combined systems notes

10 Golden Rules of SA

CUGIATCGNFC

• always client

• client doesn’t understand their own problem

• goal-centered

• client doesn’t know IP

• a* - you are analyst

• t*c* - meet time deadline and cost budget

• consider nonusers

• f* - universal computer model is fantasy

• c* - decision maker role is often confusing

standard flowchart vs systems diagrams

steps

relationships, feedback loops, unintended consequences

systems methodolgoy

concept dev, design & implementation, validation/testing, operations, disposal/replacement

msbe

digital tools/models → improve system lifecycle

how to use systems diagrams

• use diagrams which best meet need

• gain deeper understanding and identify and resolve ambiguity

• show ideal system

descriptive & normative scenario

as-is

to-be

system

interconnected

interrelated

achieve on goal

causal loop vs stock flow diagram

R&B,

+ or -

SA

1. goals

2. criteria for ranking alternative candidates

3. develop alternative solutions

4. rank alternatives

5. iterate

6. action

goal development

1. generalize question

2. descriptive scenario

3. normative scenario

4. axiological component

5. objectives tree

6. validate

7. iterate

systematic vs systemic

systematic - step by step

systemic - whole system

rand approach

SA = OR + PA

root cause analysis

5 whys, ishikawa diagram, etc

value stream map

yes/no

failure mode and effects analysis

fail → effects, how, how likely, how bad

ITS

tech to increase performance of transportation

systems integration

logical, objective procedure for applying in efficient, timely manner new and/or expanded performance req to design

SE =

SA

taylorism

hierarchal

top down vs bottom up

adjust to environment, lose focus/productivity if not done correctly

immediate evaluation of cost effectiveness, short term, ignore potential conflicts

reference behavior period

graphical rep of behavior ofver time of >1 variables in loops

observed behavior period

historical states ofv ariables at given period

cybernetic loop

various info systems that inform → possible change plan

dynamic equilibrium

inflow = outflow

input

resources that cross boundary from environment to system

E = ARP

availability x reliability x performance

stable environment

can use historical data to predict future

IP

(SCND)

sensitive

comprehensive

nonduplicative

proxy metrics

specific, numerical, indirect

streetlight effect

drunk man looks for key under light b/c that id where the light is

analysts often choose metrivcs because easy to measure not because relevant to objective

types of criteria

technical, economic, social/political, environment

pluralist view

diff stakeholders have diff goals

sensitivity

must change when system design changes

stock and flow, process flow, causal loop

accumulation

“how-to”, order of operations, sequence

feedback loops, relationships

cloud

originate of flow, sink of flow

v model

regional architecture, decomposition/definition, implementation, integration/recomposition, retirement/replacement

low then high then low technical detail

4 pillars of a good requirement

verifiable, unambiguous, atomic, traceable (VUAT)

objectives =

metrics =

goals

IP

Objective & Functions

• function - verb + object

• objective has direction but function doesn’t

• objectives/functions of the system, not the team

• solution independent - one at a time, no overlap

• structure as hierarchal tree

functional & design requirements

design - storage, tracking, access

At a high level, you look at

objectives of system - as is/to be, direction

functions of system - core things done by system, static (not directional)

At a measurable level, you look at

IPs, requirements

point of an executive summary

provide key decision makers with enough information to make informed decisions regarding original document without reading the whole thing

what to include - go back to original purpose of report

evaluation criteria

how much does it contribute to the purpose?

how extensive does the treatment have to be if you included it?

types of executive summaries

descriptive - describe structure of report/key points

informative - describe content and substance

informative abstracts

of a study

Summarizes key findings of study

• May just describe a problem and/or characterize a situation

• Example of a research report or update related to research project

decision recommendation

• Should provide decision maker with concrete advice supported by appropriate

evaluation

executive summaries

• placed at beginning of presentation

• executive is the audience

• basic context

• action based recommendations

• if then structure if selecting between multiple criteria

• >1 key number include

• WHAT and WHY of case

• summary and roadmap

writing executive summaries is hard!

PDF vs CDF

PDF - regular probabilities

CDF - cumulative/total

expected value

P(x) times value

linear vs convex vs concave

linear - E[f(x)] = f[E(x)]

convex - E[f(x)] >= f[E(x)]

concave - E[f(x)] <= f[E(x)]

poor decisions are more often due to _______ than to _________

lack of information, poor judgment

time value of money

financial principle that states that a dollar today is worth more than a dollar in the future due to its potential earning capacity or the effects of inflation.

(purchasing power)

if monetary value is the only criteria

different IPS can be converted to monetary values

multiattribute

4 scenarios to explore uncertainty

1 objective

Laplace Criterion

assume each scenario is equally likely

maximin criterion, maximax

take the “best of the worst” outcomes

take the “best of the best” outcomes

elimination by aspects

return must be at least 1

compensatory vs noncompensatory approaches

compensatory:

• weakness in one area can be “made up for” by strength in another

• considers all attributes of a choice

• choose option with highest overall utility or value

noncompensatory:

• one attribute cannot “make up for” another

• eliminate alternatives that do not meet a particular criterion

• maximin, maximax, elimination by aspects

how to develop alternative scenarios

• outscoping

• examine various elements and assemble into complete candidate solutions

• brainstorming

• brainwriting

• dynamic confrontation

drawbacks of brainstorming

competence of facilitator, extroverts dominate, uncomfortable, emotional & stressful

zwicky’s morphological box

1. Define all functional classes that make up basic subsystems

2. Exhaustive lists of tech. alternatives developed

3. combinations

4. each combination examined

steps to developing alternatives

understand problem

understand how parts are interconnected

carry out plan

options field/options profile

ordering principle (steps) with critical features

group decision support system (GDSS)

end user tech, analytic decision techniques/alternative ranking, human facilitation, customized facility layout, specialized equipment, underlying process

• HCI

• poorly design designed system can affect communication

• preference elicitation/idea expression can be hindered with poor interface

computer simulation/models

examinations of complex interactions becomes practical

validation by ge