Marketing research

BA035IU
• Lecturer: Dr. Alang Thớ
• Email: alangtho@hcmiu.edu.vn
• Workshops: Monday, 8.00– 10.30
• Consultation appointment: Skype
This week 9

• Summarizing key points from previous weeks

• Preparing data for analysis
• Strategy for data analysis
• Factors influencing the choice of statistical techniques
• An overview of statistical techniques
• Summary of this week’s lesson
• Preparing for the midterm examination
Learning objectives
Become Become familiar with the fundamental concepts of data analysis

Understand Understand the need for preliminary data preparation techniques

Describe Describe the various statistical techniques for adjusting the data

Discuss Discuss the significance of data tabulation

Identify Identify influencing factors of the selection of a data analysis strategy

Understand Understand the statistical techniques available for data analysis

Explain Explain the various multivariate techniques

Summarizing key point from previous weeks
Data analysis

A set of methods and techniques Helps avoid erroneous judgements Can constructively influence the
used to obtain information and and conclusions research objectives and the research
insights from data design
Preparing the data for analysis

DATA EDITING CODING STATISTICALLY ADJUSTING

THE DATA
 Data editing

Conducted in the field

Identifies omissions, by interviewer and
ambiguities, and field supervisor and by
errors in responses the analyst prior to
data analysis
 Problems identified with data editing

Interviewer
Omissions Ambiguity
Error

Lack of Ineligible
Inconsistencies
Cooperation Respondent
 Coding

Coding closed-ended
Open-ended questions
questions involves
(qualitative data) are
specifying how the
difficult to code
responses are to be entered

Lengthy list of
possible
responses is
generated
Examples of qualitative data coding (cont’d)
NVIVO used to code qualitative data
Examples of qualitative data coding
Examples of quantitative data coding
Using SPSS to coding
Statistically adjusting the
data: Weighting
• Each response is assigned a number according to a
pre-specified rule
• Makes sample data more representative of target
population on specific characteristics
• Modifies number of cases in the sample that
possess certain characteristics
• Adjusts the sample so that greater importance is
attached to respondents with certain
characteristics
An example of weighting adjusting

• If a study is conducted to determine the market potential of a new sports

drink, the researcher might want to attach greater weight to the opinions
of the younger people in the market, who will be the heavy users of the
product.
• This could be accomplished by assigning weights of 2.0 to persons in the
sample who are under age 30 and 1.0 to respondents over 30.
Statistically adjusting the data:
Variable Re-specification
• Existing data is modified to create new variables
• Large number of variables collapsed into fewer
variables
• Creates variables that are consistent with study
objectives
• Dummy variables are used (binary, dichotomous,
instrumental, quantitative variables)
• Use (d-1) dummy variables to specify (d) levels of
qualitative variable
An example of variable re-specification

• Suppose the original variable represented the reasons for purchasing a car,
with 10 response categories. These might be collapsed into 4 categories:
performance, price, appearance, and service.
• Respecification also includes taking the ratio of 2 variables to create a new
variables, taking square roots and log transformations, and using dummy
variable
Statistically adjusting the
data: scale transformation
• Scale values are manipulated to ensure
comparability with other scales
• Standardization allows the researcher to compare
variables that have been measured using different
types of scales
• Variables are forced to have a mean of zero and a
standard deviation of one
• Can be done only on interval or ratio scaled data
An example of scale transformation

• If sales are measured in actual dollars, and price in cent, then the actual
value of the variance for the sales variable will be higher compared to
price, because of the units of measurement.
• To compare the variances, both variables can be brought down to a
common unit of measurement
Strategy for data analysis

SIMPLE TABULATION TABULATION: FREQUENCY TABULATION: DESCRIPTIVE CROSS-TABULATION

DISTRIBUTION STATISTICS
Simple tabulation

Consists of counting the number of Use of Simple Tabulation

cases that fall into various categories
Determine empirical distribution (frequency
distribution) of the variable in question
Calculate summary statistics, particularly the
mean or percentages
Aid in "data cleaning" aspects
Tabulation: Frequency distribution

Reports the number of Shows number of

Organizes data into
responses that each observations that fall into
classes or groups of values
question received each class

Can be illustrated simply Response categories may Should result in categories

as a number or as a be combined for many with worthwhile number
percentage or histogram questions of respondents
An example of frequency table and histogram
An example
of frequency
distribution
(percentage)
Tabulation: Descriptive statistics

• Statistics normally associated with a frequency distribution to help

summarize information in the frequency table. It includes:
• Measures of central tendency (mean, median and mode)
• Measures of dispersion (range, standard deviation, and coefficient of
variation)
• Measures of shape (skewness and kurtosis)
Using SPSS to analyze frequency and statistics
Analysis for Various Population Subgroups

• Differences between means or percentages of two subgroup responses can

provide insights
• Difference between means is concerned with the association between two
questions
• Question upon which means are based are intervally scaled
Cross tabulation

• Statistical analysis technique to study the relationships among and

between variables
• Sample is divided to learn how the dependent variable varies from
subgroup to subgroup
• Frequency distribution for each subgroup is compared to the frequency
distribution for the total sample
• The two variables that are analyzed must be nominally scaled
An example of cross tabulation analysis
Using SPAA to analyze cross tabulation
Factors influencing the choice of statistical
techniques

Assumptions
Research
Type of data underlying the
design
test statistics
Type of data
Classification of data
Nominal scaling is
involves nominal, Both median and mode
restricted to the mode
ordinal, interval and can be used for ordinal
as the only measure of
ratio scales of scale
central tendency
measurement

Mean, median and

Non-parametric tests mode can all be used to
can only be run on measure central
ordinal data tendency for interval
and ratio scaled data
Research design

Number of Number of Control exercised

Dependency of
observations per groups being over variable of
observations
object analyzed interest
 If assumptions on which a statistical test is based
are violated, the test will provide meaningless
results
Assumptions
underlying
the test
statistics For example: the assumption of a two-sample t-
test are as follows:
The characteristics of
The samples are interest in each The two population have
independent population have normal equal variances
distribution
Overview of statistical technique

Why Use
Univariate Multivariate
Multivariate
Techniques Techniques
Analysis?
 Univariate techniques
Appropriate when there is a single measurement of each of the 'n' sample objects or there are
several measurements of each of the `n' observations but each variable is analyzed in isolation

Nonmetric: Measured on nominal or ordinal scale

Metric: Measured on interval or ratio scale

Determine whether single or multiple samples are involved

For multiple samples, choice of statistical test depends on whether the samples are independent or
dependent
Multivariate Techniques

A collection of procedures for analyzing association between two

or more sets of measurements that have been made on each
object in one or more samples of objects

Dependence or interdependence techniques

 Multivariate Techniques (cont’d)

Dependence techniques Independence techniques

• One or more variables can be • Whole set of interdependent

identified as dependent relationships is examined
variables and the remaining as • Further classified as having
independent variables focus on variable or objects
• Choice of dependence
technique depends on the
number of dependent
variables involved in analysis
 Why Use Multivariate Analysis?

To understand
To improve the ability
To group variables or relationships between
to predict variables
people or objects variables (such as
(such as usage)
advertising and sales)
Summary

• The first phase of data analysis includes editing, coding, and statistically preparing
the data for analysis. There are several techniques used to statistically adjust the
data such as weighting, variable re-specification, and scale transformation
• A frequency distribution provides complete information and lead to decisions to
combine response categories. Descriptive statistics such as mean, median, mode,
standard deviation, and variance can be obtained from these distributions.
• The usual step is to tabulate questions among subgroups, and this involves two of
the questions from the questionnaire. Guidelines are developed for selecting the
appropriate statistical techniques. A discussion on the overview of statistical
analysis is provided.
 Preparing for the midterm examination

When: 8:00- 11:00, 27/04/2020 How: Submitting through Type of question: individual
(check school’s exam schedules) blackboard before the deadline essay presenting as report
(11:00am 27/04/2020) format