Statistical Data Analysis and visualization

Credit:3
Total Hours: 39

Pre-requisites

● Basic knowledge of Probability (Sample space, Events).

● Basic programming skills (preferably Python or R).

Module No. Topic Total Hours

Module 1 Introduction to Data Analytics and Descriptive Statistics 6 Hours

1.1 The Data Analytics Lifecycle: Overview, Phases (Discovery,
Data Preparation, Model Planning, Model Building,
Communicate Results, Operationalize). Role of Statistics in
Data Science. (2 Hours)
1.2 Types of Data: Data types (Categorical, Quantitative),
Measures of Central Tendency (Mean, Median, Mode) and
Dispersion (Variance, Standard Deviation, Quartiles, IQR). (2 Hours)
1.3 Probability Basics in Statistics: Conditional probability,
Bayes' Theorem (Conceptual understanding and application in
models like Naive Bayes). Measures of correlation and
covariance. (2 Hours)

Module 2 Data Preparation and Exploratory Data Analysis (EDA) 7 Hours

2.1 Data Quality and Pre-processing: Data cleaning, Handling
missing values (Imputation techniques), Outlier detection, Data
transformation (Log transformation, Square root). (3 Hours)
2.2 Data Transformation and Feature Engineering:
Normalization and Standardization (z-score), Encoding
categorical data (One-hot, Label encoding), Dimensionality
Reduction introduction (e.g., PCA conceptual). (2 Hours)
2.3 Exploratory Data Analysis (EDA) Philosophy and Tools:
Summarizing datasets, Basic statistical tools of EDA (plots,
graphs, summary statistics), Data wrangling using
Pandas/NumPy. (2 Hours)
Module 3 Statistical Distributions and Inferential Statistics 7 Hours
3.1 Probability Distributions: Discrete Distributions (Binomial,
Poisson) and Continuous Distributions (Normal, Gaussian
Distribution). Central Limit Theorem (CLT) and its significance in
data analysis. (3 Hours)
3.2 Sampling and Estimation: Population vs. Sample, Sampling
methods, Point and Interval Estimation, Standard Error,
Confidence Intervals. (2 Hours)
3.3 Hypothesis Testing: Introduction, Null and Alternative
Hypothesis, Type I and Type II errors, p-value, One-sample and
Two-sample t-tests, Chi-square test (Conceptual overview). (2 Hours)
Module 4 Regression Analysis for Predictive Modeling 7 Hours
4.1 Simple Linear Regression (SLR): Model formulation, Least
squares estimation, Interpretation of coefficients, Coefficient of
determination (R2). (3 Hours)
4.2 Multiple Linear Regression: Model assumptions
(Homoscedasticity, Normality of Residuals), Multicollinearity,
Model Selection techniques (Forward, Backward, Stepwise
Regression). (2 Hours)
4.3 Introduction to Classification Models: Logistic Regression
(The Logistic Response function and logit), Generalized Linear
Model (GLM) concept, Model Evaluation Metrics (Confusion
Matrix, Accuracy). (2 Hours)
Module 5 Principles and Tools for Data Visualization 6 Hours
5.1 Fundamentals of Visualization: Importance of visualization,
Principles of effective visualization, Visualizing single and
multiple variables. (2 Hours)
5.2 Basic Visualization Techniques: Bar Charts, Histograms,
Box Plots, Scatter Plots, Line Graphs, Pie Charts, Heatmaps.
Choosing the right chart type. (2 Hours)
5.3 Visualization with Python Libraries: Introduction to Matplotlib
and Seaborn/Plotly. Creating static and basic interactive plots
for statistical insights. (2 Hours)
Module 6 Advanced Visualization and Data Storytelling 6 Hours
6.1 Advanced and Multivariate Visualization: Visualizing large-
scale and multidimensional data, Pair plots, Geographic
Information System (GIS) visualizations (Introduction). (2 Hours)
 6.2 Interactive Dashboards: Introduction to Business Intelligence
(BI) tools (e.g., Tableau, Power BI, Looker Studio -
Conceptual/Demonstration only), Creating and sharing
interactive dashboards and dynamic reports. (2 Hours)
6.3 Data Storytelling: Creating a compelling narrative with data,
Do's and Don'ts of visual presentation, Presenting and
communicating statistical findings and insights to stakeholders. (2 Hours)

Textbooks

1. Practical Statistics for Data Scientists: 50+ Essential Concepts Using R

and Python by Peter Bruce, Andrew Bruce, and Peter Gedeck. O'Reilly
Media.
2. Data Science and Big Data Analytics: Discovering, Analyzing, Visualizing
and Presenting Data by EMC Education Services, Wiley Publication.
3. The Visual Display of Quantitative Information by Edward R. Tufte.
Graphics Press.

Reference Books

1. Naked Statistics: Stripping the Dread from the Data by Charles Wheelan.
W. W. Norton & Company.
2. Data Analytics using Python by Bharati Motwani. Wiley Publications.
3. Visual Analytics Using Tableau: Structured approach for turning raw data
to powerful insights by Neha Singh Rajput and Sulabh Bhatt.
4. Data Mining: Concepts and Techniques by Jiawei Han, Micheline Kamber,
and Jian Pei. Elsevier.

Mapped
Bloom's Theory
CO No. Course Outcome Level Modules
 Understand the data analytics lifecycle and
analyze raw data using descriptive statistics to
summarize characteristics, assess quality, and
CO1 perform essential data preparation steps (EDA). Analyse(4) M1, M2
Apply the principles of probability distributions, the
Central Limit Theorem (CLT), and sampling
techniques to determine confidence intervals and
CO2 formulate hypotheses for statistical testing. Apply (3) M1, M3
Evaluate the assumptions and performance of
statistical models, specifically Linear and Logistic
Regression, and interpret the coefficients and
CO3 model diagnostics for deriving insights. Evaluate(5) M4
Analyze the principles of effective data
visualization and select appropriate chart types
(static and multivariate) based on the type of data
CO4 and the statistical question being addressed. Analyse (4) M5, M6
Formulate a compelling data narrative using visual
and statistical evidence, thereby demonstrating the M6 (and
ability to communicate complex analytical results conceptual link
CO5 to technical and non-technical audiences. Create(6) to M2, M4)

Theory
Module Tools/
Lab Alignment Topic and Key Tasks Libraries
 Pandas,
1 Module 1 Introduction to Python for Data Analysis & Data Loading NumPy
Tasks: Data structures (Series, DataFrame), Loading data from
CSV, basic data inspection.
Pandas,
2 Module 1 Descriptive Statistics and Data Summaries NumPy
Tasks: Calculate Measures of Central Tendency and Dispersion
(Mean, Median, Mode, Variance, IQR, Skewness) for data
columns.
3 Module 2 Data Cleaning: Missing Data and Outliers (Python) Pandas
Tasks: Detect missing values, apply statistical imputation, detect
and handle outliers using the IQR method.
Pandas,
4 Module 2 Data Preparation and Feature Engineering (Python) Scikit-learn
Tasks: Implement Standardization (z-score), Data
Transformation (e.g., Log), and Categorical Encoding.
NumPy,
Matplotlib,
5 Module 3 Probability Distributions and The Central Limit Theorem (Python) SciPy
Tasks: Simulate samples from distributions, empirically
demonstrate the Central Limit Theorem (CLT).
SciPy,
6 Module 3 Inferential Statistics: Hypothesis Testing (Python) Statsmodels
Tasks: Calculate Confidence Intervals, perform a Two-Sample t-
test to check for significant differences (A/B Test simulation).
Statsmodels,
7 Module 4 Simple and Multiple Linear Regression (Python) Scikit-learn
Tasks: Fit a Multiple Linear Regression model, interpret the
coefficients, and calculate R2. Export the residuals/predictions
for visualization.
Statsmodels,
8 Module 4 Regression Diagnostics and Logistic Regression (Python) Scikit-learn
Tasks: Perform diagnostics (Q-Q plot, residual plots). Apply and
evaluate a basic Logistic Regression model.
Tableau
Desktop/Publ
9 Module 5, 6 BI Tool Session 1: Introduction to Tableau for Visual Analytics ic
Tasks: Connect to data, create basic charts (Bar, Line, Scatter).
Implement Hierarchies and Filters. Perform basic calculated
fields.
10 Module 5, 6 BI Tool Session 2: Tableau Dashboards and Statistical Visuals Tableau
Desktop/Publ
 ic
Tasks: Create advanced statistical charts (Box Plots,
Histograms) in Tableau. Build an interactive dashboard for EDA.
Power BI
11 Module 5, 6 BI Tool Session 3: Introduction to Power BI and Data Modeling Desktop
Tasks: Import data, create relationships between tables (Data
Modeling). Understand and use basic DAX formulas (e.g.,
aggregate calculations).
Power BI
12 Module 5, 6 BI Tool Session 4: Power BI Dashboards and Communication Desktop
Tasks: Create key visualizations (KPIs, Gauges, Treemaps).
Apply filters/slicers. Publish a final Power BI Report to a web
service (conceptual or local export).
Python,
Comprehensive Case Study: Statistical Reporting and Tableau/Pow
13 Module 6 Storytelling er BI
Tasks: Use a new dataset. Perform statistical analysis in Python.
Export the insights and use a BI tool to create a Data Storytelling
Dashboard that communicates the findings effectively.

Software Requirements

● Python 3.x (Jupyter Notebook) with NumPy, Pandas, Statsmodels, Scikit-

learn.
● Tableau Desktop/Public
● Microsoft Power BI Desktop

Bloom's Mapped Lab

LO No. Lab Outcome Level Sessions
Execute data manipulation and cleaning
processes, including handling missing values,
managing outliers, and applying statistical data
transformations using programming libraries
LO1 (Pandas/NumPy). Apply(3) L1, L2, L3, L4
 Implement and interpret fundamental inferential
statistical procedures, such as simulating
distributions and performing hypothesis tests (t-
tests), using statistical software packages
LO2 (SciPy/Statsmodels). Apply (3) L5, L6
Construct and validate basic predictive models
(Linear and Logistic Regression) and accurately
report their statistical performance metrics and
LO3 diagnostics using Scikit-learn and Statsmodels. Apply (3) L7, L8
Develop a variety of analytical and multivariate
visualizations, including static plots (e.g., box
plots, scatter plots) and interactive dashboards,
LO4 using dedicated BI tools (Tableau, Power BI). Create(6) L9, L10, L11, L12
Design and produce a comprehensive data
analysis report or dashboard that effectively
integrates statistical results and visualizations to
LO5 answer a specific data science question. Create (6) L13 (Case Study)