2.6. Previous UCSB Experiments With Auditory Signage

The research reported in this dissertation is the latest of four research projects carried out at UC Santa Barbara, all investigating problems of transit use by those with vision impairments.   The work on the first three projects was performed as part of the California PATH Program of the University of California and the latest project was partially funded by the University of California Transportation Center.   The Principal Investigator was Dr. Reginald Golledge from the UCSB Department of Geography and the Research Unit on Spatial Cognition and Choice.   The first project consisted of interviews with 55 vision-impaired subjects from the Santa Barbara area.  
 

2.6.1. Santa Barbara Interview of Transit Use and Opinions

The purpose of these interviews was to gather data on transit use, frustrations, opinions, and improvement suggestions from a group of vision-impaired subjects (Golledge et al., 1995 ; Golledge, 1996; Golledge et al., 1997 ; Marston et al., 1997) .  These data were needed in order to frame more specific future studies.   The subjects reported a high degree of frustration when using transit, and the most important finding was that access to more information was what was needed to reduce transit-related difficulties and increase transit use.   The subjects also reported a need for more information about schedules, bus numbers, routes, and locations of bus stops and terminal amenities.   They thought that auditory messages would help them cross streets in order to make transfers and would also help if installed on buses and in terminals.  There were many elderly people in the survey who did not make many trips, and there were also 10 subjects who had access to a household car.   People with access to a household car had a much harsher view of transit use.   Their estimates on how long they wait for transit were much higher than those of subjects who had no car to rely on and who used transit more often.   Those who had no household car reported longer wait times for a ride (from friends or others) than in waiting for transit.   These findings led us to investigate the use of auditory signage (Talking Signs® ) to determine if it would provide sufficient information needed to increase path following accuracy, decrease walking and search times, and make finding a bus easier in actual field test conditions.  

  2.6.2. Path Following and Finding a Bus

This experiment consisted of two field experiments to test the effectiveness of Talking Signs® (Golledge, Marston, & Costanzo, 1998 a & b; Marston & Golledge, 1998a) .  Several path-following tests were conducted, using a square or rectangle shaped path, and then a field test was conducted where subjects tried to identify a specific bus among a group at a busy bus stop.  Subjects consisted of 10 legally blind subjects from the local blind community and 10 sighted students.   All subjects performed the tasks blindfolded, thus giving the opportunity to regard the 10 blindfolded sighted subjects as newly blinded subjects who, unlike the blind users, had no previous experience in blind wayfinding and navigation.   These sighted students also had no formal Orientation & Mobility training, during which people with vision restrictions had received many hours of training to help them follow routes, either using a long cane or a guide dog.  
 
The first experiment was set in an open field with 4 stanchions set either in a 60’ x 60’ square or a 30’ x 60’ rectangle.   Subjects were led around the shape three times and then were asked to follow the path on their own, twice in a forward direction and then once in the reverse order.  Without Talking Signs®, the 10 blindfolded sighted subjects only found 14 of 120 stanchions.   The vision-impaired subjects did better, finding 35 out of 120 attempts, although 15 of those 35 successes were accomplished by just two subjects who used echolocation to identify the targets.   When using the RIAS technology, all 20 subjects found all the stanchions in a timely fashion.   This test provided strong evidence that, with just a few minutes of training, RIAS could increase speed and accuracy in locating objects and in successful completion of a path of travel.
 
The second field test was conducted at the UCSB bus circle, where three or four buses at a time might be waiting.   Finding bus stops, and, especially, identifying the proper bus when many are present, has always been a difficult task for people with vision restrictions.  With limited or no vision, people are forced to approach each bus that they hear, find the door, and then ask the driver or others for the route or bus number.   This task can lead to missed connections and, often, unkind remarks from irate drivers or passengers.
 
The bus identification experiment started with two practice walks from the West side of the bus circle to the boarding area, subjects walked halfway around the circle and crossed two service roads.   The same two groups of nine (there was one no-show) blindfolded sighted people and 10 blindfolded people with vision restrictions were used.   All subjects tried the walk and bus identification task three times, first using Talking Signs ®, then with their regular method, and then again with TS.   All of the people with vision restrictions were able to find the proper bus after walking around the loop when using TS during both trials.   Without the system, 8 of 10 were able to find the proper bus in a timely manner.  Of the subjects who were sighted and blindfolded, without the system they were only able to find the bus two of nine times.   When using the RIAS, they found the bus five out of nine times on the first trial and seven out of nine times on a second trial.   Those who missed the bus using the system still got to the proper area but not in time to catch the bus.   The elapsed times for the trip for both groups were higher without the system than with RIAS.  The strong results of this experiment gave us motivation to test this system in more complex environments.   Post-test evaluations and comments from the subjects were very positive about the usefulness of these devices, and many people expressed a desire to have them installed; respondents also mentioned that they appreciated the fact that they did not have to ask for help.   Some of those comments led us to design more situations and questions for further exploration.
 

2.6.3. Santa Barbara MTD Bus Terminal Experiment

The previous experiment used four temporary RIAS transmitters in two different field tests.  Based on the positive results and acceptance of the system by the blind subjects, the next step was to test the system in a more extensive and “real world” environment.   Ten transmitters were permanently installed at the Santa Barbara, CA Metropolitan Transit District (MTD) bus terminal.   Using several more temporary installations, a round trip route by bus from the local Braille Institute to the main terminal and back was designed (Golledge & Marston, 1999) .  Twenty-seven people with visual impairments, recruited through the Braille Institute and other agencies, were used.  Subjects found bus stops, identified the proper bus on the street and in the bus staging area, found the location of amenities in the terminal, and simulated making several transfers between buses.  More independence was given to the subjects in this experiment.   That is, they were allowed to search and walk to locations without first being led to them several times.   This was much more akin to what happens when people with vision restrictions have to perform typical wayfinding tasks.   For example, subjects were told a bus stop was 120 feet away, and, with no further instructions or a practice walk, they searched for it on their own.  When they arrived by bus at the station, they searched for the terminal entrance without directions or practice. 
 
Dynamic spatial relationships are a constant source of problems even for the best blind traveler, who might master many spatially static environments.   Fixed locations and routes can be learned by rote practice, but an ever-changing configuration of buses at a bus terminal staging area cannot be learned and poses much uncertainty for blind travelers.   This experimental design gave much more data on how the lack of sight affects the ability to complete travel and how the use of RIAS could overcome the lack of visual cues.   For example, while leaving the station in order to find the proper bus for the return trip, those with little or no sight had to ask for help from strangers or find a bus by walking toward a sound or shape, locating the door, and then asking the driver to identify the bus.   If this was not the proper bus, they were usually “pointed” in the right direction to try again.   When using RIAS, they could scan the area, pick up the proper bus transmitter, and walk directly to the bus door.   The response times in finding proper locations were highly significant for the RIAS condition.   Subjects located two bus stops, found the terminal entrance, and located buses in much less time, with fewer mistakes, and without having to ask for help as often. 
 
Some tests of spatial knowledge acquisition of various amenities in the terminal were also conducted.   The area was often very crowded, and subjects were first led around the area three times before they tried it on their own.   A pointing task and an inter-point distance estimation task were used to record their estimation of the location of different amenities in the terminal.  Multidimensional scaling was used to determine if there was a significant difference between the two conditions; a person’s regular method, and when using RIAS .  No significant differences were found in those two tasks.   The pointing task was confounded by magnetic interference when using the compass to record direction.   The terminal was quite small, and it appeared that subjects learned the area so well with the three guided walks that their previous walk through the area overshadowed any benefit of using RIAS in that location.    
 
Based on the user comments that were obtained in the previous experiment, many more questions were asked after the test was concluded.   Data were recorded on subjects’ opinions of the usefulness of the system at various locations, their overall opinion of the system, and where they would like to see them installed.   The answers were very positive regarding the usefulness of the system.   Respondents mentioned many places they would like to see them installed and also had praise for the system in general.   In order to learn more about how difficult travel without sight could be, a series of questions was asked about the stress and difficulty of various transit tasks, such as finding a bus on a street or in a busy terminal, transferring buses, crossing streets, and finding a bus stop.   These questions were asked before the test in order to establish a baseline for subjects’ current practices and then asked again after they had used the system.  These results were highly significant.   Subject ratings were recorded on a five-point Likert scale, and, in many cases the ratings were a full two or more points better after they used the RIAS.   Many of the subjects used an agency door-to-door van service, which picked up and dropped off a large group of people.   Subjects were asked if they would switch to the fixed route bus system if it was made more accessible and easier to use.   Most subjects said they would rather use the local bus routes, however, several people said they did not mind the long van ride, that they spent the time talking with other riders.   This attitude toward the value of time prompted further inquiry into how vision-impaired travelers viewed saving time by making transfers or staying on a less direct bus.   These discussions with travelers suggested including further investigations into transfer-making decisions in the next experiment.  
 
That experiment was, until then, the most comprehensive study on RIAS, combining real world travel and wayfinding, terminal and amenity search, user input on transit and transfer difficulties, and opinions of the RIAS system.  It proved overwhelmingly that RIAS made travel and bus use easier and quicker.   Subjects unanimously stated that they did not have to ask for help when using the system, and they felt more independent.   These findings prompted the design of a further experiment with even less spatial information and path training provided to subjects.  
 

2.6.4. Findings from Previous Work that Warrant More Research

To highlight what led to the desire to conduct an even larger experiment, using more modes of transit in a larger, more urban environment, four sections of the Santa Barbara MTD experiment are summarized.
 

2.6.4.1. Finding A Bus Stop  

Subjects were asked to explain the difference between RIAS and their regular methods of navigation when finding an unfamiliar bus stop.   APPENDIX 2 lists all subject’s comments; a few are listed here as examples.
Twenty-six subjects gave 62 responses to this question; these were parsed and categorized as follows.

Table 2. 3  Bus Stop, User Response Categories

“What is different from your regular method when using Talking Signs® at a bus stop?”

Category

26 subjects

Gives direction

15

Gives positive identification

15

Confidence, assurance

14

More efficient, easier travel

13

Don’t have to ask

5

2.6.4.2. Finding the Proper Bus

A RIAS transmitter mounted on a bus sends out a signal that can reach over 100 feet.  This signal can contain the bus name, number, direction or other route information so that users know in advance which bus is coming and allows time to reach the boarding area and flag the proper bus, with complete and positive knowledge about which bus is approaching, where it goes, and the location of the entry door.  Figure 2.7 shows a bus equipped with the RIAS transmitter mounted on the front.  

Figure 2. 7   Using RIAS to Identify an Approaching Bus

Source: R, G. Golledge (2001) Reproduced with permission of the author.

Subjects were asked to explain the difference between RIAS and their regular methods of navigation when finding and boarding the proper bus.   APPENDIX 3 lists all subjects’ comments, a few are listed here as examples. Twenty-six subjects gave 89 responses to this question; these were parsed and categorized as follows.
 

Table 2. 4   Finding and Boarding Proper Bus, User Response Categories

“What is different when using Talking Signs® to find the proper bus?”

Category

26 subjects

Don’t have to ask

23

Easier, quicker

18

Positive identification of bus

14

Boarding location (door) information

10

Independent

8

Safe and secure

3

Less stress

3

2.6.4.3. User Ratings of Talking Signs® 

At the end of the MTD terminal experiment, subjects were asked to rate their approval of various installation locations for the system.   They were also asked how the system would affect their travel if it was installed on transit and at various locations.   These tables are sorted with the highest ratings first.   “Strongly agree” = 1, “Agree”= 2, through   “strongly disagree” = 5.   All responses were quite positive.   

Table 2. 5 User Opinion of RIAS : Specific Locations and Travel Behavior

User Opinion about Talking Signs® Installations

Ratings

I would like TS installed on all buses

1.1

I would like TS installed at bus stops

1.1

The TS in the terminal should be made permanent

1.1

TS on retail and other buildings would help me navigate and let me know what shopping or activities were available

1.1

TS in the MTD terminal are very helpful

1.2

I would like TS installed at street crossings that tell what street I am at and which direction I am facing.

1.2

I would like TS installed at crosswalk to keep me in the walkway and tell me the WALK/DON’T WALK signal.

1.3

 

User Opinion about Talking Signs® and Travel Behavior

Ratings

I would be more independent using TS

1.2

I would feel safer when I traveled

1.2

I could be more spontaneous when planning trips

1.2

I would take more trips

1.3

These very positive responses to the value of RIAS at bus stops and identifying the proper bus demanded more investigation into how this system would help in other transit environments.   The equally strong ratings about the efficacy of RIAS at various locations led to the examination of other specific types of locations beyond that of a bus terminal and in a more varied environment.   The strong opinions that RIAS would positively affect travel behavior led to a desire to test these sentiments in more empirical and robust experiments.   More information was needed to determine what people thought about their current trip-making behavior and what could be done to make it more equal to the general public.  
 
A pre and post-test question was asked that attempted to reveal the feelings of these vision-impaired subjects on their overall attitude toward equal access, as it is now and how it would be if RIAS was installed throughout the environment.   The results were so strong that they also demanded more research investigating why access was so limited and also to find what specific locations caused these problems and what mitigating effects the addition of environmental cues, such as location identity and direction, had on increasing access to urban opportunities.   The following two questions were asked with a five-point scale, ranging from 1= “strongly agree” to 5 = “strongly disagree.”  

Table 2. 6  ADA Compliance Measures, Pre and Post Talking Signs®  

 

Rating

Pre-test: I feel that I can get information and then find, access and use public buildings and transportation and that I enjoy the same access to buildings and transit given to the general public.

4.5

Post-Test:   If TS were installed on public buildings and transportation, I would have the same access given to the general public.

1.3


These two questions revealed that people with vision impairment do not think that they are getting the equal access that was mandated for them in 1990, and that they feel like a system that gives environmental cues would greatly help them to achieve this elusive goal of social equity and access.   Furthermore, in the survey, they also voiced a strong support for citywide RIAS installations and legislation to make installation mandatory.  
 
To better understand the problems of blind navigation, and building on past work, there appeared to be a need to collect more data on making transfers, about which locations were the most difficult, and where the use of RIAS could have the most benefit.   Subjects had mentioned many times how difficult using transit was, and this inspired a decision to ask further specific questions to determine how vision loss restricts independent travel and if RIAS could provide more access to urban opportunities and increase access to jobs, other activities, and travel.  There appeared to be a need to collect more empirical data about the problems of transit use and transferring between different transit modes.   More knowledge was also desired about how travel without vision affects trip-making behavior, limitations on activity space, and participation.   Little is known about restrictions and barriers to making transfers and how people with little or no vision perceive these barriers.   There is a need for data about how this group reacts to these barriers and if they have a different internal resistance to distance or change.   There is a need to understand if the distance decay function is different for this group than that exhibited by the general public.   Previous research on RIAS has either tested the efficacy of RIAS by itself or studied success when using the system compared to users’ regular methods.   There is also a need to test the efficacy of the system using a comparison to the walking speed of a sighted person.   Clark-Carter et al.   (1986) point out that, when testing different aids for persons who are blind, in addition to measuring errors and wrong turns, researchers must realize that the speed of the subjects reflects their ability to use an aid to increase travel skills.  By comparing the walking and search times of blind subjects using their regular methods or RIAS to a standard baseline derived from a sighted person, time penalties and their mitigation with a new aid can be clearly identified.   This method also allows for a broader understanding of which locations present the biggest problems to independent travel and which locations are more easily accessed without sight.   To be serious about understanding and then improving travel for people who are blind, researchers must be able to identify which locations cause the most problems for the blind and what aids or instruction can best be used to increase travel, independence, and quality of life for this group.  
 
Thus, the design of this research experiment evolved over many years of prior research into the needs and travel restriction of people with limited vision.  Starting with a comprehensive interview schedule to determine attitudes and needs, experiments progressed next to very controlled situations and then proceeded to the first “real world” experiments.  The empirical data, plus the comments from subjects, led to the design of the present, much more comprehensive field test.  
There has been criticism in certain academic circles that disability research is not representative of disabled people’s experiences and knowledge.   For example, Kitchin (2000) found such a lack of representation in his research with disabled people, and he stresses that research needs to be “carefully selected, presented in a way that is unambiguous, has a clear connection between theory and the lives of disabled people, and needs to be acted upon” (Kitchin, 2000, p. 29) .  Kitchin also states that disabled subjects are concerned that much research is ineffectual in transferring research results to real world improvements in their lives by helping to dismantle barriers.  A main concern was that their knowledge and experiences were “mined” by researchers, who they never heard from again and that the research made no perceivable impact on their lives.  The subjects who participated in the previously discussed UCSB experiments knew that their knowledge was cherished by the researchers and was used to frame continuing research.   Many of them made suggestions after the experiment and their comments on open-ended questions led to further research based on problems and barriers they had acknowledged.   Some subjects even refused payment for their participation, saying that the funds should go to further research and implementation.   None of the subjects quit during any experiment, even though the tasks could become quite long, or scheduling problems resulted in time allocation that was much longer than anticipated.   They all seemed determined to add to the body of knowledge about this topic that directly affects their daily lives.   This research was not some strictly academic laboratory experiment that would not directly affect them, but research into their needs to gain more information, spatial knowledge, accessibility, freedom, independence, social equity and to improve their overall quality of life.

 

BACK TO OVERVIEW BACK TO TABLE OF CONTENTS NEXT SECTION