Calibration - Safety Last! (1923)
Screen this public-domain film excerpt with the standardized 12-agent calibration panel. Focus on immediate viewer response to the visual and audio material in the attached clip.
Research group: 12-agent US calibration panel (ages 29–62; mostly rural; mixed occupations; included a Spanish-preferred respondent).
What they said: Most described the clip as tense/unsafe/chaotic and not appealing; a smaller creative/nostalgia-leaning subset enjoyed the stunt craft and clear visual storytelling. Main insights: Perceived danger suppressed appeal and intent to continue for the majority; only a few would watch more, driven by curiosity about the stunt’s resolution.
Credibility: Generally low as an informative “host/source,” though some granted credibility as authentic archival material.
Drivers: Accessibility and framing gaps (English-only intertitles, no context/safety cues) heightened discomfort and reduced engagement, especially for the Spanish-preferred viewer.
- Reframe and de-risk: Add a 5–10s context card and safety disclaimer, label as archival/staged high-tension content, consider a softer score and tighter cut, and adjust title/thumbnail to set expectations.
- Broaden reach selectively: Ship Spanish captions/subtitles (including intertitles), target nostalgia/film-literate channels while deprioritizing safety-averse cohorts, and track Appeal, Continue Intent, and Framing Clarity to validate lift.
Overall appeal
n=12"Watching it, I found it very unappealing. The whole sequence felt chaotic and risky, with people hanging off a building in a way that made me tense rather than entertained, and I do not enjoy content that feels like the..."
"I found it pretty appealing. It was easy to follow just from the visuals, and the physical comedy was clever without feeling overdone or showy."
"Wasn't much of a draw for me. Old silent stuff and all that risky climbing around just felt busy and kind of pointless. I watched it, sure, but I wouldn't say I liked it."
"Here's the thing... it was not very appealing for me. The whole clip felt tense and unsafe, and I mostly just watched people in danger. I prefer content that feels more straightforward and less stressful."
Watch intent
n=12"I probably would not keep watching. What I saw felt dragged out and kind of far-fetched, and that sort of tense, impractical setup just doesn't hold me for long."
"What I watched made me uneasy more than interested. The risky stunt material felt careless and stressful, and for me that kind of chaos is hard to enjoy, so I would probably not continue."
"I can appreciate the simple visual storytelling and the built-in suspense, and there’s a certain old-fashioned charm to it. But for me it still lands as something I respect more than something I’d really keep watching."
"I probably would not keep watching. What I saw felt tense right away, with that hanging-off-the-building suspense, and the old silent style just is not really for me. I like things that feel easier to settle into, and th..."
Host credibility
n=12"Watching it, I did not find the source credible at all. It came across as staged silent-film entertainment with exaggerated physical comedy, so for me it did not read like something meant to inform or be taken as reliabl..."
"Looked straight with me. It came off like an old film clip and didn't feel like it was trying to dress itself up as something else, so I mostly bought the source."
"It felt credible to me. What I watched had the look of a real old film clip, and nothing about it came across as fake or dressed up for effect."
"Here's the thing... it did not feel very credible to me. There was no clear host guiding it, and what I watched felt exaggerated and played for laughs more than something dependable or believable."
Sam Norstrom
62 · Rural, NE, USA · Driver
Peace Evangelista
31 · Somerville, MA, USA · Human Resources Specialist
Sandra Falcinelli
61 · Rural, PA, USA · Designer
Kaila Smith
29 · Ann Arbor, MI, USA · Business Operations Specialist
Daniel Sassaman
55 · Rural, LA, USA · Engineer
Precious Rai
40 · Rural, IL, USA · Medical Records Specialist
Brent Guevara
52 · Fort Myers, FL, USA · Personal Care Aide
Gregory Cumbo
60 · Rural, OH, USA · Brokerage Clerk
Brianna Chapman
32 · Rural, WV, USA · Hairdresser and Cosmetologist
Maribel Miller
35 · Rural, NH, USA · Retail Sales Supervisor
John Grimm
62 · Rural, IA, USA · Civil Engineer
Mario Bockus
58 · Rural, VA, USA · Retail Sales Supervisor
Sam Norstrom
62 · Rural, NE, USA · Driver
Peace Evangelista
31 · Somerville, MA, USA · Human Resources Specialist
Sandra Falcinelli
61 · Rural, PA, USA · Designer
Kaila Smith
29 · Ann Arbor, MI, USA · Business Operations Specialist
Daniel Sassaman
55 · Rural, LA, USA · Engineer
Precious Rai
40 · Rural, IL, USA · Medical Records Specialist
Brent Guevara
52 · Fort Myers, FL, USA · Personal Care Aide
Gregory Cumbo
60 · Rural, OH, USA · Brokerage Clerk
Brianna Chapman
32 · Rural, WV, USA · Hairdresser and Cosmetologist
Maribel Miller
35 · Rural, NH, USA · Retail Sales Supervisor
John Grimm
62 · Rural, IA, USA · Civil Engineer
Mario Bockus
58 · Rural, VA, USA · Retail Sales Supervisor
| Age bucket | Male count | Female count |
|---|
| Income bucket | Participants | US households |
|---|
Brianna Chapman
32 - Rural, WV
"I mean, he's still up there, and it just looks like he's gonna slip, honestly."
Mario Bockus
58 - Rural, VA
"Holy smokes, he's really still up there, looking down like he's about to take a dive."
Sandra Falcinelli
61 - Rural, PA
"It's just a wonder he hasn't fallen yet, hanging onto that clock like that."
Maribel Miller
35 - Rural, NH
"Oh, for crying out loud, he's really out there now, just hanging on."
Brent Guevara
52 - Fort Myers, FL
"He's still up there, <i>Dios mío</i>, I'm just hoping he gets down safely before something really bad happens."
Daniel Sassaman
55 - Rural, LA
"He's still trying to fix that thing. I wonder how those anchors are holding up."
John Grimm
62 - Rural, IA
"He's still moving around up there, really pushing the limits of whatever those old anchors can take."
Kaila Smith
29 - Ann Arbor, MI
"He's still in a very precarious position, and I hope he has a solid strategy to descend safely."
Precious Rai
40 - Rural, IL
"He really needs to be careful up there; it gives me the jitters just watching him."
Sam Norstrom
62 - Rural, NE
"He's still out there, just waiting for something bad to happen."
Gregory Cumbo
60 - Rural, OH
"Oh, for goodness sake, he's <b>still</b> up there! My heart can't take this, honestly. Someone just needs to get him down, or let him fall already so we can be done with it!"
Peace Evangelista
31 - Somerville, MA
"The continued exposure of this individual at such a height, without proper safety measures, remains a critical concern for potential workplace incidents."
Overview
Key Segments
| Segment | Attributes | Insight | Supporting Agents |
|---|---|---|---|
| Older rural viewers (55–62) |
|
Tended to describe the clip as chaotic, reckless and unappealing; most would not continue watching and framed the footage as spectacle rather than credible instruction. Reaction reflects low tolerance for perceived danger in source material and preference for clearer context or safer pacing. | Sam Norstrom, John Grimm, Gregory Cumbo, Daniel Sassaman |
| Visually-minded / creative respondents |
|
More likely to find the clip engaging/appealing because of stunt work, physical comedy and straightforward visual storytelling-even when acknowledging dated style; these viewers show higher tolerance for risk-on-screen when it serves craft or spectacle. | Sandra Falcinelli, Brianna Chapman, Mario Bockus, Daniel Sassaman |
| Younger, urban/college-educated women (29–31) |
|
Reacted negatively driven by aversion to perceived risk and stress; they prioritize content that feels safe, orderly and comfortable to watch and are unlikely to continue with material that creates anxiety or confusion. | Peace Evangelista, Kaila Smith |
| Lower-income, Spanish-preferred respondent |
|
Found the clip tense and less accessible; lack of language/contextual labeling reduced ability to relax into the material and lowered intent to continue-signaling a language-accessibility barrier that depresses engagement regardless of content type. | Brent Guevara |
| Majority perspective on credibility |
|
Most panelists judged the clip as staged entertainment rather than a reliable or instructive source; silent-era exaggeration and stunt framing led to low perceived real-world credibility and diminished trustworthiness for informative use. | Peace Evangelista, Kaila Smith, Precious Rai, Daniel Sassaman, Brianna Chapman, Brent Guevara, Gregory Cumbo, John Grimm, Mario Bockus |
| Archival/nostalgic readers |
|
A smaller subset interpreted the clip as authentic archival footage and therefore granted it credibility as a historical artifact; this reading raises willingness to engage despite dated production values. | Sam Norstrom, Sandra Falcinelli, Maribel Miller |
Shared Mindsets
| Trait | Signal | Agents |
|---|---|---|
| Perceived danger / tension | Across ages and incomes, viewers used language like 'tense', 'unsafe', 'chaotic' to describe the clip; this sense of on-screen risk consistently reduced enjoyment and lowered intent to continue. | Peace Evangelista, Kaila Smith, Precious Rai, Gregory Cumbo, John Grimm, Brent Guevara, Sam Norstrom |
| Appreciation for physical comedy / stunt craft | Respondents with visual or creative orientation highlighted stunt work, timing and physical performance as positive hooks that could sustain engagement even when other aspects felt dated. | Brianna Chapman, Mario Bockus, Sandra Falcinelli, Daniel Sassaman |
| Low perceived credibility due to staged conventions | Most interpreted exaggeration and silent-era staging as signs that the clip functions as entertainment rather than an instructive or reliable source, lowering its utility for informational contexts. | Peace Evangelista, Kaila Smith, Precious Rai, Brianna Chapman, Brent Guevara, Gregory Cumbo, John Grimm, Daniel Sassaman |
| Nostalgic or visual appeal can override discomfort | When viewers value vintage craft or are curious about outcome, they tolerate dated style and perceived risk; visual storytelling and curiosity about the stunt's resolution motivate continued watching for this subset. | Brianna Chapman, Mario Bockus, Sandra Falcinelli, Maribel Miller |
Divergences
| Segment | Contrast | Agents |
|---|---|---|
| Mario Bockus (older rural male, 58) | Positive engagement with stunt spectacle contrasts with many same-age rural respondents who found the clip tense and unappealing. | Mario Bockus |
| Sandra Falcinelli (older rural designer, 61) | Reads the clip as both authentic and visually appealing, differing from other older rural respondents who dismissed it as chaotic or merely entertainment. | Sandra Falcinelli |
| Brianna Chapman (younger rural female, 32) | Expressed enthusiasm and willingness to continue despite other younger women prioritizing safety/comfort and opting out early. | Brianna Chapman |
| Daniel Sassaman (high-income engineer, 55) | Offers mixed signals-dislikes the clip overall but admires the stunt craft and might watch more-showing simultaneous technical appreciation and low tolerance for dated style. | Daniel Sassaman |
| Archival-readers vs majority low-credibility view | A small group (often older or film-literate) interprets the clip as authentic archival footage and grants it credibility as a historical artifact, which contrasts with the broader tendency to dismiss the clip as staged entertainment. | Sam Norstrom, Sandra Falcinelli, Maribel Miller, Peace Evangelista, Kaila Smith |
Overview
Quick Wins (next 2–4 weeks)
| # | Action | Why | Owner | Effort | Impact |
|---|---|---|---|---|---|
| 1 | Add 5–10s context card + safety disclaimer | Reduces perceived recklessness; reframes as archival, staged entertainment before tension builds. | Editorial/Brand | Low | High |
| 2 | Reposition metadata and thumbnail | Title/tag as “Archival Classic Stunt Comedy” with an ‘Staged Stunts-Do Not Imitate’ badge to set expectations. | Programming/Product | Low | Med |
| 3 | Provide Spanish captions + modern subtitles for intertitles | Closes accessibility gap; increases comfort and comprehension for Spanish-preferred viewers. | Localization | Med | High |
| 4 | Offer alternative softer score audio track | Lower perceived tension without changing visuals; appeals to safety-sensitive viewers. | Post-Production/Audio | Med | Med |
| 5 | Shorter cut to reduce prolonged peril beats | Trims drawn-out danger that many found fatiguing; preserves key stunt moments. | Editorial | Med | Med |
| 6 | Introduce content labels: Archival, High Tension, Staged Stunts | Clarifies credibility intent and allows users to self-select comfort level. | Product/UX | Low | Med |
Initiatives (30–90 days)
| # | Initiative | Description | Owner | Timeline | Dependencies |
|---|---|---|---|---|---|
| 1 | Packaging A/B/C Test: Context, Score, and Pacing | Run controlled experiments across three variants: (A) context card + safety badge; (B) softer score; (C) trimmed pacing. Measure appeal, completion, continue-intent, and tension mentions. | Research + Programming | 4–6 weeks | Post-production resources, Analytics dashboards, Legal clearance on rescoring |
| 2 | Accessibility & Localization Rollout | Produce Spanish captions and recreate intertitles as on-screen subtitles; ensure WCAG-compliant caption styling and QA. | Localization | 6–8 weeks | Transcription assets, Budget approval, QA vendors |
| 3 | Credibility Scaffolding: 60s Historian/Host Intro | Lightweight host intro framing the scene as archival stunt comedy with brief ‘how they did it’ notes to boost trust and reduce anxiety. | Editorial/Brand | 8–10 weeks | Talent/voiceover booking, Production scheduling, Brand/legal review |
| 4 | Segmented Distribution Strategy | Prioritize nostalgia/film-literate channels and off-platform communities; deprioritize in feeds of safety-averse cohorts; adjust recs based on tension-sensitivity. | Growth/Programming | 4 weeks | Audience segmentation data, Recommendation controls, Creative variants |
| 5 | Risk & Content Labeling Framework | Build taxonomy and UI badges for Archival/Staged/High Tension across the archive; integrate into CMS and front-end filters. | Product/UX | 1 quarter | Taxonomy design, CMS updates, Frontend engineering |
| 6 | Calibration Panel Expansion + Tension Sensitivity Index | Increase N, add language diversity, and create a standardized index for perceived on-screen risk to guide programming decisions. | Research/Insights | 6 weeks | Panel recruitment, Survey instrument updates, Budget |
KPIs to Track
| # | KPI | Definition | Target | Frequency |
|---|---|---|---|---|
| 1 | Appeal Score (Top-2 box) | Percent of viewers rating appeal as positive on Q1 after exposure to packaged variants. | +15% vs baseline | Per test cycle |
| 2 | Continue Intent | Percent likely to watch more (Q2) post-packaging. | +20% vs baseline | Per test cycle |
| 3 | Framing Clarity | Share correctly identifying content as archival staged entertainment (not instructional). | ≥80% | Per test cycle |
| 4 | Completion Rate (75% VTR) | Percent of viewers reaching 75% of runtime. | +10 percentage points | Weekly |
| 5 | Accessibility Adoption | Share of Spanish-preferred viewers enabling captions/subtitles. | ≥40% | Monthly |
| 6 | Negative Affective Mentions | Coded mentions of ‘tense/unsafe/chaotic’ in qual/feedback per 100 views. | -30% vs baseline | Per test cycle |
Risks & Mitigations
| # | Risk | Mitigation | Owner |
|---|---|---|---|
| 1 | Modern packaging may alienate purists who value unaltered authenticity. | Offer dual versions (original and packaged) and clearly label differences. | Editorial |
| 2 | Perceived endorsement of unsafe behavior could trigger brand backlash. | Use pre-roll disclaimers, on-screen badges, and age-appropriate gating where needed. | Brand/Legal |
| 3 | Findings may overfit a small 12-person calibration panel. | Expand panel size/diversity; validate with quantitative A/B tests. | Research/Insights |
| 4 | Rights or music constraints may limit rescoring or edits. | Confirm public-domain status; use PD/commissioned score; document edit rights. | Legal/Post-Production |
| 5 | Localization quality or timing issues reduce trust and usability. | Professional translation + QC, style guides, and spot-checks with target users. | Localization |
| 6 | Operational complexity from new labels and variants slows delivery. | Template assets, phased rollout, prioritize high-impact titles first. | PMO/Product |
Timeline
- Week 0–2: Quick wins live (context card, metadata, labels); start subtitle production.
- Week 3–6: Run A/B/C packaging tests; monitor KPIs; iterate score and trims.
- Week 6–8: Ship Spanish captions across priority clips; initial segmented distribution.
- Week 8–10: Launch host intro pilot on top titles; refine based on feedback.
- Quarter 2: Roll out full risk/labeling framework in CMS/front-end; expand calibration panel and publish Tension Sensitivity Index.
A+E Global Calibration: Safety Last! (1923) - Immediate Viewer Response Synthesis
Objective and context. We screened a public-domain excerpt from Safety Last! (1923) to A+E’s standardized 12-agent calibration panel to capture immediate reactions to the clip’s visual and audio presentation. Across 36 total responses to questions on appeal, intent to continue, and credibility, we observed a polarized pattern: most viewers experienced the material as tense and unsafe, while a smaller, visually literate subset valued its stunt craft and vintage storytelling.
Cross-question learnings grounded in panel evidence.
- Perceived danger drives disengagement. Language like “tense,” “unsafe,” and “chaotic” appeared consistently and correlated with lower appeal and intent to continue (e.g., Peace Evangelista, Kaila Smith, Precious Rai, Gregory Cumbo, John Grimm, Brent Guevara, Sam Norstrom).
- Physical comedy/stunt craft sustains interest for creatives. Visually minded respondents pointed to timing, performance, and spectacle as positives (Brianna Chapman, Mario Bockus, Sandra Falcinelli, Daniel Sassaman).
- Low credibility as information. Most framed the clip as staged silent-era entertainment, limiting its perceived utility as an instructive source (Evangelista, Smith, Rai, Chapman, Guevara, Cumbo, Grimm, Sassaman).
- Nostalgic/archival framing can override discomfort. A smaller subset-often older or film-literate-granted it value as authentic archival material (Sam Norstrom, Sandra Falcinelli, Maribel Miller).
Persona correlations and demographic nuances.
- Older rural viewers (55–62) tended to reject the clip as reckless and unappealing, preferring clearer context and safer pacing (Sam Norstrom, John Grimm, Gregory Cumbo, Daniel Sassaman).
- Visually minded/creative respondents (32–61) were more engaged, citing stunt work and visual storytelling despite dated style (Sandra Falcinelli, Brianna Chapman, Mario Bockus, Sassaman).
- Younger urban, college-educated women (29–31) reacted negatively due to stress/aversion to risk; low likelihood to continue (Peace Evangelista, Kaila Smith).
- Spanish-preferred, lower-income viewer flagged accessibility gaps; lack of language/context limited comfort and intent (Brent Guevara).
- Notable divergences: Mario Bockus and Sandra Falcinelli buck older rural trends with positive engagement; Brianna Chapman diverges from younger female peers with enthusiasm; Daniel Sassaman shows mixed tolerance-technical admiration yet low appeal.
Implications for packaging and distribution. Without upfront framing, this content is likely to underperform among safety-averse cohorts, especially older rural and younger urban women. Conversely, archival/nostalgia and creative-leaning segments can be profitably targeted. Accessibility shortfalls depress engagement among Spanish-preferred viewers regardless of content type.
Recommendations (actionable, evidence-led).
- Add a 5–10s context card + safety disclaimer. Reframe as archival, staged stunt comedy to reduce perceived recklessness (addresses dominant “unsafe/chaotic” read).
- Reposition metadata/thumbnail. Title/tag as “Archival Classic Stunt Comedy” with a “Staged Stunts-Do Not Imitate” badge to set expectations.
- Provide Spanish captions + modern subtitles for intertitles. Closes the accessibility gap highlighted by the Spanish-preferred viewer.
- Offer an alternate, softer score. Lowers tension for safety-sensitive viewers without altering visuals.
- Create a shorter cut. Trim prolonged peril beats that many found fatiguing while preserving key stunt moments.
Strategic initiatives, risks, and guardrails.
- Packaging A/B/C test: (A) context + badge, (B) softer score, (C) trimmed pacing; measure appeal, completion, continue-intent, and tension mentions.
- Accessibility & localization rollout: Spanish captions and recreated intertitles with WCAG-compliant styling.
- Credibility scaffolding: a 60s host/historian intro to explain “how they did it” and reduce anxiety.
- Segmented distribution: prioritize nostalgia/film-literate channels; deprioritize in feeds for tension-sensitive cohorts.
- Risk/content labeling framework: standardize Archival/Staged/High Tension badges across the archive.
- Risks/mitigations: Purist backlash (offer dual versions); brand safety (pre-roll disclaimers, age gating); small-panel overfit (validate via A/B tests); rights/music constraints (confirm PD, commission score); localization quality (pro translation + QA).
Next steps and measurement.
- Weeks 0–2: Ship context card, metadata/labels; begin Spanish caption production.
- Weeks 3–6: Run A/B/C packaging tests; iterate score and trims based on data.
- Weeks 6–8: Publish captions; launch segmented distribution.
- Weeks 8–10: Pilot 60s host intro on top titles; refine.
- Quarter 2: Roll out risk/labeling framework; expand calibration panel and publish a Tension Sensitivity Index.
- Primary KPIs: Appeal (Top-2) +15% vs baseline; Continue Intent +20%; Framing Clarity ≥80%; 75% VTR +10pp; Accessibility Adoption among Spanish-preferred ≥40%.
Success looks like a measurable reduction in “tense/unsafe” mentions, higher clarity that the piece is staged archival entertainment, and improved completion/intent-especially among safety-sensitive cohorts-without eroding engagement from nostalgia/creative segments.
-
Please rate your immediate reaction to the clip on these scales: calm–tense; safe–dangerous; easy to follow–hard to follow; staged–realistic; music: subtle–overpowering.semantic differential Quantifies tone, safety, clarity, and music intensity to guide editing choices and framing.
-
Which aspects most affected your experience of this clip (positively or negatively)? Items: music intensity; perceived danger in stunts; clarity of visual storytelling; English-only intertitles; pacing/edit length; black-and-white/silent format; archival film look; shot composition/stunt craft; lack of context or safety cues.maxdiff Identifies top drivers to prioritize edits and metadata messaging.
-
If this clip started with a 5–10 second context card and safety disclaimer, how would your likelihood to continue change?single select Estimates the incremental impact of adding context/disclaimer on continuation.
-
Please rank the following potential additions by how much they would improve your experience (most to least): translated subtitles; English captions for intertitles/audio; brief historical context card; safety disclaimer; softer music; slower intertitle duration; on‑screen narrator; none.rank Prioritizes which accessibility and context features to implement first.
-
Which audio treatment would you prefer for this clip?single select Selects soundtrack direction for the next cut.
-
In one sentence, what did you understand was happening in the scene?open text Checks narrative comprehension to determine need for added context or captions.
Research group: 12-agent US calibration panel (ages 29–62; mostly rural; mixed occupations; included a Spanish-preferred respondent).
What they said: Most described the clip as tense/unsafe/chaotic and not appealing; a smaller creative/nostalgia-leaning subset enjoyed the stunt craft and clear visual storytelling. Main insights: Perceived danger suppressed appeal and intent to continue for the majority; only a few would watch more, driven by curiosity about the stunt’s resolution.
Credibility: Generally low as an informative “host/source,” though some granted credibility as authentic archival material.
Drivers: Accessibility and framing gaps (English-only intertitles, no context/safety cues) heightened discomfort and reduced engagement, especially for the Spanish-preferred viewer.
- Reframe and de-risk: Add a 5–10s context card and safety disclaimer, label as archival/staged high-tension content, consider a softer score and tighter cut, and adjust title/thumbnail to set expectations.
- Broaden reach selectively: Ship Spanish captions/subtitles (including intertitles), target nostalgia/film-literate channels while deprioritizing safety-averse cohorts, and track Appeal, Continue Intent, and Framing Clarity to validate lift.
| Participant | Response | Actions |
|---|