QMB 3602 Exam 2 UCF

Management Dilemma

What are the symptoms that are causing management to investigate an issue?

Management Question

How can management eliminate the symptoms or capitalize on the opportunity?

Research Question

- Questions that explore the various options

- These also form the hypotheses that best states the objective of the research

Investigative Question

- Represents the information that the decision maker needs to know

- Ex: what is the publics opinion on a product?

Measurement Question

- Questions that participants in the research are asked/what is observed in the research

- These questions are the actual data

Types of Management Questions:

1. Evaluation of Solutions

2. Choice of Purpose

3. Troubleshooting

4. Control

What type of management question is this:

How can we achieve the objectives we have set?

- Evaluation of solutions

What type of management question is this:

What do we want to achieve?

- Choice of purpose

What type of management question is this:

Why is our program not meeting its goals?

- Troubleshooting

What type of management question is this:

How well is our program meeting its goals?

- Control

Exploration

loosely defined research with a clear purpose

Tertiary Source

an interpretation of a secondary source

Literature Search

a review of literature related to a management dilemma

Books and journals are _________ sources

published

raw data, dictionaries, and handbooks are _________ sources

external

Data warehouse typically store large amounts of ________ data

internal

Source Evaluation for internal sources is based off:

compatibility (w/ question)

timelessness (less affected by time = better)

recentness

quality

authors

Decision Variable

- A quantifiable characteristic, attribute, or outcome on which a decision is made

- We can have one or more of these

Decision Rule

The criteria for judging the attractiveness of 2 or more alternatives when using a decision variable

Ex Post Facto Evaluation

refers to determining the value of research AFTER the research has been conducted

If a decision could have been made without research, the value of the research is _________

negative

Task Budget

- Dollars usually drawn from a discretionary reserve of an organization/budget of in-house operation.

Functional Area Budget

- Dollars drawn from a portion of the manager's business unit's operational fund.

- Manager can spend the budgeted amount how they wish

Ethical Issues in Stage 1

- Regardless of organizational structure, the manager has the most control over stage 1 so it is important that it is carried out ethically

DMP stands for

Data Management Platform

First party data is another term for ______ ______

primary data

Stratification

- Separates data so that it is easier to see and make sense of

- This can be used when there are multiple sets of data overlapping on a graph

- Ex: data may be categorized by color

(T/F) It is easier to implement stratification after data has been collected.

False

It is much easier to use stratification before collecting data

Check Sheet

- aka defect concentration diagram

- for visual refer to module 5 (ch4 textbook)

Instrument Disposition

the process(es) by which the measurement instrument is returned to the researcher

Data Entry

converts data gathered by secondary or primary methods into a medium for data viewing and manipulation

Data Fields

- Represent variables

- Ex: response to question

- Can contain numbers, words, images, etc.

Data Records

- Represent rows in a data file

- A set of data for one case (participant)

- All your answers in a survey would represent your data record

Data File

- ALL the responses by ALL participants in a study

Database

- Stores files

- Basically, a set of all records

3 Types of Data Units:

context units

sampling units

recording units

Context Units

refer to the content that directly relates to the objective of research

Sampling Units

- refer to what text elements the researcher will code

- Ex: words, phrases, sentences

Recording Units

refer to the ideas, topics, or themes that are embedded in the text

- there are 4 types:

1. syntactical

2. referential

3. propositional

4. thematic

Syntactical Units

- The exact, author defined, words and phrases

- We are interested in the meaning of the words and phrases in context

Referential Units

- Describe objects by using words or phrases

- Used to make inferences about opinions of participants

Propositional Units

- Are assertions about an object

- Ex: you will save $15 if you do this....

- "save" is the propositional unit

Thematic Units

- Topics contained within (and across) text

- These represent higher level abstraction

- Look for things that repeat

Content Analysis Software uses statistical algorithms to permit:

stemming

aliasing

excluding

Stemming

- searches for all the forms of a key word (stem)

- Ex: searching, searches, search, searched

Aliasing

- Searches for synonyms

- Ex: wise, smart, intelligent

Excluding

- Filters out trivial words

- Ex: be, is, the, of

Editing refers to

- Process for verifying that variables are used and entered correctly

- Ensures completeness and accuracy

"Don't Know" Responses

are difficult to interpret because they could mean a variety of things

Missing data may be caused by

- participants intentionally or accidentally skipping

- research error

- malfunctioning software

- changes in design

- participants dropping out of the study

MCAR

- refers to data missing COMPLETELY at random, not correlated to another variable

MAR

- refers to data missing at random, not correlated

NMAR

- refers to data missing but not at random

3 Methods to salvage missing data:

Listwise Deletion

Pairwise Deletion

Predictive Replacement

Listwise Deletion

An entire case with a missing variable (data) is deleted from all analyses that involve that variable

Pairwise Deletion

- We replace the missing data with an estimate

- May be replaced by the mean or another measure

Predictive Replacement

- Replaces missing data based off the value of another variable

Sources of bias:

- Asking misleading questions

- Forged data

- Data entry

Data Validation

- A process for verifying research protocols that avoid errors and confirm data is real

- Identifies fake or inaccurate data

Exploratory Data Analysis (EDA)

explores and reduces data using statistical methods

Confirmatory Data Analysis (CDA)

- Required when attempting to prove causation

- has something to do with significance and confidence but not really sure

Frequency Table

- tells us how many times a value occurs

- can also be in a graphical form such as bar chart or pie chart

Histogram

- groups variables into intervals

- shows a distribution

Stem and Leaf Display

shows quantitative data values in a way that sketches the distribution of the data

- a numerical version of a histogram

***(If you turn the picture sideways, it is a histogram)***

Pareto Diagram

A histogram, ordered by frequency of occurrence, that shows how many results were generated by each identified cause.

Boxplot

Boxplots display a 5 number summary which is:

- Median

- Upper quartile

- Lower quartile

- Largest observation (max)

- Smallest observation (min)

Resistant Statistics

- stats not affected by outliers

- Ex: median, upper & lower quartile

An outlier is defined as a data point that exceeds interquartile range by _______ times

1.5

Mapping

a map...

Cross Tabulation

a method of analyzing data that lets the analyst look at the responses to one question in relation to the responses to one or more other questions

Control Variable

- A variable introduced to help interpret the relationship between variables

Blue cells are _____ cells

data

Yellow cells are _______ cells

changing

Output cells

show the quantities calculated from changing cells

Orange cells are _____ cells

objective

Feasible Region

feasible region is the area that satisfies every constraint

Constraint Boundary Line

- the line of the constraint boundary equation

Constraint Boundary Equation

- the equation that defines the constraint line

- Ex: 3D + 2W = 18

3D + 2W = 18

What does D and W equal?

D:

3D ≤ 18

D ≤ 6

W:

2W ≤ 18

W ≤ 9

(T/F) Constraints are represented by _________ cells

output

5 Types of Linear Programming Problems

- Resource-Allocation Problems

- Cost-Benefit Tradeoff Problem

- Mixed Problems

- Transportation Problems

- Assignment Problem

Resource Allocation Problems

- involves the allocation of an organization's resources

- a resource can be anything from cash to production materials

- has a resource constraint

Resource Constraint (≤)

Amount of resource used ≤ Amount resource available

- this means we can only use up to a certain amount of resources until we run out

****the ≤ tells you that it is a resource constraint

Cost Benefit Tradeoff Problem

- different from resource allocation because we determine our goal and use as few resources as possible to achieve it rather than having a predetermined amount of resources

- has benefit constraint(s)

Benefit Constraint (≥)

Level achieved ≥ Minimum acceptable level

- this means we need to be greater than or equal to the minimum acceptable level that was determined

****the ≥ tells you it is a benefit constraint

Mixed Problem

- a linear programming problem that includes at least 2 of the 3 types of functional constraints (resource, benefit, fixed-req)

Fixed Requirement Constraint (=)

Amount provided = Required amount

*** the = sign is the identifying factor

Transportation Problem

- involves finding a shipping plan that minimizes total cost of transporting products to customers

- normally need 2 types of constraints

***look for words such as "shipping" or "transport"

Assignment Problem

- involves assigning people to tasks to minimize the total cost of the tasks

- Which people are assigned to which task?

****look for words like "assign" "roles"

Characteristics of any linear programming model

Linearity

Finiteness

Non-negativity

Linearity refers to

the fact that the relationship between 2 or more variables should be linear

Finiteness refers to

the fact that the solution will not be optimal if there is an infinite number of factors

Non-Negativity refers to

the fact that the variable should be positive or 0, not negative

Model Enrichment

The process of using experience with a model to identify and add important details that will provide a better representation of the real problem