Scores for adults are dropping on tests of basic skills

Study design and cross-country comparisons

• Commenters link the OECD report and note that many countries show declines, with some (e.g., Poland, Lithuania, Korea) dropping more than the US.
• Several argue cross-country comparisons are weak because questions are culturally loaded; within-country trend changes may be more meaningful.
• One notes US has high non-response bias and cautions that sampling 16–65, with scores falling with age, may leave residual bias despite weighting.

Score declines and which adults are affected

• The largest drops are in older adults (55–65).
• Some argue this doesn’t match explanations focused on current schooling methods or social media use, since those cohorts left school long ago.
• Others suggest compositional effects: increased immigration including older adults with limited formal education may depress averages. This is disputed and flagged as speculative.

Debate over the sample literacy question

• Much discussion centers on the “crackers become soft at 9% moisture” item.
• Some say this is trivial and the issue is poor schooling and reading instruction methods.
• Others focus on wording nuances: “seem” vs “become,” “about” 9%, and the mismatch with real-world reasoning, arguing tests reward pattern-matching over genuine comprehension.
• A subthread questions whether syntax like “At what moisture level…” is actually complex or just unfamiliar; some note people may parse inverted word order poorly.

Test implementation and user experience

• Multiple commenters try the OECD demo tests and describe the interface as janky, confusing, and outdated (including a Firefox-only notice).
• A few find the items straightforward once the UI is understood; others say the UX itself could depress performance.

Possible causes: COVID, media, technology, and AI

• One line of discussion proposes COVID-related neurological effects as a contributor; others see this as plausible but unproven and point instead to broader pandemic-era social and psychological disruption.
• Social media and mobile devices are blamed by some for eroding sustained attention and deep reading; others insist this needs rigorous study rather than repetition.
• Several worry that ubiquitous LLMs could further atrophy reasoning if people offload thinking too quickly, while others note that if reasoning becomes rarer it might also become more economically valuable.

Alternative interpretations and skepticism

• One commenter argues the data do not show people “getting stupider” but reflect cohort differences (e.g., schooling during turbulent vs more stable eras) and a shift toward audio/video learning that tests don’t capture.
• Another initially dismisses the test as irrelevant to modern information work, then retracts somewhat after trying it, criticizing implementation more than the conceptual goals.