Market Research
Data Analysis Methods
What is Data Analysis ?
• A set of methods and techniques used to
obtain information and insights from data.
• Helps to avoid erroneous judgments and
conclusions.
• Can constructively influence the research
objectives and the research design.
• Interpret the data using the advanced
statistical analyses like correlation, regression,
cluster analysis, factor analysis etc.
Importance of Data Analysis ?
• Convert the mountain of data into meaningful
entities.
• Reduce data into meaningful results.
• Drive decisions with better analysis.
• Turn data into opportunities.
Preparing the data for Analysis
1. Data Editing / checking.
Editing process is conducted by the data
analyst prior to the analysis process.
Editing process helps to identify the data
omissions, interviewer errors, errors in the
response, uncertainty in answering, ineligible
respondents and logical flow/routing of the
questionnaire.
Preparing the data for Analysis
2. Coding
Coding refers to the process of assigning
numerals or other symbols to answers so that
responses can be put into a limited number of
categories or classes.
Coding is necessary for efficient analysis and
through it several replies are reduced to a small
number of classes which contains the critical
information's required for the analysis.
Preparing the data for Analysis
2. Coding (Cont…)
Open ended questions are difficult to code.
A lengthy list of possible responses are generated for
this purpose and then the codes are selected from this list for
the appropriate answers.
Each response is assigned with a number (column number)
for the convenience of analysis.
Preparing the data for Analysis
3. Statistically Adjusting the data for analysis.
Weighting
Is a data projection technique used to extrapolate
the sample data representative of the target
population/Universe.
A weighting factor is derived for this purpose
[population (of the stratum)/sample (of the Stratum)] and
then it is applied to the sample data in order to extrapolate
it.
Tabulation
4. Tabulation is the process of summarizing raw data and
displaying the same in a compact form(i.e., in the
form of statistical tables) for further analysis.
a. Simple Tabulation.
Gives information about one or more groups of
independent questions.
b. Complex Tabulation.
Gives information about two or more inter-related
characteristics of data.
Tabulation (Contd..)
General Principles.
1. Every table should have a clear and concise title placed
just above the body of the table.
2. Every table should be given a distinct table number to
facilitate easy reference.
3. The column heading (breaks/captions) and the row
Headings(stubs) should be brief and clear.
4. Explanatory footnotes.
5. Units of measurement/base should be indicated under
heading.
6. Source of data can be indicated just below the table.
7. The columns can be numbered for easy reference .
Tabulation (Contd..)
General Principles(Contd…).
8. Base title should match with the base value.
9. Percentages can be kept close to the data.
10.All column figures, decimals and (+) or (-) signs must
be properly aligned.
11.Abbreviations should be avoided to the extent possible
and ditto marks should not be used in the table.
12. Miscellaneous and exceptional(DK/NS) items should be
kept at the bottom of the table.
Cross Tabulation Method
• 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
Nominal Scale
One of the Measurement Scales , which simply assigns
number symbols to events in order to label them.
These numbers are just convenient labels for the
particular class of events and there is no quantitative
value for them.
Chi-Square test is the most common test of statistical
significance that can be applied with nominally scaled
numbers.
Ex : Assignment of numbers to basketball players in
order to identify them.
Measurement Scales
• Nominal Scale
Nominal scaling is restricted to the mode as the only
measure
of central
• Ordinal Scaletendency
Rank orders represents ordinal scale and it’s frequently used
in qualitative research. Both median and mode can be used for
ordinal scale. Non-parametric tests can only be run on ordinal data.
• Interval scale
• Ratio scale
Mean, median and mode can all be used to measure
central tendency for interval and ratio scaled data
Data Analysis
Qualitative data lends itself to non-parametric tests and
Quantitative data lends itself to parametric tests in statistics
Data
Qualitative Quantitative
Nominal Ordinal Interval Ratio
variables variables variables variables
Non-Parametric Tests Parametric Tests
(Chi-Square test) t-tests and Analysis
of Variance (ANOVA)
Types of Analysis
1. Univariate analysis.
- Analysis involving a single variable
2. Bivariate analysis.
- Analysis involving two variables
3. Multi-variate analysis.
- Analysis involving more than two variables
Univariate Analysis
If the analysis involves only a single
variable, then it is called as a Univariate analysis.
Useful when each variables are analyzed separately.
Following are the examples of univariate analysis.
1. Mean
2. Variance
3. Standard deviation
4. Median
5. Mode
Etc…
Bi-variate Analysis
If the analysis involves two variables,
then it is called as a Bi-variate analysis.
Following are the examples of bi-variate analysis.
1. Linear/Simple regression
2. Correlation coefficient
3. Chi-square
4. Covariance
Etc…
Multi-variate Analysis
If the analysis involves more than two
variables, then it is called as a multi-variate
analysis.
Following are the examples of Multi-variate analysis.
1. Multiple regression
2. Correspondence analysis
3. Factor analysis
4. Cluster analysis
5. Discriminant analysis
Etc…
Why Use Multivariate Analysis ?
To group variables or people or objects
To improve the ability to predict variables
(such as usage)
To understand relationships between variables
(such as advertising and sales)
