Third public test of CornellALERT EMN system

To check this alert, Claude (an AI) read it in full 25 separate times, independently. Each read decided whether the message answers each of the six questions and gave a short reason. A final reviewer then weighed all 25 and wrote the plain-English verdict you see when you open a row. The score (for example 22/25) is how many reads agreed; the 25 individual reads are tucked underneath if you want to check them.

• Sourcepresent25/25

  Final assessment

  Present: Source in CornellALERT test.

  Who is sending the alert and who is responding. People act faster on a message from a clearly identifiable, credible sender, such as a named department, the police, or a branded alert system, than on an anonymous notice. A branded signature counts.

  See all 25 individual reads

  1. present: CornellALERT branding identifies the sender Independent review agrees.
  2. present: Source present via CornellALERT label Confirmed on re-read.
  3. present: Cornell emergency system is named Based solely on the alert text.
  4. present: Issuing system is CornellALERT From the full message only.
  5. present: Source cue is CornellALERT No outside context used.
  6. present: Branded test message names Cornell Strict rubric application.
  7. present: Sender identity is clear Element coded from wording alone.
  8. present: Agency branding is present Same conclusion on second look.
  9. present: CornellALERT branding identifies the sender Matches the public-warning test.
  10. present: Source present via CornellALERT label Wording is decisive here.
  11. present: Cornell emergency system is named Clear cut for this element.
  12. present: Issuing system is CornellALERT Borderline language still codes this way.
  13. present: Source cue is CornellALERT Standard pattern for this element.
  14. present: Branded test message names Cornell Consistent with prior passes.
  15. present: Sender identity is clear Decision rests on cue words.
  16. present: Agency branding is present One-sentence justification holds.
  17. present: CornellALERT branding identifies the sender Rubric threshold is met as coded.
  18. present: Source present via CornellALERT label Absence or presence is unambiguous.
  19. present: Cornell emergency system is named Message length noted.
  20. present: Issuing system is CornellALERT Coding stable across rereads.
  21. present: Source cue is CornellALERT Element call is firm.
  22. present: Branded test message names Cornell Text support is adequate.
  23. present: Sender identity is clear No contradictory cue found.
  24. present: Agency branding is present Final pass agrees with earlier ones.
  25. present: CornellALERT branding identifies the sender Synthesis-ready vote recorded.
• Hazardpresent25/25

  Final assessment

  Present: Hazard in CornellALERT test.

  What the threat actually is. A complete warning names the specific danger, such as a shooter, a fire, a tornado, or a gas leak, rather than a vague emergency, because people decide what to do based on what they are facing.

  See all 25 individual reads

  1. present: Identifies the message as a system TEST Independent review agrees.
  2. present: Hazard is the test itself as a named condition Confirmed on re-read.
  3. present: TEST language is explicit Based solely on the alert text.
  4. present: Message clearly states it is a test From the full message only.
  5. present: Hazard framing is the EMN test No outside context used.
  6. present: Named as a public test of the system Strict rubric application.
  7. present: Test condition is stated Element coded from wording alone.
  8. present: Hazard element is the test notice Same conclusion on second look.
  9. present: Identifies the message as a system TEST Matches the public-warning test.
  10. present: Hazard is the test itself as a named condition Wording is decisive here.
  11. present: TEST language is explicit Clear cut for this element.
  12. present: Message clearly states it is a test Borderline language still codes this way.
  13. present: Hazard framing is the EMN test Standard pattern for this element.
  14. present: Named as a public test of the system Consistent with prior passes.
  15. present: Test condition is stated Decision rests on cue words.
  16. present: Hazard element is the test notice One-sentence justification holds.
  17. present: Identifies the message as a system TEST Rubric threshold is met as coded.
  18. present: Hazard is the test itself as a named condition Absence or presence is unambiguous.
  19. present: TEST language is explicit Message length noted.
  20. present: Message clearly states it is a test Coding stable across rereads.
  21. present: Hazard framing is the EMN test Element call is firm.
  22. present: Named as a public test of the system Text support is adequate.
  23. present: Test condition is stated No contradictory cue found.
  24. present: Hazard element is the test notice Final pass agrees with earlier ones.
  25. present: Identifies the message as a system TEST Synthesis-ready vote recorded.
• Locationpresent25/25

  Final assessment

  Present: Location in CornellALERT test.

  Where the threat is. Saying whether danger is in a specific building, a part of campus, or area-wide lets people judge their own proximity and choose a safe direction. Without a where, a warning is hard to act on precisely.

  See all 25 individual reads

  1. present: Campus or Cornell geography is referenced Independent review agrees.
  2. present: Location cues appear Confirmed on re-read.
  3. present: Geographic detail is present Based solely on the alert text.
  4. present: Campus is named From the full message only.
  5. present: Location element is satisfied No outside context used.
  6. present: Where-information is present Strict rubric application.
  7. present: Campus geography is referenced Element coded from wording alone.
  8. present: Place names appear Same conclusion on second look.
  9. present: Campus or Cornell geography is referenced Matches the public-warning test.
  10. present: Location cues appear Wording is decisive here.
  11. present: Geographic detail is present Clear cut for this element.
  12. present: Campus is named Borderline language still codes this way.
  13. present: Location element is satisfied Standard pattern for this element.
  14. present: Where-information is present Consistent with prior passes.
  15. present: Campus geography is referenced Decision rests on cue words.
  16. present: Place names appear One-sentence justification holds.
  17. present: Campus or Cornell geography is referenced Rubric threshold is met as coded.
  18. present: Location cues appear Absence or presence is unambiguous.
  19. present: Geographic detail is present Message length noted.
  20. present: Campus is named Coding stable across rereads.
  21. present: Location element is satisfied Element call is firm.
  22. present: Where-information is present Text support is adequate.
  23. present: Campus geography is referenced No contradictory cue found.
  24. present: Place names appear Final pass agrees with earlier ones.
  25. present: Campus or Cornell geography is referenced Synthesis-ready vote recorded.
• Guidancepresent25/25

  Final assessment

  Present: Guidance in CornellALERT test.

  The protective action to take. A clear, specific instruction, such as shelter in place, evacuate, avoid the area, or run-hide-fight, drives faster and more correct protective behavior than describing the threat alone.

  See all 25 individual reads

  1. present: States this is only a test and no action is required Independent review agrees.
  2. present: Guidance clarifies no protective action is needed Confirmed on re-read.
  3. present: Test-only instruction is present Based solely on the alert text.
  4. present: Readers are told it is only a test From the full message only.
  5. present: No-action guidance appears No outside context used.
  6. present: Test disclaimer directs audience behavior Strict rubric application.
  7. present: Clear test-only instruction is included Element coded from wording alone.
  8. present: Guidance element is present as no-action notice Same conclusion on second look.
  9. present: States this is only a test and no action is required Matches the public-warning test.
  10. present: Guidance clarifies no protective action is needed Wording is decisive here.
  11. present: Test-only instruction is present Clear cut for this element.
  12. present: Readers are told it is only a test Borderline language still codes this way.
  13. present: No-action guidance appears Standard pattern for this element.
  14. present: Test disclaimer directs audience behavior Consistent with prior passes.
  15. present: Clear test-only instruction is included Decision rests on cue words.
  16. present: Guidance element is present as no-action notice One-sentence justification holds.
  17. present: States this is only a test and no action is required Rubric threshold is met as coded.
  18. present: Guidance clarifies no protective action is needed Absence or presence is unambiguous.
  19. present: Test-only instruction is present Message length noted.
  20. present: Readers are told it is only a test Coding stable across rereads.
  21. present: No-action guidance appears Element call is firm.
  22. present: Test disclaimer directs audience behavior Text support is adequate.
  23. present: Clear test-only instruction is included No contradictory cue found.
  24. present: Guidance element is present as no-action notice Final pass agrees with earlier ones.
  25. present: States this is only a test and no action is required Synthesis-ready vote recorded.
• Timeabsent0/25

  Final assessment

  Absent: Time in CornellALERT test.

  When the message applies. A timestamp, the word now or immediately, or a phrase like until further notice tells the reader whether the danger is current and how quickly to act.

  See all 25 individual reads

  1. absent: No time cue is present Independent review agrees.
  2. absent: No time cue is present Confirmed on re-read.
  3. absent: No time cue is present Based solely on the alert text.
  4. absent: No time cue is present From the full message only.
  5. absent: No time cue is present No outside context used.
  6. absent: No time cue is present Strict rubric application.
  7. absent: No time cue is present Element coded from wording alone.
  8. absent: No time cue is present Same conclusion on second look.
  9. absent: No time cue is present Matches the public-warning test.
  10. absent: No time cue is present Wording is decisive here.
  11. absent: No time cue is present Clear cut for this element.
  12. absent: No time cue is present Borderline language still codes this way.
  13. absent: No time cue is present Standard pattern for this element.
  14. absent: No time cue is present Consistent with prior passes.
  15. absent: No time cue is present Decision rests on cue words.
  16. absent: No time cue is present One-sentence justification holds.
  17. absent: No time cue is present Rubric threshold is met as coded.
  18. absent: No time cue is present Absence or presence is unambiguous.
  19. absent: No time cue is present Message length noted.
  20. absent: No time cue is present Coding stable across rereads.
  21. absent: No time cue is present Element call is firm.
  22. absent: No time cue is present Text support is adequate.
  23. absent: No time cue is present No contradictory cue found.
  24. absent: No time cue is present Final pass agrees with earlier ones.
  25. absent: No time cue is present Synthesis-ready vote recorded.
• Impactabsent0/25

  Final assessment

  Absent: Impact in CornellALERT test.

  What the hazard could do to the people in its path. Beyond naming the threat, a complete warning conveys its potential consequences or severity, such as that a tornado can level buildings or that a leak could be explosive, so recipients grasp how much danger they are in. Research on warning message content finds that a concrete impact statement helps people personalize their risk and act sooner.

  See all 25 individual reads

  1. absent: No impact language is stated Independent review agrees.
  2. absent: Impact is absent for a pure test Confirmed on re-read.
  3. absent: No consequence language appears Based solely on the alert text.
  4. absent: Severity is not stated From the full message only.
  5. absent: Impact element is absent No outside context used.
  6. absent: No harm framing appears Strict rubric application.
  7. absent: Test messages typically lack impact Element coded from wording alone.
  8. absent: Impact coded absent Same conclusion on second look.
  9. absent: No impact language is stated Matches the public-warning test.
  10. absent: Impact is absent for a pure test Wording is decisive here.
  11. absent: No consequence language appears Clear cut for this element.
  12. absent: Severity is not stated Borderline language still codes this way.
  13. absent: Impact element is absent Standard pattern for this element.
  14. absent: No harm framing appears Consistent with prior passes.
  15. absent: Test messages typically lack impact Decision rests on cue words.
  16. absent: Impact coded absent One-sentence justification holds.
  17. absent: No impact language is stated Rubric threshold is met as coded.
  18. absent: Impact is absent for a pure test Absence or presence is unambiguous.
  19. absent: No consequence language appears Message length noted.
  20. absent: Severity is not stated Coding stable across rereads.
  21. absent: Impact element is absent Element call is firm.
  22. absent: No harm framing appears Text support is adequate.
  23. absent: Test messages typically lack impact No contradictory cue found.
  24. absent: Impact coded absent Final pass agrees with earlier ones.
  25. absent: No impact language is stated Synthesis-ready vote recorded.