AP CSP Topic 2.3 Guided Notes - Extracting Information from Data
Big Idea 2: Data · Topic 2.3 · Guided Notes (Student)
Extracting Information from Data — Guided Notes
Fill these in during class or catch up here if you were absent. Print this page or work on paper — then check yourself with the CFUs on the Topic 2.3 page.
Print these notesTopic 2.3 lesson pageAll CSP topics
Day 1: The Numbers Are Talking
Today’s objectives
- Describe the information that can be extracted from data — trends, connections, and problems to address — and explain why a correlation alone never establishes causation (LO DAT-2.A)
- Describe the information that can be extracted from metadata, and explain why changes to metadata never change the primary data (LO DAT-2.B)
Bell ringer
A (fictional) slice of our class's music-streaming survey: Ava — 412 streams, 6.1 hrs sleep, grade 88. Ben — 95 streams, 8.4 hrs, grade 91. Cam — 388 streams, 6.0 hrs, grade 74. Dee — 120 streams, 8.1 hrs, grade 90. Eli — 401 streams, 6.3 hrs, grade 79.
Write two lists: (1) claims this table SUPPORTS, and (2) claims it CANNOT prove. Does heavy streaming cause less sleep? How would you find out for sure?
01. From Data to Information
Key Vocabulary (LO DAT-2.A)
| Term | Definition (write it) |
|---|---|
| Data | |
| Information | |
| Correlation | |
| Causal relationship |
What You Can Claim — and When You Can't
- Data provide opportunities for identifying trends, making connections, and .
- A correlation found in data does not necessarily indicate .
- When a single source can't support a conclusion, it may be necessary to .
Watch out — “The Data Proves It Caused It”
Myth: If two variables rise and fall together in a dataset, one of them must be causing the other.
Explain why this is wrong:
One Source vs. Combined Sources
A single source
Combined sources
- Often, a single source does not contain .
- Combining data from a variety of sources lets you .
AP TIP: When a stem says one dataset 'is not enough to conclude,' the credited move is combining it with a DIFFERENT source — not collecting more of the same.
Stop and think
- Using the bell-ringer table: write one piece of INFORMATION the raw numbers genuinely support, and name the data it was extracted from.
- A headline reads: 'Study finds students who stream more music get lower grades — streaming hurts GPA.' Which part is correlation, which part is the leap, and what additional research would be needed?
- A principal wants to know WHY streaming spiked in May, but the app export alone can't say. Name a second source to combine with it, and state the conclusion the combination could support.
Answer in complete sentences. Then check yourself with the matching CFUs on the Topic 2.3 page.
02. Metadata: Data About Data
Key Vocabulary (LO DAT-2.B)
| Term | Definition (write it) |
|---|---|
| Metadata | |
| Primary data |
What Metadata Is For
- Metadata are used for finding, , and managing information.
- Metadata increase the effective use of data by .
- If you edit or delete a photo's metadata, the photo itself .
Data or Metadata?
Five items from one phone. Classify each: is it the primary data, or metadata describing it?
Complete the empty cells.
| Item | Data or metadata? | How you can tell |
|---|---|---|
| The pixels of a vacation photo | It IS the content — the thing itself | |
| The photo's date taken and file size | Data ABOUT the photo, not the picture itself | |
| The audio of a song | Primary data | The content you actually hear |
| The song's artist, length, and date added | Describes the song — lets the app find, sort, and organize it | |
| You delete the photo's location tag — the picture is now... | Changes and deletions to metadata do not change the primary data |
Deep Dive · Beyond the AP Exam
Deep Dive: The Photo That Knew Too Much
- A smartphone photo's hidden metadata can include .
- Stripping a photo's metadata leaves the image .
Stop and think
- A song file's metadata lists artist, length, and date added. In two sentences, explain the difference between the song's primary data and its metadata.
- You remove the location tag from a photo before posting it. A classmate worries the picture is now damaged. Correct them, citing what changes to metadata do to primary data.
- Your music app instantly builds a 'Most played this month' playlist. Which metadata makes that possible, and which job — finding, organizing, or managing — is it doing?
Answer in complete sentences. Then check yourself with the matching CFUs on the Topic 2.3 page.
Today in one box
- Information is the collection of facts and patterns EXTRACTED from data — data is the raw material
- Correlation is never proof of causation — additional research is required
- One source often can't support a conclusion; combining sources can
- Metadata is data about data — it powers finding, organizing, and managing, and editing it never changes the primary data
Before next class: Tomorrow's disaster preview: a survey asked 'What city are you from?' — and six students from the SAME city produced five different answers. Before tomorrow, predict: what happens when a computer tries to count them?
Day 2: When Data Gets Messy
Today’s objectives
- Identify the challenges data sets pose regardless of size, and explain how cleaning makes data uniform without changing its meaning (LO DAT-2.C)
- Explain how bias, data-set size, parallel systems, and scalability shape what information can be extracted from data (LO DAT-2.C)
Bell ringer
Yesterday's teaser, delivered. The survey's 'What city are you from?' column contains: KC · Kansas City · kansas city MO · K.C. · (blank) · Kansas Citty.
A program counts students per city. What count does it report for Kansas City — and why? List every SEPARATE problem you can find in this column, then write what a fixed version would look like.
01. Why Real Data Is Messy
Key Vocabulary (LO DAT-2.C)
| Term | Definition (write it) |
|---|---|
| Cleaning data | |
| Invalid data | |
| Bias | |
| Parallel systems | |
| Scalability |
Every Dataset Poses Challenges — Regardless of Size
- Data sets pose challenges of size.
- When users type into an open field, the data may not be .
- The ability to process data depends on the capabilities of .
Cleaning the Hometown Column
The bell-ringer column, triaged. For each raw entry: name the disease, then decide what a cleaner may — and may not — do.
Complete the empty cells.
| Raw entry | What's wrong | Cleaning decision |
|---|---|---|
| KC · Kansas City · kansas city MO | Non-uniform: one city, three formats | |
| Kansas Citty | Misspelling — still unmistakably the same city | Correct the spelling to 'Kansas City' — meaning unchanged |
| (blank) | Incomplete data — the value is missing | |
| Age column: −3 | Invalid data — no age can be negative |
Stop and think
- Define cleaning data in your own words, then explain why replacing 'KC' with 'Kansas City' is cleaning but replacing a BLANK with 'Kansas City' is not.
- A birth-year column contains the value 2091. Is that non-uniform, incomplete, or invalid data? What should a careful cleaner do with it?
- Two clubs kept sign-up lists with different column names and formats, and the school wants one member count. Name the challenge this creates and what must happen before the combined list can be counted.
Answer in complete sentences. Then check yourself with the matching CFUs on the Topic 2.3 page.
02. Bias, Size, and Scale
Watch out — “Just Collect More Data”
Myth: If survey results look biased, the fix is collecting more data — with enough responses, the bias will average out.
Explain why this is wrong:
Small Data vs. Big Data
A small data set
A large data set
- The size of a data set affects .
- Large data sets are difficult to process on a single computer and may require .
- Scalability matters because a system's computational capacity affects .
AP TIP: Size changes HOW MUCH information you can extract and WHAT SYSTEMS you need. It never repairs a biased source.
Deep Dive · Beyond the AP Exam
Deep Dive: How a Billion Rows Get Counted
- The big-data pattern is: split the data into chunks, , then merge the partial results.
- Companies handle growing datasets by 'scaling out,' which means .
Stop and think
- A fitness app reports 'teenagers exercise 90 minutes a day' using data only from people who chose to install a fitness app. Explain the bias — and why doubling the number of users would not fix it.
- A five-row dataset and a five-million-row dataset both track music streaming. What does the size difference change about the information that can be extracted from each?
- A biologist's laptop cannot process a genome dataset. Name the challenge, the kind of system that addresses it, and what scalability has to do with choosing that system.
Answer in complete sentences. Then check yourself with the matching CFUs on the Topic 2.3 page.
Common AP Traps
Three ways Topic 2.3 loses points on the exam — one minute now, real points in May.
Correlation ≠ causation — in your own words:
Metadata edits touch the file — in your own words:
More data cures bias — in your own words:
Extracting Information from Data, in One Slide
- Information is the collection of facts and patterns extracted from data — data is the raw material.
- Correlation never proves causation by itself; conclusions often need combined sources.
- Metadata is data about data: it powers finding, organizing, and managing — and editing it never changes the primary data.
- Every dataset needs cleaning judgment: make non-uniform data uniform WITHOUT changing meaning; never invent missing or 'fix' invalid values silently.
- Bias comes from the source and survives more collection; size changes what's extractable and may demand parallel, scalable systems.
Exit check — I can…
- ☐ tell data from information and extract a defensible claim from a raw table (LO DAT-2.A)
- ☐ explain why a correlation alone never proves causation, and when sources must be combined (LO DAT-2.A)
- ☐ identify a file's metadata and explain what changing it does — and doesn't do — to the primary data (LO DAT-2.B)
- ☐ spot non-uniform, incomplete, invalid, and biased data, and choose the right response — from cleaning to parallel systems (LO DAT-2.C)
Get in Touch
Whether you're a student, parent, or teacher — I'd love to hear from you.
Just want free AP CS resources?
Enter your email below and check the subscribe box — no message needed. Students get daily practice questions and study tips. Teachers get curriculum resources and teaching strategies.
Message Sent!
Thanks for reaching out. I'll get back to you within 24 hours.
Prefer email? Reach me directly at [email protected]