Database Concepts

Overview

Learning outcomes

• Students will be able to explain the need for a database using a real-life example.
• Students will be able to describe the limitations of a file system: difficulty of access, redundancy, inconsistency, isolation, dependence and uncontrolled sharing.
• Students will be able to define a DBMS, name common DBMS software, and explain schema, constraint, meta-data, instance and query.
• Students will be able to use relational model terms: relation, attribute, tuple, domain, degree and cardinality.
• Students will be able to state the properties of a relation and the meaning of NULL.
• Students will be able to identify candidate, primary, alternate and composite keys in a given table.
• Students will be able to explain a foreign key and identify it between two related tables.
• Students will be able to redesign file-based records into related tables with suitable keys.

Lesson plan

Notes

# A relation stored as a header plus a list of tuples
header = ("TrainNo", "TrainName", "Source", "Destination", "Type")
rows = [
    (12951, "Mumbai Rajdhani", "Mumbai Central", "New Delhi", "Rajdhani"),
    (12301, "Howrah Rajdhani", "Howrah", "New Delhi", "Rajdhani"),
    (12002, "Bhopal Shatabdi", "New Delhi", "Rani Kamlapati", "Shatabdi"),
    (22436, "Vande Bharat Express", "New Delhi", "Varanasi", "Vande Bharat"),
]
print("Degree (number of attributes):", len(header))
print("Cardinality (number of tuples):", len(rows))

Degree (number of attributes): 5
Cardinality (number of tuples): 4

# FEE relation: neither AdmNo nor Month is unique on its own
fee = [
    (1101, "April", 2500), (1102, "April", 2500),
    (1101, "May",   2500), (1102, "May",   2700),
    (1103, "April", 2500),
]
adm = [r[0] for r in fee]
month = [r[1] for r in fee]
pair = [(r[0], r[1]) for r in fee]

print("AdmNo unique?         ", len(adm) == len(set(adm)))
print("Month unique?         ", len(month) == len(set(month)))
print("(AdmNo, Month) unique?", len(pair) == len(set(pair)))

AdmNo unique?          False
Month unique?          False
(AdmNo, Month) unique? True

CREATE TABLE PARENT (
    ParentID  INT PRIMARY KEY,
    PName     VARCHAR(40) NOT NULL,
    Phone     CHAR(10) UNIQUE,
    Email     VARCHAR(50) UNIQUE
);

CREATE TABLE STUDENT (
    AdmNo     INT PRIMARY KEY,
    SName     VARCHAR(40) NOT NULL,
    Class     VARCHAR(5),
    DOB       DATE,
    ParentID  INT,
    FOREIGN KEY (ParentID) REFERENCES PARENT(ParentID)
);

CREATE TABLE FEE (
    AdmNo     INT,
    Month     VARCHAR(10),
    Amount    DECIMAL(8,2),
    PaidOn    DATE,
    PRIMARY KEY (AdmNo, Month),
    FOREIGN KEY (AdmNo) REFERENCES STUDENT(AdmNo)
);

Key terms

File system

Storing data in separate files, each accessed by its own application programs.

Database

An organised collection of logically related data.

DBMS

Software used to create, manage and query databases, e.g. MySQL.

Data redundancy

Unnecessary duplication of the same data in more than one place.

Data inconsistency

Copies of the same data that do not match.

Database schema

The design of a database: tables, attributes, data types, constraints and relationships.

Database instance

The data stored in a database at a particular moment.

Meta-data / data dictionary

Data about data; the stored schema and constraints of a database.

Relation

A table in the relational data model.

Attribute

A column of a relation.

Tuple

A row of a relation.

Domain

The set of permitted values for an attribute.

Degree

The number of attributes (columns) in a relation.

Cardinality

The number of tuples (rows) in a relation.

Candidate key

An attribute or minimal set of attributes that can uniquely identify every tuple.

Primary key

The candidate key chosen to identify tuples uniquely; unique and never NULL.

Alternate key

A candidate key that is not chosen as the primary key.

Composite key

A primary key made up of two or more attributes.

Foreign key

An attribute whose values refer to the primary key of another relation.

NULL

A special marker for a value that is unknown or not applicable; not zero or blank.

MCQs

Tap an option to check it, or use Show answers above to reveal them all.

1. Storing the same data in more than one file is called:

   1. Data isolation
   2. Data redundancy
   3. Data dependence
   4. Data integrity
   Answer

   (b) Data redundancy

   Redundancy means duplication of data.

2. A student's address is updated in the office file but not in the transport file. This is an example of:

   1. Data inconsistency
   2. Data isolation
   3. Data dependence
   4. Controlled sharing
   Answer

   (a) Data inconsistency

   The two copies no longer match, so the data is inconsistent.

3. When the format of a date column in a file is changed, all programs using the file must be modified. This limitation is called:

   1. Redundancy
   2. Inconsistency
   3. Data dependence
   4. Difficulty in access
   Answer

   (c) Data dependence

   Programs depend on the file's structure: data dependence.

4. Which of the following is NOT a DBMS?

   1. MySQL
   2. Oracle
   3. PostgreSQL
   4. Python
   Answer

   (d) Python

   Python is a programming language, not a database management system.

5. The overall design of a database, including tables, attributes and constraints, is called its:

   1. Instance
   2. Schema
   3. Tuple
   4. Query
   Answer

   (b) Schema

   The schema is the skeleton or design; the instance is the data at a moment.

6. Data about data, stored in the database catalog, is called:

   1. Meta-data
   2. Raw data
   3. Instance
   4. Domain
   Answer

   (a) Meta-data

   The data dictionary stores meta-data such as the schema and constraints.

7. In the relational model, a row of a table is called a:

   1. Attribute
   2. Domain
   3. Tuple
   4. Degree
   Answer

   (c) Tuple

   Row = tuple; column = attribute.

8. The number of attributes in a relation is its:

   1. Cardinality
   2. Domain
   3. Size
   4. Degree
   Answer

   (d) Degree

   Degree counts columns; cardinality counts rows.

9. A table SHOP has 7 columns and 25 rows. Its degree and cardinality are:

   1. 7 and 25
   2. 25 and 7
   3. 32 and 7
   4. 7 and 175
   Answer

   (a) 7 and 25

   Degree = columns = 7; cardinality = rows = 25.

10. A relation has degree 4 and cardinality 12. After adding 3 rows and 1 column, what are the new degree and cardinality?

    1. 4 and 15
    2. 5 and 12
    3. 5 and 15
    4. 7 and 13
    Answer

    (c) 5 and 15

    Degree 4 + 1 = 5; cardinality 12 + 3 = 15.

11. The set of permitted values for an attribute is called its:

    1. Tuple
    2. Domain
    3. Key
    4. Schema
    Answer

    (b) Domain

    For example, the domain of a Gender column might be {M, F, T}.

12. Which statement about a primary key is TRUE?

    1. It can contain NULL values
    2. It can contain duplicate values
    3. It uniquely identifies each tuple and cannot be NULL
    4. A table can have many primary keys
    Answer

    (c) It uniquely identifies each tuple and cannot be NULL

    A primary key is unique and never NULL, and there is only one per table.

13. All attributes that can uniquely identify tuples in a relation are called:

    1. Foreign keys
    2. Alternate keys
    3. Composite keys
    4. Candidate keys
    Answer

    (d) Candidate keys

    Every attribute able to serve as primary key is a candidate key.

14. In EMPLOYEE(EmpID, Name, PAN, Mobile, Dept), EmpID, PAN and Mobile are all unique. If EmpID is the primary key, then PAN and Mobile are:

    1. Foreign keys
    2. Alternate keys
    3. Composite keys
    4. Non-key attributes
    Answer

    (b) Alternate keys

    Candidate keys not chosen as the primary key are alternate keys.

15. A primary key made of more than one attribute is a:

    1. Composite key
    2. Foreign key
    3. Alternate key
    4. Super table
    Answer

    (a) Composite key

    For example (AdmNo, Month) in a FEE table.

16. An attribute of one table that refers to the primary key of another table is a:

    1. Candidate key
    2. Alternate key
    3. Foreign key
    4. Domain
    Answer

    (c) Foreign key

    A foreign key links the referencing table to the referenced table.

17. In ORDERS(OrderID, CustID, OrderDate) and CUSTOMER(CustID, CName, City), the foreign key is:

    1. OrderID in ORDERS
    2. CustID in CUSTOMER
    3. City in CUSTOMER
    4. CustID in ORDERS
    Answer

    (d) CustID in ORDERS

    CustID in ORDERS refers to CustID, the primary key of CUSTOMER.

18. Which value can a foreign key take, provided it is not part of its own table's primary key?

    1. Any value
    2. NULL or a value existing in the referenced primary key
    3. Only values not in the referenced table
    4. Only unique values
    Answer

    (b) NULL or a value existing in the referenced primary key

    Non-NULL foreign key values must match an existing primary key value; NULL is allowed.

19. Which of these is NOT a property of a relation?

    1. Each attribute has a unique name
    2. No two tuples are identical
    3. Tuples must be stored in sorted order
    4. Each value is atomic
    Answer

    (c) Tuples must be stored in sorted order

    The order of tuples does not matter in a relation.

20. Which is a limitation of a DBMS compared with a file system?

    1. Higher cost and complexity
    2. More data redundancy
    3. Greater data dependence
    4. No security
    Answer

    (a) Higher cost and complexity

    A DBMS reduces redundancy and dependence but needs costly software, hardware and trained staff.

Assertion–Reason

1. Assertion (A): A DBMS reduces data inconsistency.

   Reason (R): In a DBMS, each piece of data is ideally stored only once, so an update is made in one place.

   Answer

   (A)

   Inconsistency arises from duplicate copies. Removing redundancy removes the cause, so R explains A.

2. Assertion (A): RollNo cannot be the primary key of a table storing students of all sections of Class XI.

   Reason (R): A primary key must uniquely identify each tuple, and the same roll number appears in different sections.

   Answer

   (A)

   Both are true and R gives the reason.

3. Assertion (A): A relation can have more than one candidate key.

   Reason (R): A relation can have more than one primary key.

   Answer

   (C)

   A is true. R is false: there is exactly one primary key, though it may be composite.

4. Assertion (A): A foreign key can have duplicate values in its table.

   Reason (R): Every value of a foreign key must be unique in the referenced table.

   Answer

   (B)

   A is true: many students may share one ParentID. R is also true, since the referenced attribute is a primary key and so unique, but it does not explain why duplicates are allowed in the referencing table.

5. Assertion (A): The cardinality of a relation includes the row of column headings.

   Reason (R): Cardinality is the number of tuples in a relation.

   Answer

   (D)

   Headings are attribute names, not a tuple, so A is false. R is the correct definition.

Short answer

1. 2 marks

   Define data redundancy and data inconsistency. How are they related?

   Answer

   Data redundancy is the duplication of the same data in more than one place. Data inconsistency is when such duplicate copies do not match, for example after one copy is updated and another is not. Redundancy is the cause; inconsistency is the result.

2. 2 marks

   What is a DBMS? Name any two DBMS software.

   Answer

   A Database Management System is software used to create and manage databases: to define their structure and to store, retrieve, update and delete data, and to control access. Examples: MySQL, Oracle (also PostgreSQL, SQL Server, MS Access).

3. 2 marks

   Differentiate between database schema and database instance.

   Answer

   The schema is the design of the database: tables, attributes, data types, constraints and relationships. It rarely changes. The instance is the actual data stored at a particular moment, a snapshot. It changes whenever rows are inserted, updated or deleted. One schema has many instances over time.

4. 2 marks

   Define degree and cardinality. Find them for a table with columns ItemNo, ItemName, Price, Qty and 9 records.

   Answer

   Degree is the number of attributes (columns); cardinality is the number of tuples (rows).

   Here degree = 4 and cardinality = 9.

5. 2 marks

   What is a domain? Give the domain of the attributes Gender and Marks (out of 100) in a STUDENT table.

   Answer

   A domain is the set of permitted values from which an attribute takes its value.

   • Gender: {'M', 'F', 'T'} (single characters from a fixed set)
   • Marks: whole numbers from 0 to 100
6. 2 marks

   Differentiate between primary key and alternate key with an example.

   Answer

   Both are candidate keys. The primary key is the candidate key chosen to identify tuples; the remaining candidate keys are alternate keys.

   Example: in VOTER(VoterID, AadhaarNo, Name, Ward), VoterID and AadhaarNo are both candidate keys. If VoterID is the primary key, AadhaarNo is an alternate key.

7. 2 marks

   What is a candidate key? Can a table have more than one? Explain.

   Answer

   A candidate key is an attribute (or minimal set of attributes) whose values are unique and not NULL for every tuple, so it can identify each row. Yes, a table can have several. For example, a BANK_ACCOUNT table may have AccountNo, IBAN and DebitCardNo, each unique.

8. 2 marks

   What is a composite primary key? Give an example.

   Answer

   A primary key made of two or more attributes, used when no single attribute is unique. Example: in ATTENDANCE(AdmNo, AttDate, Status), AdmNo repeats every day and AttDate repeats for every student, but (AdmNo, AttDate) is unique, so it is the composite primary key.

9. 2 marks

   What is a foreign key? Explain with an example of two tables.

   Answer

   A foreign key is an attribute of one table whose values come from the primary key of another table; it links the two tables.

   Example: DEPT(DeptNo, DName) and EMP(EmpNo, EName, DeptNo). DeptNo in EMP is a foreign key referring to DeptNo, the primary key of DEPT. Every DeptNo in EMP must exist in DEPT (or be NULL).

10. 2 marks

    Explain data isolation and data dependence as limitations of a file system.

    Answer

    Data isolation: related data is kept in separate files, possibly in different formats, with no link between them, so combining them needs special programs.

    Data dependence: programs are written for a file's exact structure, so any change in the structure (adding a column, changing a format) forces every program that uses the file to be rewritten.

11. 2 marks

    What is NULL? How is it different from zero?

    Answer

    NULL is a special marker meaning the value is unknown or not applicable, for example a phone number not yet given. Zero is a known numeric value. A student's marks being 0 means the student scored zero; marks being NULL means the marks have not been entered.

12. 2 marks

    Give any two limitations of a DBMS.

    Answer

    • High cost: hardware, software licences, and training users to write queries.
    • Increased vulnerability: data is stored centrally, so a single failure can stop all users and can lose data unless regular backups are taken.

    (Also accepted: increased complexity of managing security, consistency and recovery.)

Long answer

1. 3 marks

   Explain any three limitations of a file system with examples from a school.

   Answer

   1. Data redundancy: the office's STUDENT file and the library's MEMBER file both store each student's name and class, wasting space.
   2. Data inconsistency: when a student changes section, the office updates its file but the library does not, so the two files show different sections.
   3. Controlled data sharing is difficult: parents should only view fee details and accounts staff should update them, but plain files cannot easily give different users different rights.

   (Difficulty in access, isolation and dependence are also accepted, each with an example.)

2. 3 marks

   Consider the relation BOOK(BookID, Title, Author, ISBN, Price) with these tuples (ISBNs are sample values):

   • (B01, Wings of Fire, A.P.J. Abdul Kalam, 978-81-0001, 350)
   • (B02, Godan, Premchand, 978-81-0002, 250)
   • (B03, Malgudi Days, R.K. Narayan, 978-81-0003, 199)
   • (B04, Gitanjali, Rabindranath Tagore, 978-81-0004, 150)

   (a) Find its degree and cardinality. (b) Identify the candidate keys. (c) If BookID is the primary key, name the alternate key.

   Answer

   (a) Degree = 5 (BookID, Title, Author, ISBN, Price); cardinality = 4.

   (b) Candidate keys: BookID and ISBN, since each book has a unique ID and a unique ISBN. Title and Author can repeat (an author writes many books; two books can share a title), and Price obviously repeats.

   (c) ISBN is the alternate key.

3. 3 marks

   Consider the tables:

   COACH(CoachID, CName, Sport)
   PLAYER(PlayerID, PName, Age, CoachID)

   (a) Identify the primary key of each table. (b) Identify the foreign key and the table it refers to. (c) Can two players have the same CoachID? Can a player have CoachID = 'C99' if there is no such coach? Explain.

   Answer

   (a) COACH: CoachID. PLAYER: PlayerID.

   (b) CoachID in PLAYER is a foreign key referring to CoachID, the primary key of COACH.

   (c) Yes, many players can have the same coach, so foreign key values may repeat. No, C99 is not allowed: a non-NULL foreign key value must match an existing primary key value in COACH.

4. 3 marks

   Differentiate between a file system and a DBMS on any three points.

   Answer

   • Redundancy: in a file system, the same data is duplicated across files. A DBMS stores data once and links tables through keys.
   • Consistency: in a file system, duplicate copies may not match after updates. In a DBMS, one copy means one update, so data stays consistent.
   • Access: in a file system, a new program is needed for each new question. A DBMS answers questions with queries in a language such as SQL.
   • Security/sharing: a file system makes it hard to set user-wise permissions. A DBMS gives each user only the rights they need.
   • Dependence: in a file system, programs break when the file structure changes. In a DBMS, programs use the DBMS, which hides the storage details.

   (Any three.)

5. 5 marks

   A hospital keeps the following file:

   PATIENT_VISIT(PatientID, PName, Phone, DoctorID, DoctorName, Specialisation, VisitDate, Fee)

   The same patient visits many times, and each doctor sees many patients.

   (a) Identify two kinds of redundancy in this file. (b) Redesign it as three related tables. (c) Mark the primary key and any foreign keys of each table.

   Answer

   (a) The patient's name and phone are repeated in every visit of that patient. The doctor's name and specialisation are repeated in every visit handled by that doctor.

   (b) and (c):

   • PATIENT(PatientID, PName, Phone). Primary key: PatientID.
   • DOCTOR(DoctorID, DoctorName, Specialisation). Primary key: DoctorID.
   • VISIT(PatientID, DoctorID, VisitDate, Fee). Composite primary key: (PatientID, DoctorID, VisitDate). Foreign keys: PatientID -> PATIENT(PatientID) and DoctorID -> DOCTOR(DoctorID).

   (A separate VisitID as the primary key of VISIT is also acceptable; PatientID and DoctorID then remain foreign keys.)

   Now each patient's and doctor's details are stored once, and a correction is made in one place only.

6. 5 marks

   Define the following with reference to the relation CRICKETER(CapNo, Name, State, Role, Matches) given below: (a) relation (b) attribute (c) tuple (d) domain (e) degree and cardinality. Also identify a suitable primary key.

   • (101, Arjun, Punjab, Batter, 42)
   • (102, Kiran, Karnataka, Bowler, 37)
   • (103, Sameer, Gujarat, All-rounder, 55)
   Answer

   (a) Relation: a table of data with uniquely named columns and distinct rows. CRICKETER is the relation.

   (b) Attribute: a column. CapNo, Name, State, Role and Matches are attributes.

   (c) Tuple: a row. (102, Kiran, Karnataka, Bowler, 37) is one tuple.

   (d) Domain: the set of allowed values for an attribute. The domain of Role is {Batter, Bowler, All-rounder, Wicketkeeper}; the domain of Matches is non-negative whole numbers.

   (e) Degree = 5 (attributes); cardinality = 3 (tuples).

   Primary key: CapNo, since each player's cap number is unique and never empty. Names and states can repeat.

7. 3 marks

   Explain the three properties of a relation and state why NULL is needed.

   Answer

   1. Attributes: each attribute has a unique name, and the order of attributes is immaterial.
   2. Tuples: each tuple is distinct (no two rows identical in all attributes), and the order of tuples is immaterial.
   3. Values: all values in a column come from the same domain; each value is atomic (one indivisible value per cell, so no list of phone numbers in one cell).

   NULL is needed because sometimes a value is unknown or not applicable, for example a guardian's email not yet given, or a 'Spouse' column for an unmarried person. NULL marks this without inventing a fake value like 0 or a blank.

Case-based

Practicals

Model answer
1. Degree = 5, Cardinality = 8
2. Candidate keys: UserID, MobileNo, Email
   - each is unique for every user and must always be filled
   - Name is not (two users are named Rahul Sharma)
   - City is not (three users live in Pune)
3. Primary key: UserID - short, assigned by the system, never changes
   (mobile numbers and emails can be changed by the user)
4. Alternate keys: MobileNo, Email

# Which attributes of STUDENT can be candidate keys?
header = ["AdmNo", "RollNo", "SName", "Section", "AadhaarNo", "Email"]
rows = [
    [1101, 1, "Aditi Rao",    "A", "4521 7788 1290", "aditi@mail.com"],
    [1102, 2, "Kabir Singh",  "A", "6123 4590 7781", "kabir@mail.com"],
    [1103, 1, "Meher Gill",   "B", "7712 3409 5566", "meher@mail.com"],
    [1104, 2, "Rohan Das",    "B", "3390 1287 4410", "rohan@mail.com"],
    [1105, 3, "Aditi Rao",    "B", "8845 2210 9034", "aditi.r@mail.com"],
]

def can_be_key(col):
    values = [r[col] for r in rows]
    no_null = None not in values
    unique = len(values) == len(set(values))
    return no_null and unique

for i, name in enumerate(header):
    print(f"{name:10}", "candidate key" if can_be_key(i) else "-")

AdmNo      candidate key
RollNo     -
SName      -
Section    -
AadhaarNo  candidate key
Email      candidate key

# FEE relation: neither AdmNo nor Month is unique on its own
fee = [
    (1101, "April", 2500), (1102, "April", 2500),
    (1101, "May",   2500), (1102, "May",   2700),
    (1103, "April", 2500),
]
adm = [r[0] for r in fee]
month = [r[1] for r in fee]
pair = [(r[0], r[1]) for r in fee]

print("AdmNo unique?         ", len(adm) == len(set(adm)))
print("Month unique?         ", len(month) == len(set(month)))
print("(AdmNo, Month) unique?", len(pair) == len(set(pair)))

AdmNo unique?          False
Month unique?          False
(AdmNo, Month) unique? True

# Check the foreign key STUDENT.ParentID -> PARENT.ParentID
parent_ids = [501, 502, 503, 504]                 # primary key of PARENT
students = [                                       # (AdmNo, SName, ParentID)
    (1101, "Aditi Rao", 501),
    (1102, "Kabir Singh", 502),
    (1103, "Meher Gill", 509),
    (1104, "Rohan Das", None),
    (1105, "Ishaan Rao", 501),
]
for adm, name, pid in students:
    if pid is None:
        status = "NULL (allowed, parent not yet linked)"
    elif pid in parent_ids:
        status = "OK"
    else:
        status = "INVALID - no such ParentID in PARENT"
    print(adm, name, pid, "->", status)

1101 Aditi Rao 501 -> OK
1102 Kabir Singh 502 -> OK
1103 Meher Gill 509 -> INVALID - no such ParentID in PARENT
1104 Rohan Das None -> NULL (allowed, parent not yet linked)
1105 Ishaan Rao 501 -> OK

# File-system style: the fee file repeats the student's name every month
fee_file = [
    (1101, "Aditi Rao",   "April", 2500),
    (1101, "Aditi Rao",   "May",   2500),
    (1101, "Aditi Rao",   "June",  2500),
    (1102, "Kabir Singh", "April", 2500),
    (1102, "Kabir Singh", "May",   2500),
]
names_stored = len(fee_file)
distinct_students = len(set(r[0] for r in fee_file))
print("Names stored in fee file :", names_stored)
print("Students actually present:", distinct_students)
print("Redundant copies of names:", names_stored - distinct_students)

# Office corrects a spelling in its own file, but not in the fee file
office_file = {1101: "Aditi Rao"}
office_file[1101] = "Aditee Rao"
fee_names = set(r[1] for r in fee_file if r[0] == 1101)
print("Office file says:", office_file[1101])
print("Fee file says   :", fee_names)
print("Inconsistent?   :", office_file[1101] not in fee_names)

Names stored in fee file : 5
Students actually present: 2
Redundant copies of names: 3
Office file says: Aditee Rao
Fee file says   : {'Aditi Rao'}
Inconsistent?   : True

Model answer
MEMBER(AdmNo, MName, Class)
   Primary key: AdmNo                      Degree 3
BOOK(BookID, Title, Author)
   Primary key: BookID                     Degree 3
ISSUE(AdmNo, BookID, IssueDate, ReturnDate)
   Primary key: (AdmNo, BookID, IssueDate)  - composite
   Foreign keys: AdmNo  -> MEMBER(AdmNo)
                 BookID -> BOOK(BookID)    Degree 4

ReturnDate may be NULL until the book is returned.
Member and book details are now stored once each.

Common mistakes

• Confusing degree and cardinality. Correction: degree counts columns (attributes); cardinality counts rows (tuples). The header row is never counted in cardinality.

• Writing that a table can have several primary keys. Correction: a table can have several candidate keys but only one primary key; that key may be composite.

• Choosing Name as a primary key. Correction: names can repeat, so they cannot uniquely identify rows. Use an ID such as AdmNo.

• Saying a primary key can be NULL "if the value is not known yet". Correction: a primary key can never be NULL.

• Thinking a foreign key must be unique in its own table. Correction: foreign key values can repeat (many students, one parent); they must only match an existing primary key value.

• Treating NULL as 0 or as a blank space. Correction: NULL means unknown or not applicable; 0 and " " are actual values.

• Mixing up redundancy and inconsistency. Correction: redundancy is duplication; inconsistency is duplicates that no longer match. Redundancy leads to inconsistency.

• Calling the database schema the data itself. Correction: the schema is the structure; the data at a given time is the instance.