CSPIQ · AP Computer Science Principles · Lesson 20 of 25
CSPIQ · AP Computer Science Principles

Lesson 20: Beneficial & Harmful Effects; Crowdsourcing

Big Idea 5 (IOC) · Phase 5

Objectives

Warm-Up

GPS navigation: obviously good, right? Saves fuel, rescues the lost, gets ambulances there faster.

Also: it routes cut-through traffic onto quiet residential streets, atrophies people's own navigation skills, enables stalking via location tracking, and once directed enough tourists up a dirt fire road that a town had to petition the map company.

Every computing innovation is like this. The exam will never ask you whether a technology is good or bad — it will ask you to demonstrate that it's both. The single most reliable IOC instinct: for any innovation, you can name a benefit, a harm, and a group affected differently than the majority. Train that reflex and Big Idea 5 becomes the easiest quarter of the exam.


Core Concept

Effects: beneficial, harmful, and usually both

The CED's foundational claims about computing innovations:

  1. People create computing innovations for a purpose — but innovations routinely produce effects beyond that purpose.
  2. An innovation's effects can be beneficial and harmful at the same time, often to different people. Social media connects distant friends (benefit) and enables harassment at scale (harm) — same platform, same features.
  3. Unintended consequences are effects the creators didn't foresee or intend. Not a bug, not a design flaw — a use or outcome outside the plan. A video-call app built for business meetings becomes a lifeline for remote grandparents (unintended benefit) and a new venue for exam cheating (unintended harm).
  4. How an innovation is used matters — the same technology serves different purposes in different hands, some far from what its creators imagined. And effects change as use scales: a photo-tagging feature is cute with 100 users and a surveillance concern with a billion.

The analysis template (this is what passage questions grade):

Question GPS example
Purpose (intended) Help drivers navigate efficiently
Function (what it does) Computes routes from location + map + traffic data
Beneficial effect Faster routes, fuel savings, emergency response
Harmful effect Residential cut-through traffic; location privacy risks
Who's affected differently? Residents of "shortcut" streets; people being tracked

Effect ≠ purpose, harm ≠ malice

Two precision points the exam tests:

Crowdsourcing: many hands, via the Internet

Crowdsourcing = obtaining input, help, or resources from a large group of people via the Internet. The crowd supplies what would otherwise need employees, experts, or luck:

Citizen science is the flavor the CED names specifically: members of the public participating in real scientific research — counting birds for population studies, classifying galaxy photos, reporting rainfall, folding proteins in a game. Contributors need no credentials; scale is the superpower: a million amateur observations reach places and timescales no research team could.

Trade-offs (exam-ready): crowdsourced data is abundant but variable in quality and reliability (untrained contributors, spam, bias toward who participates — Lesson 5's collection bias returns); projects need verification mechanisms (cross-checking multiple reports, expert review, reputation systems). The correct exam answer about crowdsourcing's weakness is almost always data quality/verification, not "too few participants."


Worked Examples

Example 1 (easy): Both sides, fast

Problem: Name one beneficial and one harmful effect of ride-sharing apps.

Solution: Benefit: flexible transportation availability, especially where taxis were scarce; income opportunities with low entry barriers. Harm: increased congestion in city centers; precarious income and no traditional employment protections for drivers. (Many correct pairs exist.)

Interpretation: Practice until producing a benefit/harm pair takes ten seconds for ANY innovation. That reflex answers a plurality of IOC questions.

Example 2 (medium): Classify the consequence

Problem: A fitness-tracking app publishes anonymized heat-maps of popular running routes. Journalists later discover the maps reveal the locations of secret military bases (soldiers jog around them). Classify this outcome.

Solution: An unintended harmful consequence. The feature's purpose was helping runners find routes; the designers did not foresee that aggregated data would expose sensitive locations. It's not malice, and calling it "the app's purpose" confuses effect with intent. (This scenario closely mirrors a real 2018 incident — the exam paraphrases such cases.)

Interpretation: The classification vocabulary — intended/unintended × beneficial/harmful — is the actual tested content. Practice sorting scenarios into that 2×2.

Example 3 (medium): Citizen science or not?

Problem: Which is citizen science? (i) A university hires lab technicians to count pollen samples. (ii) Thousands of volunteers photograph and upload local moth species to a biodiversity database used by ecologists. (iii) A company pays users $5 to test its app.

Solution: (ii) — public volunteers contributing observations to genuine scientific research. (i) is ordinary employment (credentialed staff, not the public); (iii) is paid user testing — crowdsourced, arguably, but not science.

Interpretation: Citizen science needs both halves: citizens (public, typically volunteers) and science (real research use). Distractors drop one half.

Example 4 (AP-style): The scale twist

Problem: A neighborhood app lets residents report suspicious activity. With 50 users, it helps neighbors watch vacationing friends' homes. With 5 million users, which NEW concern most plausibly emerges?

(A) The app's servers use more electricity (B) Reports may increasingly reflect reporters' biases, amplifying unfair suspicion toward certain groups at large scale (C) The app can no longer transmit data using TCP (D) Users forget their passwords more often

Solution: (B). Scale transforms character: individually minor bias becomes a systematic pattern with real consequences when aggregated across millions of reports (this also previews Lesson 21's computing bias). (A) is true but trivial; (C)/(D) are nonsense-scale claims.

Interpretation: "What changes at scale?" is a signature IOC question. The answer is almost always a social effect that compounds — bias, misinformation spread, privacy erosion — not a technical resource limit.


Common Mistakes

  1. Picking "this technology is simply good/bad" answers. The exam's worldview is dual-effect. One-sided absolute answers ("eliminates all," "only benefits") are essentially always wrong.
  2. Confusing unintended consequences with design errors. A logic error is code misbehaving (Lesson 2). An unintended consequence is code working exactly as built, producing effects nobody planned.
  3. Requiring malice for harm. Harmful effects emerge from well-intentioned systems. Grade outcomes, not motives.
  4. Crowdsourcing = "asking a few experts." The definition requires a large group via the Internet — scale and openness are the essence.
  5. Overlooking who bears the harm. Benefits and harms land on different groups (drivers benefit; Elm Street residents pay). Answers that notice distributional effects are usually the credited ones.

Practice Problems

Question 1
A computing innovation's unintended consequences are:
Question 2
Which statement about computing innovations reflects the CED's view?
Question 3
Video-sharing platforms let anyone publish instructional content worldwide (benefit). Which is a harmful effect OF THE SAME capability?
Question 4
A translation app designed for travelers becomes essential for hospital emergency rooms communicating with patients. This is best described as:
Question 5
Crowdsourcing is:
Question 6
Which is the BEST example of citizen science?
Question 7
The most significant reliability concern with crowdsourced data is:
Question 8
A traffic app crowdsources congestion data from users' phones and reroutes drivers. Residents of small streets near highways see traffic surge. The residents' situation illustrates:
Question 9
(Select two answers.) Which claims about the effects of computing innovations are accurate?
Question 10
A disaster-relief agency asks the public to tag roads and buildings on satellite photos after an earthquake, producing usable maps within hours. The main advantage of this approach is:
Question 11
An app rewards users for reporting potholes. Reports cluster in neighborhoods with high smartphone ownership, so road repairs skew toward those areas. This outcome combines crowdsourcing with:

12 (short response, passage-style). A grocery chain deploys smart shelves that track which products shoppers pick up and reshelve, using the data to optimize store layouts. Write one beneficial effect, one harmful effect, and one unintended consequence a thoughtful analyst might anticipate.


Create PT Connection

Written Response 1 sometimes asks not just what your program does but who it's for and why it matters — purpose framing. This lesson's discipline improves that answer: state your program's intended benefit AND show you've considered its limits. A PT write-up that says "my budgeting app helps students track spending; it depends on honest manual entry, so its picture is only as complete as what users log" demonstrates exactly the mature dual-view the CED cultivates — and it inoculates you against overclaiming, the most common written-response weakness (Lesson 5's scoped-conclusion sentence, one more time).

Mini practice: write the benefit/harm/who-differs triple for your PT idea. Even a dice game has one (fun and probability practice / could normalize gambling mechanics / players who've struggled with gambling). If your PT can't hurt anyone, say who it simply leaves out — that's next lesson's digital divide, and it counts.


Show answer key & explanations

(g) Answer Key

1. (C). Definition: beyond purpose + unforeseen. (A) contradicts "unintended"; (B) confuses consequences with code errors.

2. (B). The dual-effect doctrine — the single most load-bearing IOC claim. (A), (C), (D) each deny a CED claim (dual effects, harm-without-malice, scale matters).

3. (B). Same mechanism (open global publishing), opposite valence — the paired-effect structure the exam rewards. (A), (C), (D) are trivial or unrelated to the capability named.

4. (A). Unplanned use, positive outcome: unintended + beneficial. The 2×2 vocabulary doing its job. (D) fails "designed for travelers."

5. (B). Definition, near-verbatim. Large group + via the Internet. (A) is small and closed; (D) is Lesson 19 wearing a similar word.

6. (B). Public volunteers + real research. (A) is professional science; (C) is a data transaction, no public participation in the research; (D) is just science.

7. (B). Quality/verification — the standard credited answer. (A) is backwards (abundance is the strength).

8. (C). Distributional analysis: beneficiaries (drivers) ≠ burden-bearers (residents). No error, no malice — an outcome pattern. This is Common Mistake 5 as a question.

9. (A) and (C). Effects exceed purpose; use shapes effects. (B) denies scale effects; (D) denies dual effects — both anti-CED.

10. (B). Scale and speed — crowdsourcing's genuine advantage. (A) overclaims accuracy (usually the reverse); (C) reverses the verification lesson — crowd data needs more checking, not less.

11. (A). Participation bias at collection (Lesson 5's pothole example, now formally in its IOC home). The crowd isn't representative; the data — and the repairs — inherit the skew.

12. (Model answer.) Benefit: better-stocked, easier-to-navigate stores; less shelf-checking labor. Harm: shopper behavior is tracked without meaningful awareness — a privacy erosion (individuals' hesitation patterns are recorded). Unintended consequence (either direction works): the layout optimization may deliberately place essentials to maximize walking past temptation items, increasing impulse spending — or the tracking data could be repurposed (sold, subpoenaed) for uses never envisioned by the shelf designers. Any well-reasoned triple with a genuine unforeseen-outcome flavor earns credit.

Answer letter distribution check: C, B, B, A, B, B, B, C, A+C, B, A — singles: A×2, B×6, C×2, D×0 + multi (A,C). B-clustered (definitional gravity again) — noted; L21–L23 keys will deliberately seat correct answers on C/D. Cumulative through L20 ≈ A 23%, B 32%, C 26%, D 19%.


Exam tip: On every IOC question, pre-filter the choices: strike anything absolute ("only," "never," "eliminates all," "guarantees") and anything requiring malice for harm. What survives is usually two candidates, one of which ignores scale or distribution — and the one that acknowledges both effects, unevenly distributed is the credited answer far more often than chance.

← All lessons
Lesson 21 ›
Score: 0/0 correct