Skip to content

improve(ch12): replace weak/off-domain MC distractors with plausible alternatives - #32

Merged
calmcoconut merged 1 commit into
mainfrom
improve/ch12-mc-distractors
Jun 30, 2026
Merged

calmcoconut merged 1 commit into
mainfrom
improve/ch12-mc-distractors

Conversation

@calmcoconut

Copy link
Copy Markdown
Owner

Summary

Audited all 15 multiple choice questions in Chapter 12 (Stream Processing) for distractor quality. The recurring issue was one obviously off-domain option per question — distractors like UDP protocols, CSV exports, encryption keys, and cookie stores that students can immediately eliminate without any subject knowledge. This allows guessing between the remaining 3 options, undermining the quiz's diagnostic value.

Changes

Question Topic Change
Q1 Batch vs. Stream Replaced "stores all input in volatile memory" with a plausible data-type distinction distractor
Q2 JMS/AMQP brokers Replaced "require UDP multicast" with a Kafka-style partition/offset distractor — the ideal wrong answer for this comparison
Q6 Dual writes Tightened Option C to a more credible performance/atomicity claim
Q7 CDC mechanism Replaced "export to CSV" with an app-level callback distractor (a real misconception about where CDC lives)
Q9 Log compaction Replaced GZIP/merge-partitions options with retention-policy-style distractors that sit closer to the correct answer domain
Q14 Watermarks Replaced "cryptographic signature" with a lateness-threshold distractor that sounds like a real watermark variant
Q16 Stream-stream join Replaced made-up "compacted log join" term with a real but incorrect join characterization
Q19 Microbatching Replaced "replicate nodes 10x" and "cookie store caching" with fault-tolerance-domain distractors
Q22 GDPR & immutability Replaced "SSD write latency" with a data-sovereignty distractor that stays in the regulatory compliance domain

Non-changes

Questions Q4, Q11, Q12, Q17, Q20, Q23 were left untouched — their distractors were already well-calibrated and required real subject knowledge to eliminate.

Design Principle

All revised distractors follow the rule used by Q12 (windows) and Q20 (idempotent operations) as benchmarks: every option should be plausible within the same conceptual domain as the correct answer, differing only in precision or applicability.

…alternatives

Audited all 15 MC questions for distractor quality. The recurring issue
was one obviously off-domain option per question (e.g. UDP protocols,
CSV exports, encryption keys, cookie stores) that students could
eliminate on sight without any subject knowledge.

Changes per question:
- Q1: Replace "stores all input in volatile memory" with a
  plausible-but-wrong data-type distinction distractor
- Q2: Replace "require UDP multicast" with a Kafka-style offset
  distractor — the ideal wrong answer for this comparison question
- Q6: Tighten Option C to a more credible performance/cost claim
- Q7: Replace "export to CSV" with an app-level callback distractor
  (a real misconception about where CDC lives)
- Q9: Replace gzip/merge-partitions with retention-policy-style
  distractors that sit closer to the correct answer domain
- Q14: Replace "cryptographic signature" with a lateness-threshold
  distractor that sounds like a real watermark variant
- Q16: Replace made-up "compacted log join" term with a real but
  incorrect join characterization
- Q19: Replace "replicate nodes 10x" and "cookie store caching" with
  fault-tolerance-domain distractors
- Q22: Replace "SSD write latency" with a data-sovereignty distractor
  that stays in the regulatory compliance domain
@calmcoconut
calmcoconut merged commit 2593f1d into main Jun 30, 2026
1 check passed
@calmcoconut
calmcoconut deleted the improve/ch12-mc-distractors branch June 30, 2026 22:42
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant