AP COMPUTER SCIENCE PRINCIPLES

COMPREHENSIVE EXAM STUDY SHEET

Big Ideas: Creative Development · Data · Algorithms & Programming · Computing Systems & Networks · Impact of Computing

1. CREATIVE DEVELOPMENT (CRD)

Core Vocabulary

Term	Definition
Program	A sequence of statements executed by a computer to complete a task
Algorithm	A finite, precise, effective step-by-step process for solving a problem
Abstraction	Reducing complexity by hiding irrelevant details; focusing on what matters
Decomposition	Breaking a complex problem into smaller, manageable sub-problems
Modularity	Dividing a program into independent, interchangeable components (procedures)
Iteration (dev)	Repeatedly revising and improving a program over time based on testing/feedback
Collaboration	Working with others using communication, consensus-building, and shared credit

Program Development Lifecycle

Typical Development Process

1. Investigate & Reflect — Define the problem; identify user needs; research existing solutions 2. Design — Plan the algorithm; create flowcharts or pseudocode; select data structures 3. Implement — Write code; use procedures and abstractions to manage complexity 4. Test — Run with normal, edge, and invalid inputs; identify and fix errors 5. Iterate — Revisit any earlier step based on test results; refine until requirements are met ★ Development is rarely linear — jumping between phases is normal and expected

Errors (Know all 3 types for exam)

Error Type	Description	Example
Syntax	Violates language rules; program won’t run/compile	Missing parenthesis; misspelled keyword
Runtime	Occurs during execution; crashes the program	Dividing by zero; accessing out-of-bounds index
Logic	Program runs but produces incorrect results	Using < instead of ≤; off-by-one in loop
Overflow	Value exceeds the maximum storable number	Integer wrapping around to negative

Abstraction Levels

Abstraction hides complexity at multiple levels:

• High-level language → hides machine code details

• Procedures/functions → hide implementation details

• Lists/data structures → hide individual element management

• APIs → hide how a service or library works internally

• Internet protocols → hide physical transmission details

2. DATA & INFORMATION (DAT)

Bits, Bytes & Number Systems

Concept	Details
Bit	Smallest unit of data; value is 0 or 1
Byte	8 bits; can represent 2⁸ = 256 different values (0–255)
n bits represent	2ⁿ unique values (e.g., 8 bits → 256, 16 bits → 65,536)
Binary (base-2)	Uses digits 0 and 1; each position is a power of 2
Hexadecimal	Base-16 (0–9, A–F); often used to represent binary more compactly
Overflow	When a value exceeds the bits available; wraps around or causes error

Binary ↔ Decimal Conversion (Must Know!)
Binary place values (right to left): 128	64	32	16	8	4	2	1 Binary 10110101 → 128+0+32+16+0+4+0+1 = 181 (decimal) Decimal 45 → 32+8+4+1 = 101101 (binary) Tip: 8-bit range = 0 to 255	Adding a bit doubles possible values Tip: If a number requires more bits than available → OVERFLOW

Data Types & Representation

Data Type	Stored As	Examples
Integer	Binary number	42, -7, 0
Float / Real	Approximation in binary (IEEE 754)	3.14, -0.001
Boolean	1 bit (0 = false, 1 = true)	true, false
String / Text	ASCII or Unicode code points	“Hello”, ‘A’
Color (RGB)	3 bytes (R, G, B each 0–255)	Red = (255,0,0)
Sound	Sampled amplitude at intervals	MP3, WAV files
Image	Pixel grid of color values	PNG, JPEG, BMP

Analog vs. Digital Data

Analog	Digital
Continuous values (infinite precision)	Discrete values (finite, sampled)
Sound waves, temperature, light intensity	MP3, integers, pixel values
Cannot be stored directly by computer	Stored as binary patterns
Converted via ADC (Analog-to-Digital Converter)	Sampling rate affects quality

Data Compression

Type	Key Facts
Lossless	ALL original data recovered on decompression. Examples: ZIP, PNG, GIF. Used for text, code, executables.
Lossy	Some data permanently discarded. Examples: JPEG, MP3, MP4. Smaller file; quality reduced. CANNOT recover the original.
Run-Length Encoding	Lossless — replaces repeated values with count+value. e.g. AAABBC → 3A2B1C
Trade-off	Compression reduces file size (storage/bandwidth) at cost of time or quality

Metadata & Data Analysis

• Metadata: data ABOUT data (e.g., file size, creation date, GPS location in a photo, author of a document)

• Metadata can reveal sensitive information even if the main content is private

• Data abstraction: using lists, databases, and structures to manage complex data sets

• Cleaning data: removing duplicates, fixing errors, handling missing values before analysis

• Visualizations (charts, graphs) reveal patterns; can also be used to mislead

• Correlation ≠ Causation: two variables moving together does not mean one causes the other

• Bias in data: if training/sample data is biased, conclusions and models will be too

3. ALGORITHMS & PROGRAMMING (AAP)

AP Pseudocode Reference — Assignment & I/O

AP Pseudocode Syntax
a ← expression // Assign value to variable a a ← b // Assign value of b to a DISPLAY(expression) // Output value to screen INPUT() // Get input from user a ← INPUT() // Store user input in a

AP Pseudocode — Arithmetic & Comparisons

Operator	Meaning	Example / Notes
a + b	Addition	3 + 4 → 7
a - b	Subtraction	10 - 3 → 7
a * b	Multiplication	2 * 5 → 10
a / b	Division (real)	7 / 2 → 3.5
a MOD b	Remainder	7 MOD 2 → 1 (use for even/odd)
a = b	Equal to (comparison)	NOT assignment in conditions
a ≠ b	Not equal to	a NOT EQUAL b
a < b, a > b	Less / Greater than	strict
a ≤ b, a ≥ b	Less/Greater or equal	inclusive

AP Pseudocode — Boolean & Selection

AP Pseudocode Syntax
NOT condition // True if condition is false cond1 AND cond2 // True only if BOTH true cond1 OR cond2 // True if AT LEAST ONE true IF (condition) // Simple conditional { } IF (condition) // If-Else { } ELSE { } IF (a > 0) // Nested conditional example { IF (a > 100) { DISPLAY("big") } ELSE { DISPLAY("small") } }

AP Pseudocode — Iteration (Loops)

AP Pseudocode Syntax
REPEAT n TIMES // Count-controlled loop { } REPEAT UNTIL (condition) // Condition-controlled loop (runs while condition is FALSE) { // Executes at least ONCE (like do-while) } FOR EACH item IN list // Traversal loop { // item takes each value in list sequentially }

AP Pseudocode — Lists

AP Pseudocode Syntax

myList ← [1, 2, 3, 4] // Create list myList[1] // Access FIRST element (AP CSP lists START AT INDEX 1!) myList[i] ← value // Assign value to index i INSERT(myList, i, value) // Insert value at index i; shifts elements right APPEND(myList, value) // Add value to END of list REMOVE(myList, i) // Remove element at index i; shifts elements left LENGTH(myList) // Returns number of elements FOR EACH item IN myList // Standard traversal pattern { DISPLAY(item) }

AP Pseudocode — Procedures

AP Pseudocode Syntax
PROCEDURE myProc(param1, param2) // Define procedure with parameters { RETURN expression // Optional — sends value back to caller } result ← myProc(3, 7) // Call procedure and store return value myProc(3, 7) // Call without capturing return value // Parameters are LOCAL — changes don't affect original variables unless returned

Common Algorithm Patterns

Patterns You MUST Recognize & Write
SUM: total ← 0	FOR EACH x IN list { total ← total + x } COUNT: count ← 0	FOR EACH x IN list { IF (condition) { count ← count + 1 } } FIND MAX: max ← list[1]	FOR EACH x IN list { IF (x > max) { max ← x } } FIND MIN: min ← list[1]	FOR EACH x IN list { IF (x < min) { min ← x } } FILTER: result ← []	FOR EACH x IN list { IF (condition) { APPEND(result, x) } } SEARCH (linear): found ← false	FOR EACH x IN list { IF (x = target) { found ← true } }

Searching Algorithms

	Linear Search vs Binary Search
Linear Search	Checks each element one-by-one from start to end. Works on ANY list (sorted or unsorted). O(n) worst case.
Binary Search	Cuts search space in HALF each step. Requires SORTED list. Much faster: O(log n). Mid = (low+high)/2.
When binary fails	List not sorted → binary search gives WRONG results. Always check sort requirement!
Binary search steps	1) Find midpoint 2) If match → done 3) If target < mid → search left half 4) Else → search right half 5) Repeat

Algorithm Efficiency & Complexity

Complexity	Meaning & Example
O(1) — Constant	Same time regardless of input size. e.g., accessing list[3]
O(log n) — Logarithmic	Time grows slowly. e.g., binary search on 1M items ≈ 20 steps
O(n) — Linear	Time grows with input size. e.g., linear search, single traversal
O(n²) — Quadratic	Nested loops over same data. e.g., selection sort, comparing all pairs
Heuristic	Approximate solution when exact is too slow; may not be optimal but is reasonable
Undecidable Problem	No algorithm can ALWAYS determine YES/NO (e.g., Halting Problem). Not solvable.
Unsolvable Problem	No algorithm exists that solves it for ALL inputs in all cases

Sorting Concepts (Conceptual — No Code Required)

• Selection Sort: Find the minimum, place it at start, repeat on remaining. O(n²).

• Insertion Sort: Build sorted portion one element at a time; insert each into correct position. O(n²) worst, O(n) best.

• AP exam tests UNDERSTANDING of how sorting works, not memorizing code

• Any correct sorting algorithm is equivalent in terms of correctness

Undecidability & Limits of Computing

The Halting Problem — Key Exam Concept
Definition: No algorithm can determine for ALL programs whether they will halt (finish) or run forever. This is UNDECIDABLE — proven by Alan Turing. Implication: There are problems computers CANNOT solve, no matter how fast or advanced. An algorithm that loops forever is still an algorithm — it just never terminates. ★ Exam tip: ‘No algorithm can solve X for all inputs’ = undecidable problem

4. COMPUTING SYSTEMS & NETWORKS (CSN)

Hardware Basics

Component	Function
CPU (Processor)	Executes instructions; performs arithmetic and logic
RAM (Memory)	Temporary storage for running programs; lost when powered off
Storage (HDD/SSD)	Permanent data storage; persists when powered off
Input Devices	Keyboard, mouse, microphone, camera — bring data IN
Output Devices	Monitor, printer, speakers — send data OUT
Network Interface	Hardware to connect to a network (Ethernet, Wi-Fi card)

How the Internet Works

Key Internet Concepts

Packet Switching: Data is broken into packets, each routed independently, reassembled at destination. Packets may take DIFFERENT paths and arrive OUT OF ORDER — protocols handle reassembly. Redundancy: Multiple paths exist between nodes; if one fails, data re-routes automatically. Fault Tolerance: The internet is designed to keep working even when parts fail. Scalability: The internet can grow (add nodes/connections) without redesigning the whole system. IP Address: Unique numerical label for every device on the internet (e.g., 192.168.1.1). IPv4 uses 32 bits (~4 billion addresses); IPv6 uses 128 bits (vastly more).

Network Protocols (Must Know All)

Protocol	Purpose & Key Facts
IP (Internet Protocol)	Assigns addresses; routes packets between devices. Handles addressing only.
TCP (Transmission Control Protocol)	Ensures RELIABLE delivery: packets acknowledged, retransmitted if lost, delivered IN ORDER.
UDP (User Datagram Protocol)	Faster but UNRELIABLE: no acknowledgment; used for video streaming, gaming, VoIP.
HTTP / HTTPS	Application protocol for web pages. HTTPS = encrypted with TLS/SSL.
DNS (Domain Name System)	Translates human-readable names (google.com) to IP addresses. Like a phone book.
SMTP / IMAP / POP3	Email protocols for sending (SMTP) and receiving (IMAP/POP3).
Wi-Fi / Bluetooth	Wireless physical layer protocols for local device communication.

Internet Layers (Simplified)

Layer	Role	Examples
Application	User-facing services	HTTP, DNS, SMTP, FTP
Transport	End-to-end communication	TCP, UDP
Internet	Addressing & routing	IP
Network Access	Physical transmission	Ethernet, Wi-Fi, fiber

Bandwidth vs. Latency

Term	Definition & Analogy
Bandwidth	Maximum data transferred per second (bits/s). Width of the pipe. Higher = more data.
Latency	Time delay for data to travel from sender to receiver (milliseconds). Length of the pipe.
Throughput	Actual data rate achieved (≤ bandwidth due to congestion, errors, overhead).
Trade-off	High bandwidth does NOT mean low latency. Satellite has huge bandwidth but high latency.

Parallel & Distributed Computing

Parallel vs Sequential vs Distributed

Sequential Computing: Tasks run one at a time, in order. Predictable but potentially slower. Parallel Computing: Multiple tasks run SIMULTANEOUSLY on multiple processors/cores. Speeds up suitable tasks. Distributed Computing: Tasks split across multiple computers/devices on a network (e.g., SETI@home, cloud computing). ★ KEY EXAM TRAPS: • Parallel is NOT always faster — overhead of splitting/merging can slow things down. • Only PARALLELIZABLE portions benefit — sequential parts remain bottlenecks (Amdahl’s Law concept). • Speedup = Sequential time / Parallel time (approximate formula for exam). • Some problems cannot be parallelized (each step depends on the previous result).

5. CYBERSECURITY & PRIVACY

Types of Attacks

Attack	Description
Phishing	Deceptive emails/sites trick users into revealing credentials or installing malware.
Malware	Malicious software: viruses (self-replicate), ransomware (encrypts files for ransom), spyware.
Denial of Service (DoS)	Overwhelms a server with requests, making it unavailable to real users.
Distributed DoS (DDoS)	DoS using thousands of compromised computers (botnet). Much harder to stop.
SQL Injection	Attacker inserts malicious SQL code into a form input to access/corrupt a database.
Man-in-the-Middle	Attacker intercepts communication between two parties without their knowledge.
Social Engineering	Manipulating people (not systems) to reveal information. Phishing is one example.
Rogue Access Point	Fake Wi-Fi network created by attacker to intercept traffic.

Security Methods & Defenses

Defense	How It Works
Encryption	Transforms readable data (plaintext) into unreadable form (ciphertext). Only decrypted with key.
Symmetric Encryption	SAME key encrypts and decrypts. Fast; key must be securely shared.
Asymmetric Encryption (Public Key)	Public key encrypts; private key decrypts. Key pair; no need to share private key. Used in HTTPS.
Digital Signature	Proves authenticity and integrity; uses sender’s private key to sign.
SSL / TLS	Protocols that encrypt data in transit (used in HTTPS). Provides authentication + encryption.
Multi-Factor Authentication (MFA)	Requires 2+ factors: something you know (password), have (phone), or are (fingerprint).
Firewall	Monitors and filters incoming/outgoing network traffic based on security rules.
Antivirus / Anti-malware	Detects, quarantines, and removes malicious software.
Software Updates / Patches	Fix known vulnerabilities; critical for security.
Principle of Least Privilege	Users/programs get ONLY the permissions they need — nothing more.

Privacy Concepts

• PII (Personally Identifiable Information): Data that identifies a person: name, SSN, address, email, phone, IP address.

• Data Privacy: Individuals have the right to control how their personal data is collected and used.

• Cookies: Small files stored by websites to track sessions and preferences.

• Third-Party Tracking: Advertisers track users across websites using embedded scripts/pixels.

• Data Broker: Companies that collect, aggregate, and sell personal data without direct user consent.

• Terms of Service: Legal agreements users accept; often grant companies broad data usage rights.

• FERPA / COPPA / GDPR: Laws protecting student records, children’s data, and EU citizen data respectively.

6. IMPACT OF COMPUTING (IOC)

Beneficial & Harmful Impacts

Positive Impacts	Negative Impacts
Improved global communication	Privacy erosion / surveillance
Access to information & education	Cybercrime and identity theft
Medical advances / diagnostics	Automation displacing jobs
Scientific collaboration & modeling	Spread of misinformation
Accessibility tools (screen readers, etc.)	Environmental impact of data centers
Economic opportunity & e-commerce	Algorithmic bias and discrimination

Equity & Digital Divide

Key Terms

Digital Divide: Unequal access to technology (internet, devices, skills) across regions, incomes, and demographics. Causes: Cost of hardware/internet, lack of infrastructure, language barriers, disability access. Crowdsourcing: Distributed data/work collection from large online groups. Ex: Wikipedia, reCAPTCHA, citizen science. Open Source Software: Source code is publicly available; anyone can modify and distribute. Ex: Linux, Firefox. Proprietary Software: Source code is private; controlled by company. Ex: Windows, Photoshop. Creative Commons: Licensing that allows creators to specify how their work may be shared/used.

Intellectual Property & Legal Issues

• Copyright: Protects original creative works automatically upon creation. Infringement = illegal use without permission.

• Fair Use: Limited use of copyrighted material without permission (education, commentary, parody).

• Plagiarism: Using others’ work without attribution — unethical even if legal (e.g., public domain).

• Open Source License: Specifies how software may be used, modified, and distributed.

• Computing Innovation: Must specify a purpose; can be beneficial or harmful depending on use and access.

Algorithmic Bias & Ethics

• Bias enters AI/ML systems through biased training data, biased labels, or biased design choices

• Facial recognition has higher error rates for darker skin tones — documented bias

• Hiring algorithms trained on historical data can perpetuate past discrimination

• Programmers have ethical responsibilities for how their code affects users and society

• Unintended consequences: programs can have effects their creators never anticipated

Legal & Safe Computing Practices

• Computer Fraud and Abuse Act (CFAA): US law criminalizing unauthorized computer access.

• Safe Browsing: Verify URLs, avoid suspicious links, use HTTPS, keep software updated.

• Strong Passwords: Long, random, unique per site; use a password manager.

• Two-Factor Authentication: Adds a second layer beyond just a password.

• Citizen Responsibility: Report security vulnerabilities responsibly (responsible disclosure).

7. BOOLEAN LOGIC & EXPRESSIONS

Truth Tables (Must Memorize)

A	B	NOT A	A AND B	A OR B	A XOR B
T	T	F	T	T	F
T	F	F	F	T	T
F	T	T	F	T	T
F	F	T	F	F	F

Boolean Logic Key Rules

AND: TRUE only when BOTH inputs are TRUE OR: TRUE when AT LEAST ONE input is TRUE NOT: Flips the value (TRUE→FALSE, FALSE→TRUE) XOR (exclusive or): TRUE when inputs are DIFFERENT (not tested directly but good to know) Short-circuit evaluation: In A AND B — if A is false, B is never checked. In A OR B — if A is true, B is never checked. De Morgan’s Laws (conceptual): NOT(A AND B) = NOT A OR NOT B NOT(A OR B) = NOT A AND NOT B

Conditional Logic Exam Patterns

AP Pseudocode Syntax
// Pattern 1: Compound condition IF (x > 0 AND x < 100) { DISPLAY(“in range”) } // Pattern 2: Negation IF (NOT (score >= 70)) { DISPLAY(“failing”) } // Pattern 3: Equivalent to above IF (score < 70) { DISPLAY(“failing”) } // Pattern 4: Nested conditional IF (age >= 18) { IF (registered = true) { DISPLAY(“can vote”) } ELSE { DISPLAY(“not registered”) } } ELSE { DISPLAY(“too young”) }

8. CREATE PERFORMANCE TASK (CPT) — EXAM REQUIREMENTS

⚠ CRITICAL: CPT is 30% of your AP Score — Know Every Requirement
You submit: a program you wrote + a written response (approx. 750 words) + a video of program running Your program must be written INDIVIDUALLY (collaboration allowed for ideation, not writing code for you) Time allocation: approximately 12 hours in class

Required Program Components (Must ALL be present)

Requirement	What Graders Look For
Input	Program receives input from user, sensor, file, web query, or another program
Output	Program produces output visible to the user (display, audio, file, etc.)
List (or other collection)	Must USE a list to store multiple elements AND use list operations (not just declare it)
Procedure with Parameter	Student-defined procedure that takes at least one parameter AND uses it in the body
Algorithm with Sequence	Procedure body must have statements in order (sequencing)
Algorithm with Selection	Procedure body must include IF/ELSE or conditional logic
Algorithm with Iteration	Procedure body must include a loop (REPEAT, FOR EACH, WHILE, etc.)
Calls the Procedure	The defined procedure must actually be called somewhere in the program

Written Response Rows (Scoring Rubric)

Row	Topic	Key Points to Address
1	Program Purpose & Function	State the PURPOSE (why it exists, the problem it solves). Describe INPUT and OUTPUT. Demonstrate functionality via video.
2	Data Abstraction	Show the list. Explain what data is stored. Explain how the list MANAGES COMPLEXITY (what would be harder without it).
3	Managing Complexity	Explain how the list helps manage complexity. Name the list, what it stores, and WHY it’s necessary.
4	Procedural Abstraction	Identify the student-defined procedure. Describe what it does. Explain how it contributes to overall program functionality.
5	Algorithm Implementation	Describe the algorithm inside the procedure. Explain sequence, selection, AND iteration — all three must be present.
6	Testing	Describe TWO different calls to the procedure. State arguments used. Identify expected output/behavior. State the result.

Common CPT Mistakes That Cost Points

✗ List is declared but never meaningfully used (must store AND retrieve data from list) ✗ Procedure has no PARAMETERS — must have at least one parameter that affects behavior ✗ Algorithm is missing selection OR iteration — need ALL THREE in the SAME procedure ✗ Procedure is not student-written (using only built-in library functions doesn’t count) ✗ Written response says ‘what’ but not ‘why’ — always explain purpose and effect ✗ Testing section only describes one test case — need TWO different calls with different arguments ✗ Program purpose described as ‘it runs’ — must state the problem being solved ✗ Video doesn’t show both input and output — record carefully

9. EXAM STRATEGIES & CRITICAL FACTS

Must-Know Exam Traps

These Are Tested Every Year

1. AP CSP LISTS START AT INDEX 1 — not 0. list[1] is the FIRST element. 2. Binary search REQUIRES a SORTED list — using it on unsorted data gives WRONG results. 3. Lossy compression PERMANENTLY discards data — you CANNOT recover the original. 4. Parallel computing is NOT always faster — overhead and non-parallelizable code can slow things down. 5. REPEAT UNTIL runs the body FIRST, then checks condition (executes at least once). 6. Correlation does NOT imply causation — two trends moving together ≠ one causes the other. 7. The Internet was DESIGNED for fault tolerance — packets re-route around failures automatically. 8. Metadata can reveal private information even when the main content is private. 9. Adding more bits DOUBLES the number of values representable (2ⁿ). 10. An undecidable problem has NO algorithm that solves it for ALL inputs — not just unsolved yet. 11. HTTPS encrypts data in transit — HTTP does NOT. 12. DNS translates domain names to IP addresses — it does NOT provide security. 13. TCP guarantees delivery and ORDER; UDP does not — UDP is faster. 14. Open-source software is NOT necessarily free (free as in freedom, not always free as in cost). 15. Parameters are LOCAL — changes inside a procedure don’t affect variables outside (unless returned).

AP CSP Pseudocode Quick Reference Card

Assignment & I/O	Control Flow	List Operations
a ← value DISPLAY(expr) a ← INPUT() a MOD b	IF (cond) { } IF / ELSE { } REPEAT n TIMES { } REPEAT UNTIL(c) { } FOR EACH x IN L { }	list[i] INSERT(L,i,v) APPEND(L,v) REMOVE(L,i) LENGTH(L)

Key Numbers & Facts to Memorize

Fact	Value	Why It Matters
1 byte	8 bits	Fundamental unit of storage
n bits → values	2ⁿ	e.g., 8 bits = 256 values
ASCII (basic)	128 characters	7-bit encoding; 0–127
RGB color depth	24 bits (3 × 8)	16.7 million colors
IPv4 address	32 bits	≈ 4 billion unique addresses
IPv6 address	128 bits	Vastly more addresses
AP list indexing	Starts at 1	NOT 0 like most languages

MCQ Strategy

• Read the ENTIRE question before looking at answers — trap answers often match partial readings

• Eliminate obviously wrong answers first, then compare remaining choices

• For pseudocode questions: TRACE the code manually with a simple example

• For algorithm questions: ask what happens with 0 elements, 1 element, and maximum elements

• For impact questions: look for the most DIRECT effect, not a secondary consequence

• For network/internet questions: remember the Internet was designed for FAULT TOLERANCE

• Time: 2 hours for 70 MCQ ≈ 1 min 42 sec per question — don’t get stuck; flag and return