QMB3602 - Exam 1

D in DMAIC Framework

D - Define, Define the problem

M in DMAIC Framework

M - Measure, Map out the current process (gathering data and measuring the process baseline aka, starting data to compare to data later on)

A in DMAIC Framework

A - Analyze, Identify the cause of the problem

I in DMAIC Framework

I - Improve, Implement and verify the solution

C in DMAIC Framework

C - Control, Maintain the solution

What are the three types of Analytics and the order in which they’re used?

1. Descriptive Analytics

2. Predictive Analytics

3. Prescriptive Analytics

Predictive Analytics

using historical data to predict “what could happen in the future?”

• Used to estimate the likelihood of a favorable event

• Ex) Transactions that are likely to be fraudulent or identifying customers who are likely to respond to a specific marketing campaign

Prescriptive Analytics

refers to using optimization and simulation algorithms to provide advice on “what should we do?”

• Suggests a course of action and several possible actions

  • Ex) Scheduling employee hours, choosing an investment portfolio, and selecting a mix of products to manufacture

Descriptive Analytics (also referred to a business intelligence, BI)

gathering, organizing, and visualizing data to summarize “what has happened?”

• Ex) Financial reports, public health statistics, enrollment at universities

Big Data

a massive volume of both structured and unstructured data that is extremely difficult to manage, process, and analyze using traditional data-processing tools. (high velocity, high volume, high variety)

Characteristics of Data

• Velocity

• Volume

• Veracity

• Variety

• Values

Unstructured data

do not conform to a predefined, row-column format

Structured data

reside in a predefined, row-column format

Velocity

how fast or slow data is generated

Veracity

credibility and quality of data

Variety

all types/form forms of data (structured & unstructured)

Volume

amount of data

Values

plan for formulating questions, curating the right data, and unlocking hidden potential

Four Scales of Measurement (least to most sophisticated)

1. Nominal

2. Ordinal

3. Interval

4. Ratio

Nominal Scale

• least sophisticated scale

• can only group or categorize variables

• differ by name or label

• Ex) Marital status, gender, type of car

Ordinal Scale

• a stronger level of measurement than nominal

• can categorize and rank according to a characteristic or trait

• Weakeness: does not provide information on the magnitude of difference between each value

• Ex) Customer Reviews: Excellent (5), very good (4), good (3)

Interval Scale

• Categorize, rank, and find meaningful differences between them

• Ex) Fahrenheit Scale

• Weakness: zero arbitrary point (ex. you CANT say 2x as hot, but you CAN say 10 f difference)

Ratio Scale

• strongest level of measurement

• categorize, rank, find meaningful differences, AND ratios

• True zero point

• Ex) Weight, sales, profits, and distance

Nominal and Ordinal scales are

categorical

Interval and ratio scales are

numerical

Interval VS. Ratio Scale

raio scales have a true zero point, thus can make meaningful ratios

Population

all items of interest in a statitical problem

Sample (also known as sample set)

subset of the population/data

Cross-sectional data

data collected by recording a characteristic of many subjects at the same time.

• Ex) Data collected on team standings at the end of the season, they might have all ended at different times and different days however that is irrelevant

Time-Series Data

data collected over several periods

• Ex) hourly body temperate

Numerical Data (Quantitative)

use numbers, either discrete or continuous

Continuous Variable

1.13452, uncountable values within an interval

• Ex) Weight (128.8) or Height (5.2ft)

Discrete Variable

assumes a countable number of values (whole numbers only)

Categorical Data (Qualitative)

Use labels or names to identify distinguishing characteristics of an observation

JSON Language

• Popular alternative to XML

• Smaller File sizes

• Faster and less resource intensive

• Supports a wide range of data types, when compared to XML

{

“Person”:{

{

“Name”:”Rich”

JSON

HTML Language

• tells the web browser how to display the data VISUALLY

• uses tags <>

<table>

<tr>

<tr> Name</th>

HTML Format

XML Language

• used for structured data

• tends to be much larger than fixed-width and delimited data files

• time-consuming for downloads

• case sensitive

<Data>

<Person>

<Names">Rich</Names>

XML Format

DFF Language doesn’t exist

true

Delimited table format

• stores tabular data

• separated by a comma

• known as comma-delimited/separated value (cvs) file

name, telephone, salary

Delimited Format

Fixed-width with hypertext

• table format

• Everything is SAME width

• HTML

• Limit of 8 characters

High Interest & High Power Stakeholder (KEY PLAYER)

• Manage closely

• Involve in projects and decisions

• Engage on a regular basis and work to maintain the relationship

High Power & Low Interest (MEET THEIR NEEDS)

• Engage and consult

• increase/maintain level of interest

• Aim to turn into key player

• could be a risk to your idea

Low Power & High Interest (KEEP INFORMED)

• Make use of interest through involvement

• consult on their area of interest

• can be a suppporter/ambassador

Low Power & Low Interest (LOW PRIORITY)

• Monitor

• communicate generally to keep updated

• aim to move to right (keep informed)

Stakeholders

individuals, groups, or other organizations that are affected by and also affect the firm’s decisions and actions.

• Ex) government agencies, activist groups, employees, shareholders, and suppliers.

Three stakeholder’s perspectives

1. Separation perspective

2. Ethical perspective

3. Integrated perspective

Separation perspective (Smith)

suggest that the manager should always strive to act in the best interest of the firm’s owners. (Focus on the OWNER FIRST)

• Does not ignore non-owner stakeholders, still focuses on them

• Rationale is different from other perspectives (managers make decisions and take actions benefiting non-owner stakeholders in order to ultimately reward the owners.)

Ethical Perspectives (Kant)

Businesses have an obligation to conduct themselves in a way that treats each stakeholder group fairly.

• takes shareholders’ interests in consideration only to the extent that their interests coincide with the greater good

• Ex) choosing not to advertise beer to underage audiences, despite loss of shareholder gain

Integrated Perspective

Firms cannot function independently without considering stakeholders, thus when managers make decisions they have to think about everyone else involved

• Self interest and interest of stakeholders (BALANCE)

• Ex) company changes operations, but they give employees a two year notice

The 5 Why’s

determine the root cause of a problem, ask WHY 5 times

• used for simple, low risk problems

Seven So What’s

thinking tool that reveals the consequences on an action, ask yourself SO WHAT 7 times

80/20 Rule

20% input leads to 80% output/results

Pros and Cons of the 5 Why’s

Pros

• Easy to learn and use

Cons

• not data driven, thus cannot be repeated

• biased, based on personal choices

• only one root cause

Fish bone diagram, what is each branch

Each branch answers WHY is this a cause?

Fish bone diagram, what is the head

The problem (there can be multiple problems)

Machine as a branch of the fish

Example: car broke down

Manpower as part of the fish

Example: boss told you the wrong time (someone else’s error)

Mother Nature as part of the fish

snowstorm

Measurements as part of the fish

alarm clock time incorrect

Methods as part of the fish

forgot to set alarm (your error)

Data preparation

The processes that ensure the accuracy of data and their conversion from their raw form into categories appropriate for analysis: includes post-collection coding of data and editing of data

How to address out-of-range erros

• summarize the data on each variable to determine if the codes not specified were entered

• reviews the original data instruments to determine what response was chosen

• determines how to address each instance of out-of-range data

• may drop a question from the survey containing too many out-of-range data errors

Anchoring

what we see or hear first feels most important = disproportionate weight to information received

Anchoring trap

• take a different perspective

• Think before asking

• seek input

• avoid anchoring others

Status quo

prefer things to stay the same, not open to change

Status quo trap

• consider objectives

• know you have options

• dont exaggerate switching costs

• dont settle

Sunk costs

decisions that justify past choices even though they no longer seem valid

Sunk costs trap

• involve outsiders

• accept that mistakes happen

• allow room for failure

Confirming evidence

seek information that supports our existing point of view and reject anything that doesn’t

Confirming evidence trap

• consider all evidence

• find a devils advocate

• avoid people who don’t agree with you

Framing

the way we state the problem influences what we decide

Framing trap

• always reframe

• think in neutral terms

• challenge others frames

Overconfidence

we tend to trust our own estimates too much

Overconfidence trap

• look at extremes

• challenge those extremes

• question estimates

Prudence

we adjust estimates to be on the safe side

Prudence trap

• aim for accuracy

• ask others to do the same

Recallability

we are overly influenced by our past experiences, prompts us to give too much weight on recent dramatic elements

Recallability trap

• check the facts

• examine assumptions

Solution Question

choices between concrete actions to solve problems or take advantage of opportunities

• Ex) How can we improve our program for product repairs?

Choice of purpose question

choice of objectives

• Ex) What GOALS should xyz try to achieve in its next round of negotiations

Troubleshooting question

diagnosing a way an organization is failing to meet its goals

• Ex) Why is our x program not meeting its goals?

Control question

monitoring or diagnosing ways an organization is not meeting its goals

• Ex) how well is our x program meeting its goals

frequency table

arrays a ctegory of codes from lowest value to highest value, with columns for count percent, valid percent, and cummulative percent

Exploration Strategy

a process that defines what exploratory sources will be used and what methods will be used to extract that information

What two questions does the exploration strategy ask

• What information sources should be used?

• What methods will be used to extract information from the sources?

The exploration strategy is composed of three subjects what are they?

1. Published sources

2. digital sources

3. interviews

Primary sources

original works of research or raw data

• ex) memos, letters, interviews, speeches

Secondary Sources

interpretations of primary data without new research

• Ex) textbooks, encyclopedias, handbooks, magazines

Tertiary sources

aids to discover primary or secondary sources

• Ex) indexes, internet, search engines

How to evaluate internal sources

• compatibility

• timelessness

• recentness

• quality

• authors experience (credibility)

Internal sources can both primary and secondary

true

Internal sources

• published sources

• digital sources

Through exploration, researchers can:

• Establish range and scope of management decisions

• Establish major dimensions of the research task

• define asset of sub diary questions that can guide research design

• develop hypotheses about possible causes of management dilemma

• learn which hypotheses can be safely ignored

• conclude additional research is not needed or feasible

How to evaluate external sources

• purpose

• scope (depth of topic coverage)

• authority

• audience

• format

Data Cells

cells in the spreadsheet that show the data of the problem (LIGHT BLUE)

Objective cell

total profit (ORANGE)