|
Index |
Results The telephone discussions focused on a set of eight questions or topic areas. Before turning to these specific findings, however, it is of interest to present some general results from the discussions: There were several bugs in the PBCAT version that was sent to the beta testers. These bugs were not debilitating and did not materially influence the results of the test. PBCAT generally made an extremely favorable impression on the participants. There was unanimous agreement that the program has utility for anyone who has to analyze pedestrian or bicycle crashes. It was apparent that PBCAT served as an excellent vehicle for teaching the concept of crash types and their relationship to countermeasures. PBCAT was considered easy to use. As discussed below, the beta testers were a computer literate group. Nevertheless, the discussions clearly suggested that even less experienced computer users should have no problem with the program. How much time did you spend working with PBCAT? Half the participants
spent about three hours working with PBCAT while the balance spent approximately
two days with the program. The difference was whether only the beta test
assignment was completed or the tester also attempted to use the software
for a real task. In either case, PBCAT was well liked and it met or exceeded
expectations.
Do you think you were at the plateau on the learning curve for PBCAT when you coded the test crash reports? If yes, how long did it take? If no, why not and how long do you think it would take to get proficient? All nine respondents felt they were at or near plateau on the learning curve for the PBCAT modules they used, primarily the crash typing module. All believed the program was easy to learn and to use. Do you
think the software functioned as if it were release quality? In other
words, was the test realistic with respect to software maturity?
There was also unanimous
agreement that the software had the look and feel of a release quality
product. The respondents did not have the impression that they were dealing
with a test version. Several volunteered that, absent the bugs that they
had reported, PBCAT was ready for release.
How would you characterize your computer skills on a scale of expert to novice? Eight of the nine
participants in this follow-up evaluation judged their computer literacy
as at least intermediate or higher. None encountered significant problems
installing PBCAT, although several mentioned that it would only install
on a "C" drive. There were numerous positive comments concerning the graphical
interface of the program. There was also general agreement that it did
not take a computer expert to install or operate PBCAT. The one participant,
who rated himself as "close to novice" in computer skills had no problems
running the program but did ask a "computer specialist" in his organization
to install it for him.
Was the
test representative of a real use situation? If no, why not (spent more
time, spent less time, wouldn't have done it yourself)?
All participants
said the test was representative of a real use situation of inputting
and crash typing. Few used the output modules because their use was not
part of the beta test. A few participants indicated that they would probably
assign the routine use of PBCAT to support personnel in their organization.
What is
your job responsibility?
The nine participants
represented a broad range of job responsibilities potentially related
to an interest in looking at pedestrian and/or bicycle crash data. These
included: Researcher/program
evaluator/data analysis specialist
Bicycle/pedestrian
coordinator
Planner
Coordinator of a
broad-based community safety program.
Varying levels of
management responsibility were also represented in the sample.
Would you or your organization use PBCAT? If so, for what? If not, who should use it (what is its niche)? All participants
indicated that they, their organization or the appropriate member of a
partnership or coalition they belonged to would use PBCAT if and when
the need to analyze pedestrian or bicycle crashes arose.
How could
PBCAT be improved functionally?
In discussing this
topic, it was made clear to the participants that the focus was on functions
and not on more microscopic design issues such as the appearance of screens
or the methods of navigating through the program. Other than this delimitation
of the scope of the question, the participants were given no specific
choices for additional functionality. Their responses were totally unaided
and therefore can be considered as stronger arguments for the suggested
additional functions than if the choices arose from a selection list.
This topic prompted
the suggestion that a link be created to a mapping or graphical information
system (GIS) from virtually every respondent. The participants have become
accustomed to using mapping programs to display and analyze data. They
saw great potential in linking PBCAT to a GIS program so that double entry
of data would be avoided. They also expressed their belief in the benefits
of graphical analyses for crash data.
Many of the participants
also suggested a stronger, more definitive link to recommended countermeasures.
There was even the suggestion to extend this link to the single crash
level. This would allow the analyst to obtain countermeasure recommendations
on a crash-by-crash basis. Since few of the testers were pedestrian or
bicycle countermeasure implementors, it is not surprising that they wanted
additional and more precise links between the crashes they coded and the
countermeasures to be employed.
Other functions that were mentioned at least once in response to this question included a module to permit the scanning of police crash reports for entry (to eliminate or minimize keying) and a link to the NHTSA standard laptop police crash reporting software. |