The sample was not representative of the total blind population. The percentage of people with disabilities, including blindness, increases with age. Many people with vision impairments are quite advanced in age, and they also might have other physical disabilities that would have made this type of testing impossible to conduct. Not included in this sample were those types of people who do not make independent trips, such as reported by Clark-Carter et al. (1986) . As has been made clear, the sample population required active and independent travelers to get to the test site, and, thus, they were not representative of all blind people. A true representative sample, even if condoned by human subject protocol and restrictions, would have probably not been able to complete many of the tasks using their regular methods. Fatigue and stress would also have taken their toll on subjects.
It is doubtful that this type of sample had any negative impacts on the validity
of this study. This research was designed to be a “real world”
experiment, and the group tested was likely to be representative of active blind
travelers who make independent trips into new environments. It is assumed
that, if a more representative sample had been used, the control group, using
their regular methods, would have made many more mistakes and not been able
to complete many more tasks than was the case with these people. It is
expected that the results would have been even stronger for the efficacy of
the auditory signs if the elderly and more dependent blind people had been included.
When asking blind subjects to locate and find 20 locations in a new environment, there would be many times when the goal could not be found in a reasonable amount of time. In addition, a valid experiment must keep stress, fatigue, frustration, and other negative feelings at a low level. For those reasons, some type of time constraint must be placed on these tasks. A four-minute limit was used on each of the 20 tasks. As many more people “timed out” when using their regular method, the effect of this limit was to actually reduce the time difference between the two conditions. A higher limit would have made the differences in the results even more robust but would have led to other problems. In a route learning experiment at the UCSB campus (Golledge, Marston, & Costanzo, 1998a) , it was noticed that, blindfolded, sighted subjects did not hesitate to give up on a task when they could not easily locate a target, whereas the blind subjects were very reluctant to ever give up before the time limit. In the Santa Barbara MTD experiment, (Golledge & Marston, 1999) , many subjects kept searching even when totally disoriented and were also very reluctant to stop their task. This is a vital and necessary search tactic for an independent blind person, and it was felt that this tenaciousness in terms of completing a task and not being seen to “fail” would have made the test much longer, stressful, and frustrating if more time had been allowed.
One problem with using the same limit for all tasks is that the performance differences were constrained by the upper bound of four minutes, although all routes were not of equal length. Thus, a route that would take a sighted subject or RIAS user two minutes could, at the best, be only twice as “effective” when compared to a subject who timed out at four minutes. In contrast, a shorter route, such as crossing a street or walking to a nearby location, could have a much higher effectiveness rating. One way to solve this problem would be to have each limit based on a multiple of the actual sighted control walking time. However, this could cause other problems. A walk that would take two minutes might need to have a limit of 20 minutes to make it comparable with another shorter control trip of 24 seconds and having an upper limit of four minutes. If the purpose of this experiment was strictly to measure the relative difference between conditions at different locations, then all distances should have been made equal to avoid this confound. However, the real world motivation demanded that people navigate though an environment and learn routes that did not consist of artificially equal distances.
7.7.3. Requests for Assistance
In most blind research, subjects are not allowed to ask for any help from others.
However, previous research (Golledge & Marston, 1999) showed that having
to rely on others was a major frustration in blind navigation and a very common
tactic that sometimes must be used. Again, it seemed that not allowing
people to use this everyday and normal technique to transverse this complicated
environment would have added to stress and frustration and not been representative
of how the blind travelers actually explore new spaces. The principal
researcher did not give any help to locate destinations, but subjects were allowed
to ask others for verbal information to help locate objects, although they were
not allowed to be led or guided to the location. This technique influenced
the results in two ways, both of which actually reduced the time and performance
difference between the two conditions, and, as such, actually subtracted from
the relative performance advantage of RIAS. First of all, no one using
RIAS for the first trial ever asked for help, and so, in that condition, subjects
actually had less input than available to those using their regular ( NRIAS)
method. Secondly, without the ability to ask for help, many more of the
disoriented people using their regular method would have failed to complete
the task in the allotted time. When searching for locations like concession
counters, track doors, or the bus stop, many of the regular method ( NRIAS
) subjects got very close but, without asking for help or identification, would
not have been able to complete the task. If subjects had not been able
to ask for information, the results would have been much more robust for the
RIAS condition.
At the present time, there are no urban areas that are fully equipped with RIAS. If there were, comparisons between actual travel data from blind people with and without the system could be made. In order to gather data about travel behavior and activity participation with RIAS, estimations of how they thought their actions would change were used. After using the RIAS in the field test, they were asked to estimate future travel activities if the environment had the same kind of coverage as did the test area. When these data were compared to pre-test responses of current behavior, the differences in increased activity participation and travel were very large. As explained in Chapter 6, there is little that is more important to a blind person than achieving access to independent travel and activity participation. For most blind people, it is assumed that there are daily affirmations of the negative impact of vision loss on independent travel and access to opportunities, so it was fully expected that they would be able to give well-informed opinions on how this system could affect their quality of life.
This methodology might raise questions of ecological validity; i.e., did their estimations accurately measure how RIAS would affect their travel behavior? All but one subject thought they would make more trips with RIAS, and all subjects thought they would waste less time on transit trips by using cues from RIAS to help them make transfers. These strong results leave little doubt about the perceived effect that RIAS would have on increasing access, activity participation, and travel efficiency and comfort. Actual trip data using RIAS, if and when possible to collect, would simply better quantify the magnitude of these effects.
Such a strong indication of missed and desired trips might be considered a
desire to please the researcher. There is also a tendency for respondents
to give “socially desirable” answers that has been well documented
in the literature, especially those concerning environmentally friendly or political
actions. More people report voting in an election or buying environmentally
favorable products than the actual participation data can support (Lam &
Cheng, 2002) . However, this desire by some to impress upon others that
they make socially desirable actions has little to do with the questions asked
in this experiment. There is nothing socially desirable about admitting
that activities are denied by the difficulty caused by one’s vision limitations.
Likewise, there is no social reason to exaggerate one’s willingness
to stay on a vehicle for long periods of time to avoid making a transfer.
Keeping in mind the strong sense of independence and self-worth exhibited by
most active blind people, it is just as easy to think that they would be hesitant
to admit how “restricted” their current lives were, or how much
they missed, by not having access to all that urban life has to offer.
The fact that they admit to so many missed opportunities is a strong indication
of how difficult travel without vision can be and how it affects every aspect
of daily living. However, in any survey, there always exists the possibility
that respondents try to influence the overall results by exaggeration.
| BACK TO OVERVIEW |
BACK TO TABLE OF CONTENTS
|
NEXT SECTION |