0% found this document useful (0 votes)
12 views18 pages

Operations Research Course Overview

This document outlines the course details for an Operations Research course taught at Cairo University. It includes the course outline, resources, and key topics that will be covered over the duration of the course. The course aims to teach students a quantitative approach to business management. It will cover business applications of various quantitative techniques including parametric and non-parametric methods, univariate and multivariate analysis, and how to interpret and apply quantitative outputs to business decisions. Students will learn through lectures, textbooks, assignments, and presentations. Key topics that will be discussed include the philosophy of management science, categories of quantitative techniques, and the appropriate use of parametric vs non-parametric methods based on the measurement scale of business data.

Uploaded by

Randa ElAbd
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views18 pages

Operations Research Course Overview

This document outlines the course details for an Operations Research course taught at Cairo University. It includes the course outline, resources, and key topics that will be covered over the duration of the course. The course aims to teach students a quantitative approach to business management. It will cover business applications of various quantitative techniques including parametric and non-parametric methods, univariate and multivariate analysis, and how to interpret and apply quantitative outputs to business decisions. Students will learn through lectures, textbooks, assignments, and presentations. Key topics that will be discussed include the philosophy of management science, categories of quantitative techniques, and the appropriate use of parametric vs non-parametric methods based on the measurement scale of business data.

Uploaded by

Randa ElAbd
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Operations Research 1

Cairo University

Faculty of English Commerce

Business Administration Major

Year 2016

Operations Research

Professor/ Wael Kortam


Operations Research 2

Course Outline

1- Philosophy of quantitative approach to business & management.

2- Business applications of parametric & non-parametric quantitative techniques.

3- Business applications of inferential VS derivative quantitative techniques.

4- Business applications of uni-varities VS multi-varieties.

5- Interpreting, reporting, and incorporating quantitative outputs in business

decisions & policies.

Course Resources

1- Lecture.

2- Textbook.

3- All digital library & discussion resources.

4- Assignments & presentations.

5- EGO.
Operations Research 3

1) A philosophy of quantitative approach to management:

Objective of management science:

1- Avoidance of type 1 error (rejecting the null hypothesis while it is true).

2- Avoidance of type 2 error (accepting the null hypothesis while it is false).

Null hypothesis: scientific assumption that theres no existence for a business

phenomenon variable relationship of impact.

Example:

Theres no relationship between TV Ads & market share.

Answer:

1- Reject the null hypothesis so invest L.E. 5M Idle investment Type 1 error.

2- Accept the null hypothesis no spending on TV Ads Lost a business

opportunity Type 2 error.

Phases of management science (Steps to ensure not to make any time of

error):

1- Selection of appropriate quantitative techniques. There should be a cooperation

between business analyst (they know business objectives, they know all previous

research phases) & statistical expert (aware of two techniques & understands

assumptions).

2- Administration/computation of selected quantitative techniques methodologies

(quantitative experts).

3- Interpretation of quantitative analysis output (what those figures are saying

business wise) (business analyst/decision maker).


Operations Research 4

2) Categories of management science (quantitative techniques) in

business techniques

1- Categorization One:

Depending on kind of data:

Metric Data (Age) Non-Metric Data (Gender)

Two kinds of scales: interval/ratio Coming from nominal or ordinal

scales. scale (ID)

For them we use metric techniques Non-parametric techniques (not

(correlation + regression) strong as parametric)

Most common (Chi square x2

discriminant analysis)

Frequency analysis (pie charts)

2- Categorization two:

Depends on purpose of analysis (how customers are satisfied).

If you ask 400 people & generalize on Egyptian people = Inferential.

Just the 400 = Descriptive.

Purpose of analysis:

1- Descriptive:

o Central tendency age of certain group mean.

o Dispersion standard deviation tells how different you are


Operations Research 5

2- Inferential:

o Correlation/T-Test/F-Test/Factor Analysis generalize what you find in a

sample to a whole population.

3- Category Three: Number of variables.

o Uni-variance techniques (test each variable affecting you to buy a car)

statistical tests of significance (x2-T-Z-F-KW Tests)

o Multi-variance techniques multi variance analysis (all variables tested

simultaneously, not one by one) multiple regression/multiple

discriminance/factor analysis/path analysis/cluster analysis/structural

equation modeling.
Operations Research 6

3) Business application of parametric VS non-parametric quantitative

techniques

Measurement & scale of business phenomenon and variables.

The purpose of measurement & scale: To assign numbers to business phenomena

and variables in a way that allows safe generalization.

Scaling: is developing a continuum that reflects the maximum and minimum

values of a business phenomenon or variable and all other values in range in

between/it is an application of selected measurement theory.

What are the kinds of business data?

o Facts = I am Egyptian.

o Opinions = I like dogs.

Facts:

o Quantitative = Ratio scaling.

o Qualitative = Nominal scaling.

Opinions:

o Attitude/Absolute (about one thing) = Interval scaling.

o Comparative (preference) = Ordinal or ranking scaling.


Operations Research 7

How can we judge the measurement & scaling of quality?

Validity Reliability

Is the ability of the measurement scale Is the ability of the measurement scale

to express the true essence of the to reflect consistently the differences

business phenomenon or variable among sample units or across time.

under research in an exclusive and

inclusive manner.

We measure it by factor analysis. We measure it by Crombach alpha.


Operations Research 8

4) Kinds of Scales

1- Nominal Scaling:

o Deals with qualitative facts.

o Mutually exclusive = Christian or Muslim.

o Produces categorical data - non-metric = numbers are symbols not values.

Number are only for identification not description - frequency analysis =

70% married (non-parametric/very weak). E.g. Code of Cairo is (2),

Alexandria (3) = numbers here doesnt mean one is higher than the other.

Example:

If we have a bank which has four branches; Heliopolis, Maadi, Alexandria, Mohandseen.

Please indicate your home branch where you opened your account. (By definition youll

check one answer only.

1) Heliopolis 40%

2) Maadi 20%

3) Mohandseen 10%

4) Alexandria 30%

*If you give a code (1) or (2), these are only numbers that identify them.
Operations Research 9

Chart Title
0.45

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0
Heliopolis Maadi Mohandseen Alexandria

Column3 Column1 Column2

Example:

Please indicate which of the following branches youve dealt with:

1) Heliopolis.

2) Maadi.

3) Mohandseen.

4) Alexandria.

This is a nominal scale as they are four questions not one.

2- Ordinal/Ranking Scales:

o Deal with comparative opinions which are preferences = numbers are

stronger than others; numbers mean something.

o Numbers reflect values in the form of power of preference.

o Problem in in unequal intervals among ranks categorical data non-

metric allows only frequency analysis non-parametric.


Operations Research 10

Example:

Please rank the following cities as tourists destinations according to your preference.

- (3) New York

- (1) London

- (2) Paris

- (4) Rome

Lets say 20% of people around the world chose New York as number one. 30% chose

New York as number two. 40% chose New York as number three. 10% chose New York

as number four.

Chart Title
0.45

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0
New York-1 New York-2 New York-3 New York-4

Column1 Column2 Column3


Operations Research 11

3- Interval Scaling

o We look at absolute opinions which we call attitudes & it gives equal

intervals among ranks which means that you have metric data so you can

use all kinds of parametric techniques ( Very strong). Not an interval

scales e.g. Liker Scale.

Example:

Please indicate to what extent you agree with the following statements:

Statements Strongly Agree (4) Neutral Disagree Strongly

Agree (5) (3) (2) disagree

(1)

1) After sale

service is

very

important

to buy a

certain

brand.

2) TV Ads

influence

my

decision

to buy a
Operations Research 12

certain

brand.

We should say Please because everyone is doing you a favor even if youre

paying money.

We remove words & put numbers (Strongly agree (5), strongly disagree (1)) = this

is an interval scale.

4- Ratio Scale:

o Measures quantitative facts (very strong) e.g. Money, number of children.

o Metric data = you can use all forms of parametric techniques.

Example:

How much do you spend on mobile services every month = you have to give numbers.

Why we do this (give a range for the salary):

o Please indicate you monthly income (500/500-1000/1000-3000/3000-

5000/5000-10, 000/More than 10, 000) = we give a range because it is a

confidential data that no one likes to tell or talk about it.

Example:

If I want to know the impacts of level of education on monthly spending on fast food,

favorite chain(s) and most important quality standards. What are the scales that should

be used to collect data for this research?

1) Level of education = qualitative fact = nominal scale.

2) Monthly spending on fast food = quantitative fact = ratio.


Operations Research 13

3) Favorite chains = preference (comparative) = ordinal or ranking scale.

4) Quality standards = absolute (attitude) = interval scale.

Validity = people understand what you are asking them.

Make sure people really mean what they say!

Factor Analysis

Judges the validity of measurements.

Factor analysis is a parametric technique + inferential (you can use it to

generalize) + multi variance technique (deals with many variables).

Test of construct validity.

Data reduction technique (what is the real underlying dimensions of my story?).

Factor analysis: calculate all possible mathematical combinations of correlations

through an infinite algorithm called VARI MAX rotation (Max. Number of

rotations; very huge) ((

Factor analysis brings the highest evidence correlations classified into number of

factors found.

o Rule 1 = all the values range from 0.00 to 1 (No negative in FA)

o Rule 2 = Consider the values of 0.5 or more only with exceptions!

Example:

The following table represents a factor analysis of perceived quality of marketing mix for

an Egyptian fast food chain. This table result from a questionnaire based survey on

marketing mix variables that collected data from a representative sample of 1, 200

sampling units using a five point liker interval scale.


Operations Research 14

Factors \ Factor One Factor Two Factor Factor Four

Variables Three

1- Food quality 0.91 0.33 0.07 0.03

2- Menu variety 0.67 0.49 0.18 0.31

3- Packaging 0.75 0.25 0.29 0.12

4- Value for 0.12 0.88 0.31 0.09

money

5- Price 0.23 0.71 0.16 0.41

discounts

6- Geographical 0.42 0.12 0.86 0.22

dispersion of

branches

7- Layout of 0.20 0.16 0.56 0.26

branches

8- TV Ads 0.18 0.36 0.26 0.77

(quality)

9- Outdoors 0.32 0.22 0.34 0.66

10- Staff 0.27 0.13 0.17 0.91

courtesy

11- Eigen values 3.25 2.20 1.80 2.75

12- % of 30% 19% 18% 20%

variance Total: 87%


Operations Research 15

Required:

1- Use factor analysis to judge the construct validity of used measurement scales for

perceived quality of marketing mix in the fast food chain.

2- Interpret the four factors & their relative importance as underlying dimensions of

perceived quality of fast food chain.

Answer:

Requirement (1) Mark all values above 0.5

Factor One:

Quality of food = 0.91 (highest)

Packaging = 0.75

Menu variety = 0.67

Theres a high construct validity of measuring the product factor.

Factor one can be labelled product with the underlying dimension of:

1- Quality of food.

2- Packaging

3- Menu variety

Respectively based on factor loadings (number).

Factor Two:

Value for money = 0.88

Price discounts = 0.71


Operations Research 16

Theres a high degree of construct validity for measuring the price factor

Factor two can be labelled price with the underlying dimensions of:

1- Value for money.

2- Price discounts.

Respectively based on factor loadings.

Factor Three:

Geographical coverage of branches = 0.86

Branches layout = 0.56

Theres a high degree of construct validity for measuring the place factor.

Factor three can be labelled place with the underlying dimensions of:

1- Geographical coverage of branches.

2- Branches layout.

Respectively based on factor loadings.

Factor Four:

Staff courtesy = 0.91

TV Ads = 0.77

Outdoors = 0.66

Theres a high degree of construct validity for measuring the promotion factor.

Factor four can be labelled promotion with the underlying dimensions of:
Operations Research 17

1- Staff courtesy

2- TV Ads

3- Outdoors.

Respectively based on factor loadings.

Requirement 2:

The most influential underlying dimension of perceived quality of marketing mix

for the fast food chain is the factor of product (e.g. value = 3.25), followed by the

factor of promotion (e.g. value = 2.75), then the factor of price (e.g. value = 2.2),

and finally the factor of place (e.g. value = 1.8) respectively.

According to factor loading:

o The product factor is mainly composed of underlying dimensions of:

Food quality

Packaging

Menu variety (in terms of relative importance)

o The promotion factor is mainly composed of underlying dimensions of:

Staff courtesy

TV Ads

Outdoors (in terms of relative importance)

o The price factor is mainly composed of underlying dimensions of:

Value for money

Price discounts (in terms of relative importance)


Operations Research 18

o The place factor is mainly composed of underlying dimensions of:

Geographical coverage of branches

Branches layout (in terms of relative importance)

The accumulated % of variance indicates that the four factors and their

underlying dimensions explain & influence 87% of perceived quality of marketing

mix of the fast food chain.

The most influential factor in explaining & influencing the perceived quality of

marketing mix for the fast food chain is the:

o Product 30%

o Promotion 20%

o Price 19%

o Place 18%

You might also like