Technical Interview Questions for Data Tracks at ITI

Answers of questions in this link

https://datapost73.blogspot.com/2020/08/iti-40-datamanagement-data-
science.html?fbclid=IwAR0e6VCxxcNVzaEMuaCqP0KVSJJIkFl9Ob9Juw
GtEwhxbDgRYrz_IcwN240

Index Question Answer

SQL + PL / SQL + Database (Very Important topic)
DCL (Data Control Language) is used to manage
permissions and access control. DML (Data
Manipulation Language) is used for data
manipulation like INSERT, UPDATE, DELETE.
DDL (Data Definition Language) is used to define
1 What are DCL, DML, and DDL in SQL? and manage database structures like CREATE,
ALTER, DROP. Example: DCL - GRANT SELECT
ON table TO user; DML - INSERT INTO table
(column1, column2) VALUES (value1, value2);
DDL - CREATE TABLE table (column1 datatype,
column2 datatype);
GROUP BY is used to group rows based on a
column's values, typically used with aggregate
What is the difference between group by and functions. HAVING is used to filter grouped results.
2
having? Example: SELECT department, AVG(salary)
FROM employees GROUP BY department
HAVING AVG(salary) > 50000;
ORDER BY is used to sort query results. You can
order by one or more columns by specifying
What is the order by? Can we order more than one
3 multiple column names in the ORDER BY clause.
column?
Example: SELECT name, age FROM students
ORDER BY age, name;
UNION combines the result sets of two or more
SELECT queries into a single result set, removing
duplicates. JOIN combines rows from two or more
tables based on a related column. Example:
4 What is the difference between union and join? UNION - SELECT name FROM table1 UNION
SELECT name FROM table2; JOIN - SELECT
customers.name, orders.order_date FROM
customers JOIN orders ON
customers.customer_id = orders.customer_id;
There are various types of joins: INNER JOIN
(returns matching rows), LEFT JOIN (returns all
rows from the left table and matching rows from
the right), RIGHT JOIN (returns all rows from the
right table and matching rows from the left), FULL
5 What is the different type of join?
OUTER JOIN (returns all rows when there is a
match in either table). Example: INNER JOIN -
SELECT customers.name, orders.order_date
FROM customers INNER JOIN orders ON
customers.customer_id = orders.customer_id;
Aggregate functions perform calculations on a set
of values and return a single result. Common
6 What is the aggregate functions? aggregates include COUNT, SUM, AVG, MAX,
and MIN. Example: SELECT COUNT(*) FROM
orders;
The typical sequence of SQL statements in a
7 What are the SQL statements Sequence?
query is: SELECT (columns) FROM (table)
WHERE (conditions) GROUP BY (columns)
HAVING (conditions) ORDER BY (columns);
A view is a virtual table based on the result of a
SELECT query. It simplifies complex queries,
provides security, and hides underlying table
8 What is the view? + Why we use it?
structures. Example: CREATE VIEW
employee_view AS SELECT name, salary FROM
employees WHERE department = 'HR';
A SQL transaction is a sequence of one or more
SQL statements treated as a single unit of work. It
follows ACID properties (Atomicity, Consistency,
9 What is the SQL transaction? Isolation, Durability) to ensure data integrity.
Example: BEGIN TRANSACTION; UPDATE
account SET balance = balance - 100 WHERE
account_number = '123'; COMMIT;
DELETE removes specific rows from a table based
on a condition and can be rolled back. TRUNCATE
What is the difference between delete and
10 removes all rows from a table and is not reversible.
truncate?
Example: DELETE FROM employees WHERE
department = 'IT'; TRUNCATE TABLE employees;
You can use the ALTER TABLE statement to add
a column to an existing table. Example: ALTER
11 How can insert a column to the table?
TABLE employees ADD COLUMN address
VARCHAR(255);
Use the INSERT INTO statement with multiple
How can insert multi rows in only one insert value sets in parentheses. Example: INSERT
12
statement? INTO students (name, age) VALUES ('Alice', 25),
('Bob', 22), ('Charlie', 28);
A database is a structured collection of data.
DBMS (Database Management System) is
software that manages databases. RDBMS
13 What is the database, DBMS, and RDBMS?
(Relational DBMS) stores data in tables with
relationships. Example: Database: CompanyDB;
DBMS: MySQL; RDBMS: PostgreSQL;
Attributes in a database represent properties of
entities. They can be classified as simple (atomic)
or composite (composed of sub-attributes) and
14 What are the kinds of attributes? derived (calculated from other attributes).
Example: Simple - Age, Composite - Address
(Street, City), Derived - TotalPrice (Quantity *
Price);
ERD (Entity-Relationship Diagram) is a visual
representation of database entities, their attributes,
15 What is the ERD?
and relationships between entities. It helps in
database design. Example: ;
Constraints enforce data integrity rules. Common
types include PRIMARY KEY, FOREIGN KEY,
UNIQUE, CHECK, and NOT NULL. Example:
16 What is the type of constraints?
PRIMARY KEY (employee_id), FOREIGN KEY
(department_id) REFERENCES
departments(department_id);
A primary key uniquely identifies rows in a table. A
foreign key establishes a link between tables,
What is the difference between primary key and ensuring referential integrity. Example: PRIMARY
17
foreign key? KEY - employee_id in employees table; FOREIGN
KEY - department_id in employees table
referencing departments table;
(Repeated question) DELETE removes specific
What the difference is between delete and
18 rows; TRUNCATE removes all rows and is not
truncate?
reversible.
DELETE SET NULL sets foreign key values to
NULL when referenced rows are deleted. DELETE
19 What is delete set null and delete cascade?
CASCADE deletes rows in related tables when the
referenced row is deleted. Example: DELETE SET
NULL - Set employee_id to NULL in orders when
an employee is deleted; DELETE CASCADE -
Delete all orders when an employee is deleted.
Normalization is the process of organizing data in
a database to reduce redundancy and improve
data integrity. It prevents update anomalies and
What is the normalization and why are we making
20 ensures efficient data storage. Example: 1NF -
it?
Ensure each column has atomic values; 2NF -
Remove partial dependencies; 3NF - Remove
transitive dependencies.
Common normalization forms include 1NF, 2NF,
3NF, BCNF, and 4NF. Each eliminates specific
21 What are the types of normalization? types of data redundancy. Example: 1NF - Each
column has atomic values; 2NF - No partial
dependencies; 3NF - No transitive dependencies.
Update anomalies occur when inconsistencies
arise due to data redundancy, such as when
updating data in one place but not another.
22 What are the update anomalies?
Example: In a denormalized table, updating an
employee's salary in one row but not in another for
the same employee.
SQL is a query language for managing and
querying data in databases. PL/SQL is a
procedural extension of SQL used for writing
23 What is the difference between SQL and PL/SQL? stored procedures and functions. Example SQL:
SELECT * FROM employees; Example PL/SQL:
CREATE PROCEDURE getEmployee (emp_id
NUMBER) AS BEGIN ... END;
PL/SQL provides loops like FOR LOOP, WHILE
24 What are the types of loops in PL/SQL? LOOP, and LOOP-END LOOP for repetitive tasks.
Example: FOR i IN 1..10 LOOP ... END LOOP;
Cursors are database objects used to retrieve and
manipulate data. Types include Implicit (used for
single-row queries) and Explicit (used for multi-row
queries), which can be further categorized as
What are the cursors and what are the cursors
25 Static, Dynamic, and Scrollable. Example: Implicit
types?
Cursor - SELECT name INTO employee_name
FROM employees WHERE id = 123; Explicit
Cursor - DECLARE emp_cursor CURSOR FOR
SELECT name FROM employees;
A procedure is a named collection of PL/SQL
statements that can be stored in a database and
executed as a single unit. It can take parameters
26 What is the procedure?
and return values. Example: CREATE
PROCEDURE calculate_salary (employee_id
NUMBER) AS BEGIN ... END;
A procedure doesn't return a value, while a
function does. Functions can be used in SQL
queries, whereas procedures cannot. Example
What is the difference between procedure and Procedure: CREATE PROCEDURE
27
function? update_employee (emp_id NUMBER) AS BEGIN
... END; Example Function: CREATE FUNCTION
get_employee_name (emp_id NUMBER)
RETURN VARCHAR2 AS BEGIN ... END;
Triggers are PL/SQL blocks executed
automatically in response to specific database
events. Types include BEFORE and AFTER
What are the triggers and what are the triggers
28 triggers for INSERT, UPDATE, DELETE events.
types?
Example: BEFORE INSERT Trigger - Prevent
inserting records with invalid data; AFTER
UPDATE Trigger - Log changes to a table.
SQL statements depend on specific requirements
29 Write SQL Statements and tables. For example, to insert data: INSERT
INTO employees (emp_id, emp_name) VALUES
(1, 'John Doe'); To update data: UPDATE products
SET price = price * 0.9 WHERE category =
'Electronics'; To delete data: DELETE FROM
customers WHERE last_purchase_date < '2022-
01-01';
Business Intelligence
Business Intelligence (BI) refers to the
technologies, processes, and tools used to
analyze and present business data to support
decision-making. It helps organizations gain
30 What is Business Intelligence?
insights, make informed decisions, and improve
business performance. Example: Using BI to
analyze sales data to identify trends and optimize
product offerings.
The typical steps in BI include data collection, data
integration (ETL - Extract, Transform, Load), data
storage, data analysis, and data visualization.
Example: 1. Collecting sales data from multiple
31 What are the steps in BI? sources. 2. Integrating and transforming the data
into a unified format. 3. Storing it in a data
warehouse. 4. Analyzing it to discover sales
trends. 5. Creating dashboards to visualize the
trends.
ETL (Extract, Transform, Load) tools include
Apache NiFi, Talend, and Informatica. Analysis
tools include Tableau, Power BI, and QlikView.
What are the tools we use in BI (for ETL, Analysis,
32 Visualization tools include D3.js, Google Data
and Visualization)?
Studio, and Looker. Example: Using Tableau for
data analysis and visualization to create interactive
sales reports.
Data Warehouse
A data warehouse is a centralized repository that
stores, integrates, and manages data from various
sources to support business reporting and
analysis. It is designed for query and analysis
33 What is the data warehouse?
rather than transaction processing. Example:
Storing historical sales data, customer information,
and product data for business intelligence
purposes.
Characteristics include subject-oriented (focus on
specific business areas), integrated (combines
data from diverse sources), time-variant (stores
34 What are the characteristics of a data warehouse? historical data), non-volatile (data is not updated
frequently), and supports complex queries.
Example: Analyzing sales trends over the last five
years.
A database is designed for transactional
processing, while a data warehouse is designed
for analytical processing. Databases support real-
What is the difference between a database and a time data updates, while data warehouses store
35
data warehouse? historical data and support complex queries for
reporting and analysis. Example: A database for
online order processing vs. a data warehouse for
sales analysis.
A data warehouse stores structured data in a
highly organized manner, while big data
encompasses vast volumes of structured and
unstructured data. Data warehouses are well-
What is the difference between a data warehouse
36 suited for structured data analysis, whereas big
and big data?
data technologies like Hadoop handle unstructured
and semi-structured data. Example: A data
warehouse for analyzing sales data vs. using big
data tools to analyze social media posts.
OLTP (Online Transaction Processing) systems
are used for day-to-day transactional operations,
supporting real-time data entry and retrieval. OLAP
37 What is the difference between OLTP and OLAP? (Online Analytical Processing) systems are for
complex data analysis and reporting. Example:
OLTP for processing bank transactions, OLAP for
analyzing customer spending patterns.
Data warehousing is the process of designing,
building, and maintaining data warehouses. It
involves data extraction, transformation, loading
38 What is Data Warehousing?
(ETL), and providing a platform for business
intelligence and reporting. Example: Setting up a
data warehousing system for a retail company.
Processes include data extraction, data
transformation, data loading (ETL), data storage,
What are the processes that can be done in the data retrieval, data modeling, and data analysis.
39
data warehouse? Example: Extracting customer data, transforming it
into a standardized format, and loading it into the
data warehouse for analysis.
Data modeling is the process of defining the
structure and relationships of data in a database or
data warehouse. Types include conceptual
modeling (high-level representation), logical
40 What is Data Modeling? + Types of Data Modeling?
modeling (entity-relationship diagrams), and
physical modeling (designing database tables).
Example: Creating an entity-relationship diagram
for a customer database.
Data warehouses are typically designed for read-
intensive operations, and updates are infrequent.
Updates can be performed, but they often involve
41 Can we update a record in a data warehouse?
complex ETL processes to maintain historical data.
Example: Correcting a customer's address in the
data warehouse.
A data mart is a subset of a data warehouse that
focuses on specific business areas or
departments. It contains a smaller, more
42 What is a data mart?
specialized set of data for targeted analysis.
Example: Creating a sales data mart for the Sales
department to analyze sales performance.
A data cube is a multi-dimensional representation
of data that allows for efficient querying and
analysis. It contains dimensions (attributes) and
43 What is a Data Cube?
measures (facts) and is often used in OLAP
systems. Example: Analyzing sales data with
dimensions like time, product, and region.
ETL (Extract, Transform, Load) is a process used
to extract data from source systems, transform it
into a desired format, and load it into a data
44 What is ETL? warehouse or data mart. Example: Extracting
sales data from a CRM system, transforming it to
match the data warehouse schema, and loading it
into the data warehouse.
In a star schema, dimension tables are directly
linked to a central fact table. In a snowflake
schema, dimension tables are normalized into
multiple related tables. Star schemas are simpler
What is the difference between snowflake and star
45 but can be less space-efficient, while snowflake
schema?
schemas save space but can be more complex.
Example: Star schema for sales analysis vs.
snowflake schema for complex product
hierarchies.
Fact tables contain numerical measures and
What is the difference between fact and dimension
46 foreign keys to dimension tables. Dimension tables
tables?
contain descriptive attributes about dimensions
such as time, product, or location. Example: Fact
table with sales revenue vs. dimension table with
product details.
Big Data
Big data is important because it enables
organizations to gain valuable insights from vast
and diverse datasets that were previously too large
and complex to manage and analyze effectively. It
47 Why is big data important? can uncover patterns, trends, and opportunities for
better decision-making. Example: Analyzing
customer behavior across social media, online
purchases, and offline interactions to enhance
marketing strategies.
Big data is characterized by the three V's: Volume
(large amounts of data), Velocity (high-speed data
generation and processing), and Variety (diverse
data types, structured and unstructured). Some
48 What is big data? (V's of Big Data) also add Veracity (data accuracy) and Value
(extracting insights). Example: Social media
platforms processing massive volumes of tweets
(Volume) in real-time (Velocity) with text, images,
and videos (Variety).
Data types in the context of big data can include
structured data (e.g., numbers, dates), semi-
structured data (e.g., JSON, XML), and
unstructured data (e.g., text, images, videos).
49 What are the data types?
Example: Structured data - Sales revenue as
numbers; Semi-structured data - Customer data in
JSON format; Unstructured data - Text reviews
from customers.
A Data Lake is a central repository that stores vast
amounts of raw and unprocessed data from
diverse sources. It allows for flexible and scalable
50 What is Data Lake?
data storage and analysis. Example: Storing log
files, sensor data, and social media posts in a Data
Lake for future analytics.
ETL (Extract, Transform, Load) involves extracting
data from source systems, transforming it before
loading it into a data warehouse. ELT (Extract,
Load, Transform) loads data into the data
warehouse first and then performs
transformations. ELT is often used in big data
51 What is the difference between ETL & ELT?
scenarios where data may not fit the traditional
ETL model. Example (ETL): Extracting sales data,
aggregating it, and loading it into a data
warehouse. Example (ELT): Loading raw log data
into a Data Lake, then transforming it into a
structured format for analysis.
Databases are designed for structured data
storage and transaction processing, while big data
encompasses both structured and unstructured
data. Big data technologies like Hadoop and
What is the difference between a Database and Big
52 NoSQL databases are built to handle massive
data?
volumes and varieties of data. Example: A
relational database for storing customer
information vs. Hadoop for processing social
media data.
Big data tools include Hadoop (for distributed
storage and processing), Spark (for fast data
processing), MapReduce (for data processing in
53 What are the tools in big data?
Hadoop), Hive (for querying and data
warehousing), Impala (for SQL queries on
Hadoop), Kafka (for real-time data streaming), and
more. Example: Using Spark for analyzing large
datasets in real-time.
- Hadoop is a distributed storage and processing
framework for big data. - Spark is a fast and
versatile data processing engine. - MapReduce is
a programming model used in Hadoop for parallel
processing. - Hive is a data warehousing and SQL
What are querying tool for Hadoop. - Impala is an open-
54
(Hadoop/Spark/MapReduce/Hive/Impala/Kafka/...)? source SQL query engine for Hadoop. - Kafka is a
distributed streaming platform for real-time data.
Example: Using Hadoop to store and process
large log files, Spark for real-time analytics, Hive
for querying structured data in Hadoop, and Kafka
for ingesting streaming data.
Data Science + Machine Learning + Data Mining (Data Science Track)
Data science is an interdisciplinary field that uses
scientific methods, algorithms, processes, and
systems to extract knowledge and insights from
structured and unstructured data. It combines
55 What is data science?
aspects of statistics, computer science, and
domain knowledge to solve complex problems.
Example: Using data science to analyze customer
behavior and recommend personalized products.
Data scientists focus on designing and
implementing complex algorithms to solve
business problems, often requiring programming
and machine learning expertise. Data analysts
What is the difference between data scientists and
56 primarily work on data exploration, visualization,
data analysts?
and basic statistical analysis to answer specific
questions. Example: A data scientist develops a
predictive model, while a data analyst creates
reports and dashboards.
Data cleaning involves identifying and correcting
errors, inconsistencies, and inaccuracies in
datasets. It includes tasks like handling missing
values, removing duplicates, and correcting
57 What is data cleaning? How do we clean the data?
outliers using statistical methods and domain
knowledge. Example: Replacing missing age
values in a dataset with the median age of known
values.
Data mining is the process of discovering patterns,
relationships, and valuable insights from large
datasets. It involves techniques like clustering,
58 What is Data Mining? classification, regression, and association rule
mining. Example: Analyzing retail sales data to
identify product associations for marketing
strategies.
Applications include fraud detection,
recommendation systems (e.g., Netflix), medical
diagnosis, sentiment analysis in social media,
What are the real-life applications of data mining
59 predictive maintenance in manufacturing, and
and machine learning?
autonomous vehicles. Example: Using machine
learning to predict disease outbreaks based on
historical health data.
The process involves data selection, data
preprocessing, data transformation, data mining,
pattern evaluation, and knowledge presentation.
What is the Process of Data Mining/Knowledge
60 Example: In e-commerce, selecting sales data,
Discovery Process?
preprocessing it (cleaning and transforming),
mining customer purchase patterns, and
presenting these patterns for business decisions.
Challenges include handling large datasets, data
61 What are the Challenges of Data Mining?
quality issues, selecting appropriate algorithms,
overfitting, interpretability of complex models, and
ensuring privacy and security of sensitive data.
Example: Dealing with skewed data distribution in
fraud detection, where fraudulent transactions are
rare.
Machine learning is a subset of artificial
intelligence that involves the development of
algorithms that enable computers to learn patterns
62 What is Machine Learning?
and make predictions or decisions from data.
Example: Training a machine learning model to
recognize handwritten digits in images.
Deep learning is a subset of machine learning that
uses artificial neural networks with multiple layers
(deep architectures) to automatically learn and
63 What is deep learning?
represent data. It excels in tasks like image and
speech recognition. Example: Training a deep
neural network to recognize objects in images.
Tasks include clustering (K-Means), classification
(Decision Trees), regression (Linear Regression),
association rule mining (Apriori), and anomaly
64 What are the data mining tasks/algorithms?
detection (Isolation Forest). Example: Using K-
Means to group customers based on purchasing
behavior.
Supervised learning uses labeled data to train
models (e.g., classification or regression), while
What is the difference between Supervised and unsupervised learning uses unlabeled data to find
65
Unsupervised learning? patterns or groupings (e.g., clustering). Examples:
Supervised - Spam email detection; Unsupervised
- Customer segmentation.
Classification assigns labels to data based on
predefined classes, while clustering groups data
What is the difference between Classification and into clusters based on similarity. Examples:
66
Clustering? Classification - Identifying email as spam or not;
Clustering - Grouping customers into market
segments.
K-Means, Hierarchical Clustering, and DBSCAN
67 Examples of clustering algorithms
are examples of clustering algorithms.
Decision Trees, Logistic Regression, and Support
68 Examples for classification algorithms Vector Machines (SVM) are examples of
classification algorithms.
Association rules identify relationships between
items in a dataset, often used in market basket
69 What is an association rule? analysis to find item associations in transactions.
Example: "If a customer buys bread, they are likely
to buy butter."
Provide brief explanations of how each algorithm
How does this algorithm work (K-Mean, works. Example: K-Means clusters data points into
70 Regression, SVM, association rule, decision tree, K clusters based on proximity; Decision Trees
KNN...)? make decisions by following a tree-like structure of
if-else conditions.
Recall measures the ability of a model to identify
all relevant instances. Precision measures the
ability of a model to return only relevant instances.
71 What is recall and precision, F1? F1-score is the harmonic mean of precision and
recall, balancing them. Example: In a medical test,
recall is the percentage of actual sick patients
correctly identified by the test.
The bias-variance trade-off refers to the balance
between model complexity and model
performance. A model with high bias (underfitting)
72 What is the bias-variance trade-off? has low complexity and may not capture
underlying patterns. A model with high variance
(overfitting) fits the training data too closely and
may not generalize well to new data. Example: In
polynomial regression, increasing the polynomial
degree leads to lower bias but higher variance.
A confusion matrix is a table that visualizes the
performance of a classification algorithm. It shows
true positives, true negatives, false positives, and
73 What is the confusion matrix?
false negatives. Example: In a binary classification
problem, the confusion matrix may look like this:
TP: 120, TN: 80, FP: 10, FN: 5.
The ROC (Receiver Operating Characteristic)
curve is a graphical representation of a classifier's
performance, showing the trade-off between true
74 What is the ROC Curve? positive rate and false positive rate at various
thresholds. Example: In medical diagnosis, plotting
the ROC curve helps assess the accuracy of a
diagnostic test.
Cross-validation is a technique used to evaluate
the performance of a machine learning model by
dividing the dataset into multiple subsets (folds). It
trains and tests the model on different
75 Explain cross-validation? combinations of folds to assess its generalization
ability. Example: Using k-fold cross-validation to
train and test a model on five subsets of the data,
rotating which subset is used for testing in each
iteration.
A validation set is used during the model training
phase to tune hyperparameters and assess
performance. A test set is a separate dataset used
What is the difference between a validation set and to evaluate the final model's generalization
76
a test set? performance. Example: Using a validation set to
adjust the learning rate in gradient boosting and a
test set to estimate the model's accuracy on
unseen data.
Missing values can be imputed using methods like
mean, median, or interpolation. Outliers can be
identified and removed or transformed using
77 How do you treat missing/outlier values? statistical techniques. Example: Replacing missing
age values with the median age of known values;
Detecting outliers using the Z-score and removing
extreme values.
Data preparation involves data cleaning, feature
selection/engineering, handling missing values,
scaling/normalizing features, and splitting data into
78 How do you prepare the data for the ML Model? training, validation, and test sets. Example: Scaling
numerical features to have a mean of 0 and a
standard deviation of 1 for better model
convergence.
Statistics (Data Science Track)
Variance measures how individual data points
deviate from the mean. Standard deviation is the
square root of the variance and measures the
What is the difference between standard deviation average deviation of data points from the mean.
79
and variance? Example: Variance calculates the average squared
difference from the mean, while standard deviation
provides a more interpretable measure in the
original units of the data.
Mean is the average of a set of numbers. Median
is the middle number when the numbers are
80 What are Mean, Median, and Mode? ordered. Mode is the value that appears most
frequently. Example: For the set of numbers {2, 3,
3, 5, 7}, Mean = 4, Median = 3, Mode = 3.
Variance measures the spread or dispersion of
What is the difference between variance and data by calculating the average of squared
81
standard deviation? differences from the mean. Standard deviation is
the square root of the variance and provides a
more interpretable measure in the original units of
the data. Example: Variance = 9, Standard
Deviation = 3 for the set {1, 2, 3, 4, 5}.
A box plot (box-and-whisker plot) is a graphical
representation of the distribution of data. It shows
the median, quartiles, and potential outliers. The
box represents the interquartile range (IQR), and
82 What is the Box plot?
the whiskers extend to the minimum and maximum
values within a defined range. Example: A box plot
showing the distribution of test scores, with the
median, quartiles, and any outliers.
Skewed data can be positively skewed (right-
skewed) where the tail extends to the right, or
negatively skewed (left-skewed) where the tail
83 What are the types of skewed data? extends to the left. Example: Positive skew in
income distribution data due to a few high earners;
Negative skew in test scores with many high
scores.
The Z-score (standard score) measures how many
standard deviations a data point is from the mean.
It standardizes data, making it possible to compare
84 What is the Z-score?
values from different datasets. Example: A Z-score
of -1.5 indicates a data point is 1.5 standard
deviations below the mean.
The P-value measures the evidence against a null
hypothesis in hypothesis testing. It indicates the
probability of observing a test statistic as extreme
as, or more extreme than, what is observed in the
85 What is the P-value?
sample, assuming the null hypothesis is true.
Example: In a medical trial, a P-value of 0.03
suggests a 3% chance of observing the results if
the treatment has no effect (null hypothesis).
The Pearson correlation coefficient (Pearson's r)
measures the linear relationship between two
continuous variables. It ranges from -1 (perfect
negative correlation) to 1 (perfect positive
86 What is the Pearson correlation coefficient?
correlation), with 0 indicating no linear correlation.
Example: Pearson's r of 0.75 between hours
studied and exam scores suggests a strong
positive correlation.
A/B testing (split testing) is a controlled experiment
where two versions (A and B) of a webpage, app,
or product are compared to determine which
87 What is A/B Testing? performs better in terms of user engagement or
conversions. Example: Testing two different
website layouts to see which one results in higher
click-through rates.
Hypothesis testing is a statistical method used to
make inferences about population parameters
based on sample data. It involves formulating a
null hypothesis (no effect) and an alternative
88 What is hypothesis testing?
hypothesis (an effect exists) and testing the null
hypothesis using data and statistical tests.
Example: Testing whether a new drug is more
effective than an existing one in a clinical trial.