952 - authority to obtain clarification.  They are told to submit written
320953 - questions to the boss.  This can take more time than is available to
320954 - prepare the report.  Lengthy questions that are clear to the analyst
320955 - who has studied the issue, and can be made clear by the analyst
320956 - talking to the consumer who needs support, is often not clear to the
320957 - manager, who is reluctant to bother superiors with questions they do
320958 - not understand, and so strives to avoid criticism by being a "team
320959 - player," reviewed previously on 040203. ref SDS 18 XI41  This leads to
320960 - failure to obtain clarification.  The analyst is told to destroy
320961 - written questions, so there is no documentation that management did
320962 - not act.  The analyst is told to do the best that can be done based on
320963 - what is available.
320965 -  ..
320966 - Gary said these dynamics can lead to abysmal analysis in reports,
320967 - noted in Congressional testimony, per above. ref SDS 0 LX5F
320969 -  ..
320970 - Consequences of cursory analysis that fails to apply requirements for
320971 - "connecting the dots" to verify accuracy occurred on 010911, when the
320972 - United States was attacked by terrorists. ref SDS 10 UP5K   Another
320973 - example was collapse of Enron reported on 020204. ref SDS 12 0001  On
320974 - 030826 testimony before the US Congress reported that the space
320975 - shuttle Columbia exploded on return from space because executives and
320976 - engineers ignored requirements for good management. ref SDS 17 8K4G
320977 -
320979 -  ..
320980 - National Security Suffers Cursory Erroneous Intelligence Analysis
320981 -
320982 -
320983 - Submitted ref DIT 1 0001 to Gary saying...
320984 -
320985 -    1.  We talked yesterday about recent testimony before the US Senate
320986 -        on cursory support for national security analysis.  Very thick,
320987 -        heavily footnoted analyses are mostly stream-of-conscious work
320988 -        product from personal memory, ref SDS 0 LX5F, without review
320989 -        nor support of original sources for accuracy and context,
320990 -        because this part of knowledge management, i.e., finding
320991 -        critical details, is a lot of hard work, see briefly the report
320992 -        on 000307, ref SDS 8 767G), because people lack command and
320993 -        control of the record with conventional tools.
320995 -  ..
320996 - Failure of national security due to poor intelligence analysis was the
320997 - problem that caused tragic loss on 010911. ref SDS 10 YNGH
320999 -  ..
321000 - Letter to Gary continues...
321001 -
321002 -    2.  Cautionary testimony before Congress noted that even teams and
321003 -        departments of professionals assigned to perform analysis and
321004 -        who collaborate on preparing work product, fall prey to working
321005 -        primarily through conversation, i.e., who remembers thus and
321006 -        so, rather than checking the records.  He made a point that
321007 -        Congress and the president should require submission of a
321008 -        bibliography, and that footnotes should be at least spot
321009 -        checked.
321011 -  ..
321012 - Cursory analysis is common in findings for medical tests, and
321013 - practice which nominally should get a high priority for accuracy, yet
321014 - suffer from constant error due to failure to "connect the dots" that
321015 - convert good "data" into useful "knowledge," see the study on 040416.
321016 - ref SDS 22 PTSX
321018 -  ..
321019 - Letter to Gary continues...
321020 -
321021 -    3.  If time permits, you might prepare a record to generally
321022 -        explain how formal analysis is performed on the job, along the
321023 -        lines of your record on 060222 that discussed communication
321024 -        practices using email, as reviewed in my record on 060408.
321025 -        ref SDS 33 0001, and contrasted with analysis in an SDS record,
321026 -        illustrated by the study on 050812. ref SDS 26 0001
321027 -
321028 -
321029 -
321030 -
321031 -
321032 -
321033 -
321034 -
321035 -
321036 -
321037 -
321038 -
321039 -
321040 -
321041 -
321042 -
3211 -