Chapter 8 Quiz: Algorithm Literacy

Instructions: Choose the best answer for each question. The answer key with explanations follows the questions.

1. Which of the following most accurately describes how a recommendation algorithm works?

a) It evaluates content quality on a scale of 1–10 and distributes the best content to the most users b) It uses input signals to predict which content will generate the engagement behaviors the platform wants to maximize c) It distributes content proportionally based on how many followers a creator has d) It randomizes distribution and then adjusts based on which content does well by chance

2. TikTok's For You Page algorithm treats which of the following as its most important initial ranking signal?

a) The number of likes a video receives in its first 24 hours b) The creator's follower count c) Video watch completion rate from the initial seed audience d) The number of hashtags used in the caption

3. On YouTube, the relationship between click-through rate (CTR) and watch time can best be described as:

a) CTR is more important than watch time for all channel sizes b) Watch time is the only metric that matters for algorithmic distribution c) Both signals matter: CTR affects initial distribution, watch time determines whether distribution continues d) Neither matters much; subscriber count is the primary driver of recommendations

4. Why do platforms typically refuse to publish the exact mechanics of their recommendation algorithms?

a) Transparency laws in most countries prohibit it b) The algorithms are too simple to be worth publishing c) Competitive advantage, abuse prevention, and legal liability concerns d) Platforms want to maintain the mystique that drives creator motivation

5. A creator posts two videos. Video A is 15 seconds long with 88% completion rate. Video B is 90 seconds long with 42% completion rate. On TikTok, which video is likely to receive broader distribution?

a) Video B, because it has more total watch time per view b) Video A, because completion rate is weighted more heavily than total watch time on TikTok c) They will receive equal distribution because TikTok doesn't measure completion rate d) Video B, because longer videos signal higher content investment

6. Which of the following engagement actions is generally considered the highest-weight signal by most social media platforms?

a) Leaving a comment with three or more words b) Clicking the like button while watching c) Sending the content to another person via direct message d) Visiting the creator's profile after viewing

7. Marcus Webb's finance content was algorithmically depressed partly because YouTube grouped it with what category of content?

a) Content that had been flagged for copyright violations b) High-return/get-rich-quick finance content that had been penalized c) Content with low production quality d) Content that had previously had community guidelines strikes

8. According to the chapter, what is the main difference between how podcasting and TikTok handle content discovery?

a) Podcasts have more sophisticated algorithms than TikTok b) TikTok focuses on follower count, while podcasts use engagement rate c) Most podcast apps have no meaningful discovery algorithm; TikTok's FYP actively pushes content to non-followers d) Podcasts use geographic targeting while TikTok uses interest-based targeting

9. The term "engagement bait" refers to:

a) Content that performs unexpectedly well due to an algorithm change b) A tactic used by platforms to encourage creators to use paid promotion c) Content specifically designed to generate engagement signals without necessarily delivering genuine value d) A type of content that encourages viewers to engage with other creators

10. Which of the following best represents the equity concern raised in this chapter about recommendation algorithms?

a) Algorithms favor large creators over small creators, which is unfair but unavoidable b) Algorithms embed the values of their designers and produce documented disparate impacts, including suppression of content from creators of color and amplification of misinformation c) Algorithms are perfectly neutral and equity concerns are about platform policies, not the algorithms themselves d) Equity concerns only apply to political content, not to creator economy issues

Answer Key

1. B — Recommendation algorithms are optimization systems that predict which content will generate the engagement behaviors (watch time, shares, saves, etc.) that keep users on the platform. They don't evaluate "quality" in any subjective sense — they predict engagement outcomes. Option A describes an ideal but not the reality. Options C and D are inaccurate descriptions of how these systems work.

2. C — TikTok's documented primary signal for initial FYP distribution is watch completion rate from the seed audience. TikTok is famous for prioritizing completion rate above almost all other signals in the early distribution phase. Follower count, likes, and hashtags all matter but are secondary to completion rate.

3. C — YouTube explicitly uses both CTR (which determines whether people click when shown your thumbnail) and watch time (which determines whether they stay after clicking). A video needs both to achieve sustained distribution. High CTR with low watch time signals misleading packaging; high watch time with low CTR means the content is good but not being discovered. YouTube's algorithm accounts for both.

4. C — All three reasons are legitimate and real: competitive advantage (the algorithm is a core product differentiator), abuse prevention (publishing the algorithm would enable gaming at scale), and legal liability (documented suppression of certain content categories could create legal exposure). Option D is not a real reason platforms cite.

5. B — On TikTok, completion rate is the primary signal for initial and expanded distribution. A 15-second video with 88% completion outperforms a 90-second video with 42% completion because completion rate, not total watch time, drives the FYP evaluation cycle. This is one of the most practically important things to understand about TikTok's algorithm.

6. C — Sending content to another person via DM is consistently treated as the highest-weight signal across most platforms because it represents an active, intentional act of personal endorsement. No one sends their friends content they disliked. Comments, likes, and profile visits are all lower-weight signals by comparison.

7. B — The chapter describes how Marcus's content was algorithmically grouped with get-rich-quick finance content that had been heavily penalized after a wave of fraudulent financial advice went viral. His content was not this type, but algorithmic classification errors can depress legitimate creators by association. This illustrates the equity problem of algorithmic systems affecting creators without transparency or recourse.

8. C — Most podcast apps operate on an RSS feed model with no meaningful discovery algorithm; listeners find podcasts through word of mouth, search, and cross-promotion. TikTok's FYP actively pushes content to non-followers through its recommendation algorithm. This fundamental difference requires entirely different growth strategies.

9. C — Engagement bait is content explicitly designed to generate engagement signals (like, comment, share) without necessarily delivering genuine value. "Drop a like if you agree!" is a classic example. The algorithm is fooled into thinking the content is high-quality because engagement metrics look good, even though viewers didn't find it genuinely interesting or useful.

10. B — The chapter cites documented research showing that algorithms embed the values and blind spots of their creators, producing disparate impacts including suppression of content from creators of color (documented in investigations of TikTok and Instagram) and amplification of misinformation (MIT Media Lab research). Option C (algorithms are neutral) is explicitly argued against in the chapter. Option A describes a real pattern but is not the primary equity concern raised.