AP CSP Unit 2 FRQ Practice: Written Response

Written Response · Unit 2 · Data

Note: The AP CSP exam is all multiple-choice; extended writing appears in the Create Performance Task. This question gives you practice with that style of written response.

A student has a dataset of songs. Each record includes the song's title, artist, genre, duration (in seconds), and number of plays. The audio for each song is also stored as a separate file. The student writes a program to analyze the dataset.

Describe how the program could be used to extract information from this dataset to answer a specific question. Name one data-processing step it would use, and identify one piece of metadata in the records.

✓ Model answer

The program could answer a question such as "Which genre is the most popular?" by filtering and aggregating the data — for example, grouping records by genre and adding up the number of plays in each group, then finding the largest total. This turns the raw data into information (the most-played genre). A piece of metadata in the records is the song's genre (or duration/artist) — descriptive data about each song rather than the audio itself.

Why it scores: States a specific question, names a real data-processing step (filter/group/aggregate), AND identifies a valid piece of metadata. A vague "the program looks at the data" without a step or a metadata example would lose credit.

The audio files are large. Explain the difference between lossless and lossy compression, and state which would be more appropriate for compressing the audio files, with a justification.

✓ Model answer

Lossless compression reduces file size without losing any data, so the original can be reconstructed exactly; lossy compression discards some data permanently to achieve a much smaller size. For the large audio files, lossy compression (such as MP3) is usually more appropriate, because it greatly reduces the storage needed and a small loss of audio quality is generally acceptable to listeners — an important benefit when storing many songs.

Why it scores: Correctly contrasts lossless (recoverable) with lossy (permanent loss, smaller), chooses lossy for the audio, AND justifies it with the size-vs-quality tradeoff in context. Choosing lossless without justification, or reversing the definitions, would lose credit.

The student notices that songs with longer durations tend to have more plays, and concludes that making a song longer will cause it to get more plays. Explain why this conclusion is not justified, and describe one responsible way to present the finding.

✓ Model answer

The conclusion is not justified because correlation does not imply causation: longer duration and more plays are associated in the data, but that does not prove that length causes the plays. A third factor — such as the songs being by more popular artists, or in a genre people replay — could explain both. A responsible way to present the finding is to report it as an association, not a cause ("longer songs in this dataset tended to have more plays"), and to note the limitation and the need for further investigation before making any claim of causation.

Why it scores: Explains that correlation is not causation and gives a plausible lurking factor, AND describes presenting the result honestly as an association with its limitations. Simply restating "correlation isn't causation" without the reasoning or the responsible-presentation part would earn less.

How to write strong AP CSP responses

Name a concrete data-processing step. Filtering, sorting, grouping, or aggregating — not just "looking at the data."
Define compression precisely. Lossless = fully recoverable; lossy = smaller but permanent loss.
Justify your choice in context. Tie lossy vs. lossless to the size-vs-quality tradeoff for the specific files.
Distinguish data from information and identify metadata correctly (data about data).
Never confuse correlation with causation, and present findings responsibly with their limitations.