I’m delighted to be presenting on the panel of Hazards Research by New Professionals.
You’ll note that this is an extremely simple presentation, with extra cat pictures. This is because there’s evidence (1, 2) that duplicating text and audio information is unhelpful at best. But presentations are still the industry standard, accessible in different ways, and provide a quick point of reference for people arriving late or leaving early. Since they are very simple slides, and my roommate’s cats are very cute, they are there for visual interest.
I’ve made multiple presentations about my Station fire research – two of them were taped and are viewable on my Research page. That was also the subject of my Master’s thesis, and multiple other papers are currently under consideration, so I won’t go into extensive detail here.
My newer research, and the focus of this talk, is on the Beverly Hills Supper Club Fire in Kansas in 1972. It adds new complexity to the way I’m studying fires, based on several factors, some of which inform wildly different parts of the approach.
- Multi-story building (programming)
- Now working from an extensive paper dataset rather than a csv
- Aggregating the data
- Making judgement calls about group identification
- Matching evacuation maps with surveys and police statements
- Fire burned a much longer time before it was noticed, and there were more internal walls, which in turn impacts several factors
- Warnings and alarm
- Evacuation speed
I do not yet have a computer model of this evacuation, but it will come eventually. So far, the greatest insights have come from reading the statements and surveys, where people discussed why and how and when they did things in some detail. Particularly, people’s awareness of the fire as a hazard to be taken seriously has been thrown into sharp relief. There is a persistent thread of gender as it impacts hazard awareness throughout the literature, and this is borne out in some cases, such as all of the female musicians in the Cabaret Room packing up and leaving as soon as they were notified, while none of the male musicians did, but there are additional factors showing up, as well.
The way the warning worked in that particular room is there was a comedy act ongoing while people ate dinner, and a busboy came in and got up on stage to announce that the building was on fire. Initially, some people considered this as part of the comedy act, which definitely influenced how seriously people took the warning.
But, before he got on stage, the same busboy – a young man, generally considered part of a demographic that is less risk-averse – alerted the head waitress of that room – a somewhat older adult woman, generally considered part of a demographic that is more risk-averse – that there was a fire. She dismissed him, and dismissed concerns about the fire. This raises questions that are harder to isolate on a broader scale over how social status specifically – as opposed to gender generally – influence risk perception.
There are also interesting pools of information in how employees acted. Specifically, many identified the people at their tables as their people and proceeded to shepherd them out, while some employees who were in the kitchen or in less customer-facing positions were more oriented towards the fire or other employees.
These are preliminary findings, of course, because in March we all went home and pivoted to COVID-19 research, and since these boxes are one of a kind irreplaceable data they’ve been in lockdown ever since. I miss them.
I am also looking into wildfire evacuation in California this summer, as soon as it’s approved by IRB.