Category Archives: databases

Psychology Podcasts

Those who know me know I am a fiend for podcasts. Since I’m also a fiend for psychology, I can’t help but notice when it pops up in a podcast, even one not focused on psychology. I use many of them in my courses: for example, the This American Life episode on what having schizophrenia sounds like is a must listen when I hit the Abnormal Psychology chapter in Intro. The Radiolab Memory and Forgetting is a staple in my Cognitive class and I take advantage of the multi-disciplinarity of Human Factors to play clips from every area. Startup had a good one that illustrates what human factors looks like to a client.

Over the years I’ve compiled a list of my favorites relating to psychology. Some are clips from longer podcasts while some are dedicated to psychology (e.g., Invisibilia). Each one has a general area of psychology noted (although some hit two or more areas) and if it’s a clip I put the start and end time of the most related audio.

I hope you enjoy the resource and I will keep updating it as I find more. If you know of any I don’t have listed, please link to it in the comments for the blog and I’ll add it to the spreadsheet.

Are we too trusting of GPS automation?

A GPS certainly makes life easier — and although I think many of us might consider what would happen if we were without it or it was unable to identify where we were, it is less often we consider how it may lead us astray.

One of our early postings on the Human Factors Blog was about a bus driver following GPS directions that led under a too-short bridge. His case was augmented by the fact that he had chosen the “bus” setting on the GPS and assumed any route produced was therefore safe for buses. The actual model of the GPS under the bus setting was only to add routes that only buses could take, such as HOV exits, rather than to limit any route.

NPR just posted stories of people in Death Valley who got lost from following GPS directions down roads that no longer existed. In one of the cases, their car got stuck for 5 days and resulted in the death of a child. After hearing numerous stories about inaccurate GPS directions from lost drivers, a ranger investigated the maps used by the GPS systems and found roads included in them that had been closed for years. How accurate and updated do GPS systems need to be to be considered safe? How can they address over-trust in potentially dangerous situations (e.g., death valley)?

 

Mining Tragedy Update

There is new information on the West Virginia coal mine tragedy where the methane detectors were disabled to prevent automatic shut down of the machinery. This comes from NPR:

Methane monitors are mounted on the massive, 30-foot-long continuous miners because explosive gas can collect in pockets near the roofs of mines. Methane can be released as the machine cuts into rock and coal. The spinning carbide teeth that do the cutting send sparks flying when they cut into rock. The sparks and the gas are an explosive mix, so the methane monitor is designed to signal a warning and automatically shut down the machine when gas approaches dangerous concentrations.

Because the monitor continually shut down the machine:

On Feb. 13, an electrician deliberately disabled a methane gas monitor on a continuous mining machine because the monitor repeatedly shut down the machine.

Three witnesses say the electrician was ordered by a mine supervisor to “bridge” the automatic shutoff mechanism in the monitor.

There is some discussion as to whether the monitor was malfunctioning and shutting the machine down when it should not have or whether it was shutting down due to actual methane in the air. People in many industries willfully disable aids meant to keep them safe and malfunction is only one of the variables that affects the behavior (granted, it’s likely a big one). Here is one example from agriculture, collected for NIOSH through the FACE database program*:

A 26-year-old Hispanic male knitting machine operator died when he was crushed by moving parts within the knitting machine as he tried to make an adjustment.  The victim opened a safety gate and jammed a needle in the “on” button that allowed the machine to operate with the safety gates open.

Last, in at least this one case the safety cut-off contributed to an accident.

On June 4, 2004, a 47-year-old co-owner of a recycling business was run over and killed by a Gradall telescopic boom lift (rough-terrain forklift) while he was working underneath it. He had been operating the Gradall, and had shut it down when he momentarily exited the vehicle. When he returned to the machine, he found it would not restart. The Gradall had a safety interlock that prevented starting from the ignition switch while in gear. The contractor was apparently unaware of this safety feature. He checked the batteries, and then crawled underneath the cab area and reached up into the engine compartment with a screwdriver. The screwdriver made contact between the two terminals on the starter, effectively jump-starting the engine and bypassing the safety mechanism that prevented ignition while in gear. The Gradall started and moved forward. The parking brake was not set. The back left tire rolled over the contractor.

In short, I admire but do not envy the designers who have to create these dangerous systems. Their users are inventive, under pressure, and different from each other in countless ways. Designing safety sounds easy (one can imagine  “just make it shut off when they aren’t using it,”) but the answers seem far from being so simple. Many of the examples I have seen from other industries show quick and easy ways to bypass a safety system.

  • Machinery automatically cuts off after 8 seconds when there is no weight in the driver’s seat. Worker keeps a heavy tool bag nearby to put on the seat when the worker wants to check on things outside the cab.
  • Same system as above – worker tries to jump out of cab and complete task in less than 8 seconds.
  • Worker cannot reach objective with lap safety bar in place, a bar that must be down for machinery to operate. Worker lifts bar then puts it back down across empty seat and reaches for objective with machinery running.

There does seem to be a difference in premeditation in the examples I’ve come across and the idea of hiring an electrician to specifically and more permanently remove a guard from a safety system.

*I have posted on the FACE program before. It is a valuable repository.

Photo Credit NIOSH on Flickr

Database Three: Mining safety data

This post is part of a series on free data available to curious researchers and professionals:

Mine Safety and Health Administration (MSHA)

This database contains more searching capability than browsing. You need the specific geographic location of a mine before searching (state, county).

Once you zero in on a particular mine, the types of reports available are:

  1. Overview
  2. Inspections
  3. Accidents
  4. Violations
  5. VPID ( Violations Per Inspection Day)
  6. Health Samples
  7. Mine Employment and Coal Production Totals for This Mine

Here is an example of what the database returns when queried for “Health Samples”:

phs

Health samples

Here is an example of “Accidents”:

acc mine

Accidents

This database lacks much of the rich narrative present in the others and is harder to search. The only way to get complete information would be to go through each state and then each county in that state, then generate reports for each mine. The information is there, but it is hard to get into a useful state.

Database Two: Aviation safety data

This post is part of a series on free data available to curious researchers and professionals:

The Aviation Safety Network Database

The ASN Safety Database, updated every week, contains descriptions of over 12,200 airliner, military transport category aircraft and corporate jet aircraft safety occurrences since 1943.

Here is a link to their recording of the recent landing on the Hudson river by a US Airways airbus.

Included in the report:

  1. a summary of the occurrence
  2. multiple photos of damage
  3. video from reports
  4. maps
  5. statistics

Statistics include:

19th loss of a Airbus A320

This is the 19th Airbus A320 plane that was damaged beyond repair as result of an accident, a criminal act or a non-operational occurrence (hangar fire, hurricanes etc.).

11th worst accident involving a Airbus A320 (at the time)
When the accident happened it was, at the time, the 11th worst accident involving a Airbus A320. This includes crashes as a result of criminal acts (shoot down, sabotage etc.) and does also include ground fatalities.

11th worst accident involving a Airbus A320 (currently)
Currently this is the 11th worst accident involving a Airbus A320. This includes crashes as a result of criminal acts (shoot down, sabotage etc.) and does also include ground fatalities.

1095th worst accident in United States of America (at the time)

When the accident happened it was, at the time, the 1095th worst accident in United States of America. This includes crashes as a result of criminal acts (shoot down, sabotage etc.) and does also include ground fatalities. Please note that accidents that have occurred more than 12 miles / 19,3 km offshore in open waters are classified under one of the three Oceans (Atlantic, Indian, Pacific).

1096th worst accident in United States of America (currently)
Currently this is the 1096th worst accident in United States of America. This includes crashes as a result of criminal acts (shoot down, sabotage etc.) and does also include ground fatalities.

Please note that accidents that have occurred more than 12 miles / 19,3 km offshore in open waters are classified under one of the three Oceans (Atlantic, Indian, Pacific).

When available, you can also find recordings and transcripts from the “black boxes” on airplanes. In all, a great database for many kinds of data.

Pssssst! Free occupational safety data!

Do you love databases? Especially if you are interested in safety, there are a number of carefully archived databases of events out there. A couple of years ago, I found one of these while trying to answer the question “What kinds of human factors interventions would be most increase agricultural safety?” Six months of coding later, I had some answers and a good direction for my research program*.

The posts describing the databases I have found are long, so I will spread them out over multiple posts under the category “Databases.” Today’s database is…

The Fatality Assessment and Control Evaluation (FACE) Progam – NIOSH

This database from the National Institute of Occupational Safety and Health covers fatal accidents in a number of areas with a standard interview and evaluation for each case. These are rich descriptions with the added benefit of having an expert in that field try to assess what went wrong. Some of the areas they cover are highway work zones, agriculture, commercial fishing, commercial aviation, construction, and logging. Each case includes a summary, descriptions, pictures, analysis, and recommendations to prevent a re-occurrence. Here is an excerpt of one of their cases:

Summary

Mid-spring 2005, a 29-year-old man died when the tractor he was operating overturned upside down pinning him underneath (Photo 1). The 40-year-old tractor had a narrow (tricycle) front axle. It did not have a rollover protective structure (ROPS). A front-end loader was attached to the tractor’s frame but no counter-weights had been installed for ballast. The loader with its bucket full of rocks was raised to nearly hood height. The tractor leaned to the right as the man steered it forward at a slight upward angle on a slope. The position of the heavy load, the absence of ballast, the tractor’s configuration, the dynamics of the tractor-loader combination and its load in transport on the sloping, uneven terrain contributed to the sudden overturn of this tractor. ROPS and use of the seat belt would likely have prevented this man’s death.

Photo 1 – Tractor without ROPS at scene of overturn. Note the rocks spilled from the loader bucket and height of the loader bucket relative to the tractor’s hood.

Photo 1 – Tractor without ROPS at scene of overturn. Note the rocks spilled from the loader bucket and height of the loader bucket relative to the tractor’s hood.

RECOMMENDATIONS based on our investigation are as follows:

  • Agricultural tractors should be equipped with a rollover protective structure (ROPS) and the seat belt should be used, except with foldable or retractable ROPS in their down or retracted position.
  • ROPS should be designed and readily available for all makes and models of agricultural tractors in common use..
  • Front-end loaders should not be installed on narrow-front (“tricycle”) tractors that do not have a ROPS.
  • Operators handling heavy loads should follow manufacturer recommendations for proper equipment, set up, ballasting, and safe operating practices.

And that’s just the summary! You can look at cases by state, by cause and any number of other variables.

John presenting poster at HFES 2009

John presenting poster at HFES 2009

Data visualization tools

Foreshadowing Anne’s upcoming series of posts on large, public, and free data sets, here are two interesting tools to help you visualize massive quantities of data. First, my grad student Margaux informed me of Google Fusion Tables (shown above). The site lets you upload data and visualize it in different ways. The website has some samples.

From the website:

Look at public data.

Get started with an interesting data set from the Table Gallery.

Import your own.

Upload data tables from spreadsheets or csv files. During our labs release, we can support up to 100MB per table, and up to 250MB per user. You can export your data as csv too.

Visualize it instantly.

See the data on a map or as a chart immediately. Columns with locations are interpreted automatically, and you can adjust them directly on a map if necessary. Use filter and aggregate tools for more selective visualizations

Pivot

Microsoft Pivot

Microsoft Pivot

Microsoft Pivot

Next, is Microsoft’s Pivot. Instead of being a web-based service, it is a program that runs on your computer (Windows only). It is currently available by invite-only (invite code available from TechCrunch) and I just installed it but the installer encountered an error so I am not yet able to play with it. But from what I’ve seen, it really provides a type of faceted browsing front-end for disparate sources of data.

Pivot makes it easier to interact with massive amounts of data in ways that are powerful, informative, and fun. We tried to step back and design an interaction model that accommodates the complexity and scale of information rather than the traditional structure of the Web.

TechCrunch got a sneak preview and wrote up a more detailed description:

The best way to understand the importance of Pivot is through a real-world example of how this technology would work. So let’s say I wanted a visualization of all the Wikipedia links to TechCrunch, Pivot would essentially crawl all of Wikipedia and create a map of the Wikipedia pages that are connected to TechCrunch, such as Michael Arrington’s Wikipedia page.

Another real-world use of Pivot is extracting data from Facebook. For example, you can use Pivot to crawl Facebook and break down friends by various data points like relationship status or college. Microsoft has an interesting example of Pivot being used to sort through Sports Illustrated covers, where you can break down covers into verticals by type of sport, team, athlete and more.