Separation of Lanthanides/ Lanthanides and Actinides
Data analysis and interpretation
1. Session 2
Data Analysis & Interpretation
(Social Science)
Welcome to Teachers Mitraa
Online Talk on
Basics of writing research papers
2. Over View of the Session
• Part 1 – Qualitative Analysis
Data Analysis of Review of Literature
Data Analysis of text in research questionnaire
(Suggestions/Opinions)
• Part 2 – Quantitative Analysis
Descriptive Statistics
Scales of Data Measurement
Test Chart and Some Statistical Analysis for Parametric
Tests (Inferential Statistics)
Chi-square Test
Correlation and Regression
Independent Sample T-Test
One Way ANOVA
3. Key takeaways
By the end of the session
• To understand the various forms of data or data sets
(Qualitative/Quantitative, Data measurement types)
• To understand the Right Tests that can be computed for
right combination of data (MsExcel and SPSS)
(Inferential Statistics using Parametric Test)
• To understand the nomenclatures for setting hypothesis
• To understand the key aspects to be considered while
writing interpretation
4. Key Takeaway - 1
• To understand the various forms of data or data
set
(Qualitative/Quantitative) (Data measurement types)
5. Types of Data
Qualitative Data
Unstructured
Quantitative Data
Structured
•Literature (Research Papers)
•Text (Opinion / Suggestions)
•Content (Social Media)
• Categorical data
• Metric data
Descriptive Statistics
Text Analysis
Word Cloud
List of words
Descriptive Statistic
Inferential Statistics
7. Qualitative Data
– This data type is non-numerical in nature and it does
not refer to the aspects that can be numbered or
measured.
– Example: Text / Script / Names… (Content Analysis)
– Forms of Collection: Research Papers, Questionnaires,
Interview or Observation technique (Writing), Video
recording, Voice recording.
Analytical Tools: MsExcel, NVIVO, (https://www.wordclouds.com/)
Type of Analysis: Descriptive Analysis, SENTIMENT / TEXT ANALYSIS - List the
highest repeating words, Word Cloud
9. Review of Literature
Review of
literature in
context to Indian
Scenario
Review of
literature in
context to Global
Scenario
Total
India 16 NA 16
Developing Countries NA 14 14
Developed Countries NA 11 11
Global prospective NA 21 21
Total 16 46 62
Descriptive Statistics for Review of Literature
10. Example of Qualitative Data
Sl. no Words Count
1 Supply 11676
2 Chain 10820
3 Management 8410
4 Environmental 6249
5 Green 6212
6 Performance 4299
7 Production 2688
8 Product 2725
9 Practices 2485
10 Business 2474
11 Supplier 4439
12 Cost 2123
13 Time 2015
14 Customer 1890
15 Sustainability 4944
16 Process 1669
17 Manufacturing 1703
18 Distribution 998
13. • Quantitative Data
– Quantitative data are measures of values or counts and are
expressed as numbers. Quantitative data are data about
numeric variables.
– Example: Sales, Revenue, Expenses, Number of people
visit a outlet, Temperature, Height, Weight, Age,
Types of Data
1. Major Classification of Statistics
2. Parametric and Non-Parametric Tests
3. Scales of Data Measurement
14. 1. Major Classification of Statistics
• Descriptive Statistics
– Describes the nature of
data
Measures of Centre of Tendency
Measures of Dispersion
Normality Check
Reliability Check
Data Duplication
Missing Data Check
Add flavors through Plots / Charts
• Inferential Statistics
₋ Generalizing the results
derived from the
sample to entire set of
Population
Toothpaste Consumers in
Karnataka - 5.2 Crore
People
(Population)
Sample
15. Few Takeaways
• For Normality Check the value of Skewness
and Kurtosis needs to be between:
-1.96 and +1.96
• For Reliability Check the Cronbach's Alpha
value needs to be between:
0.70 and 0.95
16. 2. Parametric and Non-Parametric Tests
• Parametric Tests
– The data needs to be normally distributed (Independent Metric
Variable)
– The data needs to be collated through Random Sampling
Technique (Probability Sampling)
• Non-Parametric Tests
– The data need not be normally distributed (Independent Metric
Variable)
– The data need not be collated through Random Sampling
Technique (Probability Sampling)
17. 3. Scales of Data Measurement
• Nominal Data – Names or numbers assigned to
identify the data unit.
– Example: Jersey number for football players, Roll numbers,
Gender, Color (Group), Class room number, Country, Religion.
These are also called as: Categorical or Grouped data
• Ordinal Data – The data set which follows a
particular order and the difference between the
two numbers or groups is not exact
⁃ Example: Education Qualification, Ranking for range of
products, Social class, likert Scale (Emotions).
18. • Interval (Scale) Data – The difference between
two numbers is exact and there is no absolute
zero. (Zero has some meaning)
– Example: Number of customer footfalls, Marks of
students, Age, Time, Year, Temperature.
2. Scales of Data Measurement
These are also called as:
Continuous or Interval data or Metric
• Ratio Data – The difference between two
numbers is exact and there is a absolute zero.
– Example: Income, Height, Weight, Sales,
unemployment rate.
19. Same Data as Categorical and Metric
Age
Age of the
Respondent
25
36
45
20
15
40
42
38
52
65
28
54
Ratio
Metric
Age of the
Respondent
20 to 25
26 to 30
31 to 35
36 to 40
41 to 45
Above 46
Ordinal
Categorical
Sales
Sales of Product X
251
368
456
201
159
412
426
385
521
658
286
548
Interval
Metric
Sales of Product X
200 to 250
251 to 300
301 to 350
351 to 400
401 to 450
Above 451
Ordinal
Categorical
20.
21.
22. Very Vital Terms
• Each individual statement of the “Likert scale”
is called as a “Item”
(Parameter/Factor/Element/Antecedent/Attribute)
• When these items are added it is called as a
“Construct” or “Dimension”
Item is a Categorical data
Construct is a Metric data
23. Key Takeaway – 2,3 &4
• To understand the Right Tests that can be
computed for right combination of data (MsExcel
and SPSS)
(Inferential Statistics using Parametric)
• To understand the nomenclatures for setting
hypothesis
• To understand the key aspects to be considered
while writing interpretation
24. Chart for Parametric Tests
Dependent
Variable
Dependent Variable
Categorical Metric
Independent
Variable
Categorical Chi Square
1) Independent Sample T-test
(2 Groups) Example:
(High/Low), (Girl/Boy)
2) ANOVA (More than two
groups)
Independent
Variable
Metric
Logistic
Regression
1) Paired T-Test
2) Correlation
or
Regression
26. Prerequisite for Chi Square Test
• Both the variable need to be categorical
• Hypothesis
– H0: There is no association between X and Y (Both are
independent)
– H1: There is association between X and Y (Both are
independent)
• The P-Value considered is 0.05 level of significance
• Reject null if P-Value is less than 0.05
27. Example of Category on Category
1. Gender of respondents on Number of brands
used
Hypothesis:
– H0: There is no association between Gender of
respondents and Number of brands used
– H1: There is association Gender of respondents
and Number of brands used
Test used is – Chi square
29. Prerequisites for Correlation and Regression
• Correlation is used as Bivariate analysis
• Regression is used as Uni A, Bi A, and Multivariate
Analysis
• Regression is likely to be computed only if there is positive
correlation
• Correlation Values
– Less than 0 (Negative Value) No Correlation
– 0 to 0.29 is Low Correlation
– 0.3 to 0.49 is Moderate Correlation
– 0.5 and above is High Correlation
• P Value
- less than 0.05 is Statistically Significant
- More than 0.05 is Statistically Insignificant
30. Difference Between Correlation and Regression
Students Group 1
Students Group 2
Students Group 3
Faculty "X"
Faculty "X"
Faculty "X"
0.8
0.75
0.85
Objective for the study:
To assess the relationship between Students and Faculty "X"
31. Difference Between Correlation and Regression
Students Group 1
Students Group 2
Students Group 3
Faculty "X"
0.80.55
0.6
0.7
0.65
0.45
Objective for the study:
To assess the relationship between Students and Faculty "X"
32. Example of Metric on Metric
2. Consumer Behavior on Post Purchase Behavior
Hypothesis:
– H0: There is no relationship between Consumer
Behavior and Post Purchase Behavior
– H1: There is relationship between Consumer
Behavior and Post Purchase Behavior
Test used is – Correlation and Regression Analysis
34. Prerequisite for T-Test
• It is a Bivariate Analysis
• Independent Variable is a Categorical data and
Dependent Variable is a Metric Data
• The Independent Variable (Categorical Data) must
have only two groups
• Example: If we consider state of residence we
need to select only 2 states or groups
35. • Reject null
– If T Statistics value is More than T Critical
Or
– If P value is less than 0.05
Prerequisite for T-Test
Objective of the study:
• To assess if the respondents based on their Gender
have different understanding on the aspect of
Consumer Behaviour
36. Example of Category on Metric
And there are two groups in the category
3. Hypothesis for the study
H0: There is no difference in mean scores of
Gender on the aspect of Consumer Behaviour
H1: There is difference in mean scores of
Gender on the aspect of Consumer Behaviour
Test used is – T Test
H0: Boys Girls
H1: Boys Girls
37. Interpretation
• To assess the uniformity of responses towards the aspect of
consumer behavior considering gender as categorical
variable, T-Test is carried out. The t statistics value is
identified to be 1.12 which is less than T critical, resulting
in accepting the null hypothesis. However if we refer the P-
value, it is identified that the value is more than 0.05 level
of significance. Hence it is observed that there is no
difference in mean scores of Gender on the aspect of
Consumer Behavior.
39. Prerequisites for
One Way ANOVA (Analysis of Variance)
• It is a Bivariate Analysis
• Independent Variable is a Categorical data and Dependent
Variable is a Metric Data
• The Independent Variable (Categorical Data) must have
more than two groups
• Example: Family occupation on Consumer Behaviour
Agriculture, Business, Government Employee, Others,
Private Employee, Self Employeed (Free Lancer)
Test used is – One Way ANOVA
40. • Reject null
– If F Statistics value is More than F Critical
Or
– If P value is less than 0.05
Prerequisite for One Way ANOVA
Objective of the study:
• To assess if the respondents based on their Family
occupation have different understanding on the
aspect of Consumer Behaviour
41. Example of Category on Metric
4. Family occupation on Consumer Behaviour
Hypothesis:
– H0: There is no difference in mean scores of Family
occupation on Consumer Behaviour
– H1: There is difference in mean scores of Family
occupation on Consumer Behaviour
H0: Agriculture Business Government Employee
Others Private Employee Self Employed (Free Lancer)
H1: Agriculture Business Government Employee
Others Private Employee Self Employed (Free Lancer)
42. Linking the Analysis with Marketing Research Process
• Determining the Statement of Problem
• Defining Research Objectives
• Setting Research Hypothesis
• Forming Research Design
• Framing Research Instrument
• Collection of Data
• Analyzing the Data
• Communicate Results
Title for the Study
"Consumer Behavior towards Toothpaste Consumption - Empirical Study
in Reference to Indian Youth".
43. Statement of Problem
• Tooth paste market is dominated by few of the
companies and it is difficult for the rest of the
market players and new entrants to create
footholds.
• Consumer behavior in today’s scenario is
unpredictable as they tend to shift variants or
brands
If you are doing research for an underdog in the segment (Low Market Share)
44. Research Objective (Using Bloom’s Taxonomy)
• To Identify the consumer preference towards the
existing toothpaste products amidst Indian youth.
(Consumption pattern)
• To understand the factors influencing the aspect of
Consumer Behavior
• To examine if there is difference in the opinion of
respondents based on their varied demographic profile
towards the aspect of Consumer Behavior
• To estimate if there is relationship between Consumer
behavior and Post purchase behavior amidst Indian
youth towards toothpaste consumption