C S L
Notes
Number 21/22/23 December/March/June 1997
A quarterly newsletter of the
Cognitive Science
Laboratory, The
Catholic University of America, Washington, D. C. 20064.
Letter from the Editor
by David Hardy
It is a time of change at the Cognitive Science Laboratory.
New people are coming and old people (like myself) are going.
See the 'CSL News' section for details about new members.
Many structural changes are also in the works. A
new testing room has been built inside the basement room G7 where
EEG experiments will now be conducted. Room G8, which houses the
flight simulator and ATC simulator, is being renovated. In the
main office suites (Rooms 250-260), Room 257 mitosed into two
rooms (257 and 259), and Room 260 has been transformed into a
conference room (with new carpet and window blinds). I think CSL
members agree that these room changes greatly improve the look
and function of the lab.
As you probably know, this issue is far behind schedule.
As editor my mark on CSL Notes is, in my opinion, improved
quality but slower production rate--a tradeoff. Carol Cairns takes
over as editor after this issue and I'm sure she will leave
her own mark on CSL Notes. I've enjoyed editing
this newsletter. The CSL is a vibrant laboratory and I never had
to look far for submissions or worthwhile things about which
to write.
A Fond Farewell
By Helena Chandler
I'm embarrassed to say that the last time
I contributed to the CSL newsletter was at Evan Byrne's
request that I introduce myself as a new member of the CSL team
in October, 1995. This time I am writing because my tenure here
is coming to a close and I want to take the opportunity to reflect
on what I have learned over the past couple of years. I searched
back to my entry in the November/December 1995 edition of CSL
notes, hoping that it would jar my memory as to what my goals
were in joining the lab. Alas, I found no such insight and so
cannot say whether my original expectations were met. Instead,
I'll share an unexpected development, enthusiasm for the
process of doing research.
Over the past couple of years, I have found increasing
satisfaction in the details of research design, data collection
and preservation, and statistical analysis. Many researchers seem
to view these tasks as lowlier than hypothesizing and creating
new theory about behavior. But reckless story-spinning does science,
and particularly the more vulnerable science of psychology, a
great disservice. My work in the lab complemented and informed
my classes in statistics and methods, and I would like to thank
members of the lab for showing me the application of these courses
to their research. Here in the CSL, I have witnessed how important
it is to break the larger questions down into their component
details, and then piece the answers together to create an accurate
description of the construct of interest.
Before I fully transition out my Research Assistant
position in the lab, and into my Data Analyst position at Bell
Atlantic, I wish to thank everyone for their willingness to share
information and ideas, and encouragement of my attempt to find
my niche in psychology. I'd like to end with a quote from
Mary Nicol Leaky (1913-1996) who conducted a great deal of the
painstaking tasks of collecting data and eventually unearthing
evidence of human origins in Africa with her husband Louis Leaky
who was more famous because of his theories of human origins (now
being questioned): 'Theories come and go, but fundamental
data always remains the same'.
Report on the Ninth International Symposium
on Aviation Psychology
By Scott Galster
On April 27-May 1, 1997 several members
of the CSL attended the Ninth Biennial International Symposium
on Aviation Psychology held in Columbus, Ohio. Raja Parasuraman,
Jackie Duley, and Tony Masalonis all presented papers at the conference.
'Human Use and Abuse of Automation', 'Effects
of Display Superimposition on Monitoring of Automation',
and 'Instrument Failure Detection and Workload in Simulated
General Aviation Flight During Manual and Automated Lateral Tracking'
were the respective titles of the presentations. Authorship
of the papers presented were not limited to the speakers and can
be obtained by contacting the CSL.
The conference was attended by over
seven hundred guests representing aerospace corporations, private
research organizations, universities, airlines, NASA, FAA, and
other domestic and international entities. CSL members agreed
that the conference was a productive and stimulating venue for
the discussion of current and future aviation trends.
The Human Factors and Ergonomics Society
is the next major conference CSL members involved in human factors
research are planning to attend. The conference will serve as
the inauguration for reporting the results of the CSL research
activities devoted to Free Flight issues in the National Airspace
System.
Calendar of Events
May 23-26, 1997. American Psychological Society
Ninth Annual Convention. Washington, DC.
August 15-19, 1997. American Psychological Association
105th Convention. Chicago, IL.
September 22-26, 1997. Human Factors and Ergonomics
Society 41st Annual Meeting. Albuquerque, NM.
October 15-19, 1997. Society for Psychophysiological
Research 37th Annual Meeting. Cape Cod, MA.
October 25-30, 1997. Society for Neuroscience
Annual Meeting. New Orleans, LA.
News from the Laboratory System Manager
by Tony Masalonis
Surfing CSL
The CSL Web page has undergone modifications, especially
in the Staff section. The newest group picture, taken in the Library
in the renovated Room 260, is now viewable on the site. The staff
list is also updated and reorganized for a more streamlined appearance.
Clicking on an individual name takes you to that person's
resume/vita. CSL staff members who have not provided a vita
should, at their earliest convenience, provide an electronic copy
of the document . The goal is for the CSL Website to have everyone's
resume, including undergrad research assistants! If you don't
have a full-fledged resume, you may want to write a paragraph
or two about yourself, what you do at CSL, and your professional
activities and interests.
The Web page can be viewed at:
http://www.acad.cua.edu/as/psy/csl/
It is also linked from the CUA home page, under
Departments,
and as of this writing is also under the What's New section
on the CUA page.
E-Mail
The lab now owns Claris E-mailer, a point-and-click
application that will retrieve your e-mail from your Vax account
without having to log on. It retrieves your mail for you at an
interval you pick (say, ten minutes), and keeps you from using
up your Vax storage space (because the mail messages come to the
Mac's hard drive) and your Vax money (since it's
not necessary to log on to send and get messages).
Backing Up
Another important acquisition for the lab is the
Jaz drive, which holds 1 gigabyte (that's over a billion
bytes, which is a lot of bytes) on each cartridge. All lab members
with their own desktop computers are advised to back up the newest
contents of their hard drive at least every two weeks. Important,
irreplaceable files, such as data, should be backed up in three
places:
- on the Jaz,
- on some other media of your choice (e.g., optical
drive),
- in a third location (either on the hard drive
or in hard copy).
At least one of your backups should be in ASCII 'text
only' (for documents) or tab-delimited text-only (for data
tables and spreadsheets), so that when you need it years from
now, it won't be dependent on some software that might
no longer work on the computers of the future (like Norton Backup,
or even Word - Bill Gates is, amazingly enough, not immortal).
Feel welcome to consult your system manager for help
with backing up, or with any questions about the Website or e-mail
software.
CSL Publications
Clark, V. P., Parasuraman, R., Keil,
K., Kulansky, R., Fannon, S., Maisog, J., Ungerleider, L. G.,
Haxby, J. V. (1997). Selective attention to face identity and
color studied with fMRI. Human Brain Mapping.
Duley, J., Molloy, R., Parasuraman,
R. (1997). Display configuration
in adaptive automation: integration and dynamic presentation.
In M. Mouloua & J. Koonce (Eds.), Human-automation interaction:
Research and practice. (pp 109-118). Mahwah, NJ: Lawrence
Erlbaum Associates.
Greenwood, P. M. & Parasuraman, R. (1997). Attention
in aging and Alzheimer's Disease: Behavior and neural systems.
In J. A. Burack & J. T. Enns (Eds.), Attention, development,
and psychopathology. New York: Guilford Press.
Greenwood, P.M., Parasuraman, R., & Alexander,
G.E. (1997). Controlling the focus of spatial attention during
visual search: Effects of advanced aging and Alzheimer's disease.
Neuropsychology, 11, 3-12.
Hilburn, B., Jorna, P. G., Byrne,
E. A., Parasuraman, R. (1997).
The effect of adaptive air traffic control (ATC) decision aiding
on controller mental workload. In M. Mouloua & J. M. Koonce
(Eds.), Human-automation interaction: Research and practice.
(pp. 84-92). Mahwah, NJ: Lawrence Erlbaum Associates.
Molloy, R. J., Byrne, E. A.,
Masalonis, A.
J., Parasuraman, R. (1997).
Laboratory flight simulator for automation and display design
research. In M. Mouloua & J. Koonce (Eds.), Human-automation
interaction: Research and practice. (pp. 151-156). Mahwah,
NJ: Lawrence Erlbaum Associates.
Parasuraman, R. (1997). Human use and
abuse of automation. In M. Mouloua & J. Koonce (Eds.),
Human-automation
interaction: Research and practice. (pp. 42-47). Mahwah, NJ:
Lawrence Erlbaum Associates.
Parasuraman, R., & Riley, V. (in press). Humans
and automation: Use, misuse, disuse, abuse. Human Factors.
Singh, I. L., Molloy, R., & Parasuraman, R. (1997).
Automation-related monitoring inefficiency: The role of display
location. International Journal of Human-Computer Studies,
46, 17-30.
Guest Column: Timing Issues of Superlab
By Mike Miller
While research has progressed considerably since
the days in which performance-based psychometrics depended primarily
upon a stopwatch and a good eye (or ear), the increased use of
computers in psychological research has not yet freed us from
timing accuracy concerns. Indeed, many new concerns have arisen.
There are a wide variety of issues depending upon the platform,
IBM-compatible or Macintosh, the software used, and the type of
input/output (I/O) methods employed. The purpose of this article
is to address the timing issues involved when using the Mac, SuperLab
software, and a standard keyboard and microphone. The difficulties
essentially center around stimuli display and reaction input.
However, before addressing these issues it is first necessary
to briefly describe Mac video display and I/O.
The Mac has two means of I/O: RS-422 serial ports,
commonly called the 'printer'; and 'modem'
ports, and the Apple Desktop Bus (ADB), used to connect the keyboard
and mouse. Timing issues when using the serial ports are negligible
and limited to the internal efficiency of SuperLab, because the
ports communicate at a sufficiently rapid 57,000 bits per second.
The ADB is somewhat more problematic. It is essentially a small,
self-contained network, designed to support up to seven daisy-chained
devices such as keyboards, mice, and graphics tablets, each with
an identifying address. The devices might be best thought of as
'clients' and the Mac as the 'server.' The
standard Apple keyboards, as well as the Extended and Extended
II models, employ a 2 MHz Motorola 68CH11 microcontroller to gather
and send the keypress and mouse click information to an I/O chip
in the Mac. The 68CH11 has several 2-8 byte registers in which
it stores keypresses and mouse clicks, and the addresses of the
devices from which the information was obtained. It does not send
the contents of these registers to the Mac, and this is crucial,
until it is requested to do so by the I/O chip. The 68CH11 cannot,
however, send information to the Mac any faster than every 2 ms
because of its clock speed and for that reason timing can never
be more accurate than 2 ms when using the ADB keyboard.
So far, the interaction between the devices and the
Mac have been purely in the hardware domain. Enter software: The
Mac I/O chip is controlled by the ADB Manager, which is a special
type of event manager in the Mac operating system. The ADB Manager
polls the ADB devices for information approximately every
11 ms. The interval may vary depending upon CPU activity when
the ADB chain is polled. The MacOS, for instance, uses valuable
CPU cycles to continually polls media devices for insertion of
new disks and CD's. Normally, this is an advantage because the
Mac automatically recognizes new disks. However, this polling
is partly responsible for the variation in ADB timing. Additional
variation is caused by the growing number of extensions and control
panels used in most Macs, which patch the MacOS and cause
considerable slowing noticeable even in daily use.
There are three methods of bypassing the ADB Manager:
1) Use an ADBS resource in the program, 2) An ADB INIT, and 3)
Inclusion of an ADB driver library or header in the source code
of the program. Regarding the ADBS resource, this method was officially
abandoned by Apple as of MacOS v6.0.4, circa 1991, and is no longer
recommended. It is easily detected by viewing the resource fork
of an application and examining it for an ADBS resource. The ADB
INIT is an extension that loads on startup and replaces the MacOS
ADB Manager. The final method requires access to the source code
to verify. However, even if this last method is used, ADB device
timing will have no better than 2 ms accuracy.
SuperLab does not contain an ADBS resource, nor does
it include any extensions. There are ADB drivers available that
provide accuracy at about 2 ms. SuperLab's author assures that
he has included an ADB driver in the software as of v1.5, but
also warns that accuracy with the keyboard is no better than 16.67
ms. Based on the available information, either the driver is no
better than the ADB Manager or was simply not included and SuperLab
depends upon the MacOS for ADB information. Although doubt about
the inclusion of an accurate driver is disconcerting, dependence
on the MacOS may actually be worse because the timing interval
between polls will vary, also causing timing errors to vary.
In addition to the ADB timing accuracy issue, there
exists the possibility for errors in measurement to occur in the
interval between the initial presentation of a color visual stimulus
and when SuperLab starts and stops the reaction time to that stimulus.
SuperLab's author recommends using a feature called 'instant
switching' to minimize these errors. Instant switching enables
a stimulus to be completely loaded into memory and then presented.
A stimulus is displayed more quickly and timing does not begin
until the screen is completely drawn, and the next screen is begun.
Since the screen is typically drawn at about 67 Hz, or about 67
times each second, a small image 20 pixels high will have a delay
between presentation and initialization of timing of about 6 us
(.0006 sec). The actual delay will depend on the size of the image.
A full screen image would require about 2 ms. Black and white
images require significantly less time to draw and are not subject
to the same penalty.
Timing accuracy issues with the microphone are more
easily resolved. The Mac uses the Apple Sound Chip (ASC), manufactured
by Sony, to drive the microphone port. The ASC captures sound
at a resolution of up to 44 kHz, depending upon the device used.
The standard Omni microphone included with most Macs has a resolution
of only 22 kHz, and the PlainTalk microphone has a resolution
of 44 kHz. AV series Macs have a different chip set which allows
sound capture at up to 66 kHz with a PlainTalk microphone. When
using the microphone on a Mac, the hardware (both the internal
chip set and microphone) determines the resolution of sound capture.
Almost any combination of hardware will provide accuracy better
than 1 ms. The only real issue is, again, in how timely and efficiently
SuperLab uses this information.
In conclusion, certain timing accuracy caveats apply
when using the Mac and SuperLab. First, ADB (keyboard) timing
resolution is no better than about 16 ms. Microphone timing resolution
is better than 1 ms when used as the sole source of input. In
conjunction with the keyboard, resolution drops to about 16 ms.
Finally, instant switching minimizes, but does not eliminate,
the delay between stimulus presentation and initialization of
the reaction time clock. SuperLab is an easy to use method of
presenting auditory and visual stimuli and recording reaction
times, but certain measures must be taken to avoid the issues
mentioned here, or at least to minimize their effects.
1) To avoid unnecessarily using valuable CPU cycles
that will affect timing regardless of whether you use an I/O board,
start up the Mac with extensions off (start the Mac and hold down
the shift key until the 'Welcome to Macintosh. Extensions
Off' appears). Using an extension manager is insufficient
because several types of extensions and control panels are not
recognized by these managers (including Apple's!). To be 100%
sure, manually move ALL extensions and control panels from their
location in the System folder to another folder. Usually, these
folders are named 'Extensions (Disabled)' and 'Control
Panels (Disabled).'
2) Insert floppy disks and CD's into all available
devices to avoid having the MacOS poll for new media. The MacOS
stops polling after identifying new media.
3) Use the microphone alone whenever possible to
obtain better than 1 ms timing accuracy.
4) Avoid using the keyboard or mouse. For the best
results, use a digital I/O board, such as the National Instruments
LabLC board, or the SuperLab button box to obtain better than
1 ms accuracy with keypress type responses.
5) Keep stimuli images as small in size as possible
and use instant switching. The background can be changed to whatever
color necessary using the 'Background, Color...' menu
option. Also, use a program that allows you to save images with
an indexed color palette. This palette contains fewer colors and
reduces considerably the time required to display images. Adobe
PhotoShop is one such program that has this capability.
6) Avoid using the Mac II and earlier models. Unless
an additional video card is installed, these models rely on static
RAM for display memory. This introduces additional uncertainty
about how memory is managed and images are displayed. PowerMac
models are fast enough to avoid display memory timing issues altogether
and should be the platform of choice when using SuperLab. National
Instruments will make available, if it is not already, a PCI version
of their LabLC board for newer model PowerMacs.
CSL News
Welcome!
Welcome to Scott Galster, Carol Cairns, Mike Le,
Jenny Engle, Chris Wilson, Danielle Yates, and John Perez. Scott
is in the Ph.D. applied/experimental psychology program and joined
the CSL full time in January. Scott comes to us from the University
of Cincinnati (he has worked in Joel Warm's laboratory)
and will be specializing in the free flight/ATC human factors
research. Carol is research officer for the CSL. Carol's
is a familiar face to us from the psychology department where
she was assistant to the chair for over six years. Mike, Jenny,
Chris, Danielle, and John are undergraduates (all psychology majors
except for Mike who is an electrical engineering major) here at
Catholic University. Mike is computer programming for various
human factors projects. Jenny is an assistant to Carol. Chris
is being trained for the EEG research. And Danielle and John are
helping with the cognitive/aging research on Superlab.
Congratulations!
Congratulations to Sangy on her successful oral defense
of her doctoral dissertation March 21, 1997. The title of her
dissertation was 'Cholinergic Substrate of Visual Selective
Attention.'
Congratulations also to Jackie Duley, Tony Masalonis,
and Sean Fannon. Jackie and Tony passed their comprehensive exams
this past Fall and Spring semester respectively. Sean graduated
this Spring semester 1997 garnering multiple honors, including
the psychology department's 'Distinguished Psychology
Major' award, Phi Beta Kappa, and Magna Cum Laude of his
class. Way to go Sean!
New Grants
In December 1996, the CSL was awarded a three year
$1.1 million grant from NASA Ames Research Center for human
factors studies of Free Flight.
In April 1997, the CSL was awarded a follow-on grant
of $75,000 from NASA Langley Research Center for continued study
of adaptive automation.
Other continuation grants include an NIA grant for
studies of attention in aging and Alzheimer's disease.
Journal Club
The Cognitive Aging Journal Club will begin some
time in the first part of June. The purpose of the journal club
is to discuss current journal articles in the area of cognitive
aging and other related topics under cognitive psychology. A paper
will be chosen and discussed at every meeting. Everyone who plans
on attending the meeting will receive a copy of the journal article
a week in advance. Those interested are encouraged to choose papers
that they may be interested in to discuss at the meeting.
If anyone is interested in joining the journal club
please let Stephanie Johnson know what days of the week you are
available. Meetings will most likely take place once a month.
Presentations
The human factors CSL members in collaboration with
Brian H to Dynamically Adjust the Attentional Focus from Youth to Old
Age to Alzheimer's Disease', at the Society for
Neuroscience meeting at New Orleans.
Jackie Duley, with Scott Galster, Tony Masalonis,
Brian Hilburn, and Raja Parasuraman, are presenting a paper, 'En
Route Controller Information Requirements from Current ATM to
Free Flight', at the Free Flight conference in Amsterdam,
October 20-21.
Farewell!
Good-bye and good luck to David Hardy. David is moving
out west to Los Angeles (his home) for a post doc with Paul Satz
at the Neuropsychiatric Institute at UCLA. He's not gone
for good however. He'll be back (soon!) to defend his doctoral
dissertation.
A Quote to Note
'Applied psychology is much more than cleverness and common
sense using the facts and principles found in the standard texts.
It is scientific work, research on problems of human nature complicated
by conditions of the shop or school or army, restricted by time
and labor cost, and directed by imperative needs'.
Edward L. Thorndike (1919)
Science, Vol. 49, No. 1255, p. 60
Reading for Work . . .
- Handbook of Human Factors and
Ergonomics
(2nd ed.)
G. Salvendy (Ed.), Wiley & Sons, Inc., 2137 pp.,
$165, 1997.
- Handbook of Human Factors and the Older
Adult
A. Fisk & W. Rogers (Eds.), Academic Press, 419
pp., $65, 1997.
- Attention: Selection, Awareness, and Control.
A Tribute to Donald Broadbent
A. Baddeley & L. Weiskrantz (Eds.), Oxford University
Press, 436 pp., $45, 1993.
- From Jackie,
Flight to the Future
Human Factors in Air Traffic Control
National Research Council, National Academy Press,
368 pp., 1997.
. . . or Pleasure
- The Night is Large
Martin Gardner, St. Martin's Press, 586 pp.,
$29.95, 1996. Collected essays from 1938-1995 by 'one of
the great intellects produced in this century or any century.'
Topics bounce from 'Superstrings' to 'Klingon
and Other Artificial Languages' to 'The Popperism
of Sir Karl'. There's even an essay about William
James.
- From Pam,
How the Irish Saved Civilization
T. Cahill. Explains how Irish monks preserved the
remnants of Western civilization after the fall of Rome.
- From Jackie,
Annapolis
William Martin. Historical novel about the power
and politics of early Annapolis.
Dog Attention -- A New Line of Research for the CSL?
CSL research (and reference to Raja Parasuraman)
is mentioned in the May issue of New Scientist. The article,
'Pay attention Rover', is about the brain mechanisms
of selective and sustained attention in dogs who have been trained
and bred to sniff out drugs and explosives. Copies of the article
are available. Check it out--an amusing side application of attention
research?.
CSL Notes
Editor: David J. Hardy
CSL Notes is published
quarterly (September, December, March, and June) by the Cognitive
Science Laboratory, The Catholic University of America. Inquiries
concerning subscriptions, submissions, suggestions, or for
general information on the CSL, should
be sent to: Carol A. Cairns, The Catholic University of America,
Cognitive Science Laboratory, 620 Michigan Ave., Washington, DC
20064; or call 202-319-5825; fax 202-319-4456; or e-mail cua-csl@cua.edu.
Cognitive Science Lab
Home Page
