Investigating the Feasibility of Personalized Rapid Transit: an Experimental Approach

Citation:

Richard J. Lutz (1972) ,"Investigating the Feasibility of Personalized Rapid Transit: an Experimental Approach", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 800-806.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 800-806

INVESTIGATING THE FEASIBILITY OF PERSONALIZED RAPID TRANSIT: AN EXPERIMENTAL APPROACH

Richard J. Lutz, University of Illinois

One of the most serious challenges of the 1970's is that of cleaning up the air in major U. S. cities. At the present time, the more than 87-million automobiles on the road are considered to be the cities' worst single source of air pollutants.

For years the primary alternative to the automobile in most U. S. cities has been antiquated and highly inefficient public transportation. However, new technology now makes it possible to automate public transportation, thus greatly reducing operating costs. A revamping of public transportation modes will do much to help clean up American cities.

One of the more exciting new developments in public transportation is the concept of "personalized rapid transit," which provides an alternative to the automobile for short-haul service within major metropolitan areas. Personalized rapid transit (PRT) calls for a network of computer-controlled cars, each of which carries 20 or fewer passengers. During non-peak load times, a passenger would be able to use the PRT vehicle to reach a specific destination directly-the vehicle would not follow a pre-designated route. During rush hours, the PRT system could be adapted to operate like a regular mass transit system, with trains of vehicles following specific schedules.

One of the problems which has faced designers developing PRT vehicles has been the size of the vehicle. While a small vehicle would be desirable, designers fear the reaction of passengers to being enclosed, perhaps with strangers, in a vehicle of the dimensions of the automobile.

This vehicle size problem is especially evident in the design of a PRT system for an existing tunnel network in a large Midwestern city. Shortly after the turn of the century, a tunnel network was completed which was used for freight delivery to major buildings in the city's downtown district. The company that operated the system went bankrupt in 1959, and since that time the city has been searching for a new use for the tunnels.

One recent proposal which is being considered by the city is to develop the tunnel network into a PRT system. As one segment of the feasibility studies being conducted regarding this proposal, the Research and Design Department of the University of Illinois developed a mock-up of the proposed PRT vehicle.

A unique problem which was encountered in the design of the vehicle resulted from the dimensions of the tunnels for which it was being designed. With a height of only 7'6", the tunnel would allow a vehicle of no more than 5' interior height. This would mean that passengers in a moving vehicle would be Unable to stand. Feeling that this feature might be a strong deterrent to potential passengers, it was decided to consider an alternative vehicle which would allow passengers to stand if they desired. If this alternative vehicle were found to offer a significant advantage over the smaller vehicle, the tunnels would have to be enlarged by some two feet in height, at a cost of several million dollars. Thus, an important decision in the design of the PRT vehicle are passengers' reactions to its size. The primary purpose of this investigation, therefore, was to determine the effects of vehicle size on passenger attitudes toward a mock-up of the proposed PRT vehicle.

A secondary concern was the load factor--or how the presence of other passengers affects passenger reactions to the vehicle. Recent research on "crowding" (Little, 1965; Griff and Veitch, 1971; Stokols, et al., 1972) has shown that people tend to react negatively toward situations in which their "personal space" is limited. It was felt that a passenger in a fully loaded PRT vehicle might attach a lower evaluation to the vehicle than would a single passenger.

THEORETICAL ORIENTATION

A general class of attitude models--labeled "expectancy-value" models has seen increasing use in marketing in the past few years. These types of models are most often direct applications or hybrids of models developed by Fishbein (1963) and Rosenberg (1956), although some have been developed within marketing (Cohen and Houston, 1970; Sheth, 1970). Essentially, these models are all based on the premise that a person's attitude toward any object, action, or brand can be represented by his cognitions or beliefs about the attitude object. While some issues remain unresolved as to the appropriate measures and combination rules implied by these models, the most general form of the model can be represented algebraically:

"_o = EB_ia_i

where

"_o = attitude (affect) toward the object, action, or brand under investigation

B_i = the belief that the object possesses some attribute i or will help reach some goal i

a_i = attitude toward or satisfaction with the it attribute or goal.

The expectancy-value model was chosen as the theoretical orientation which guided the investigation of passenger reactions to the proposed PRT vehicle. In addition to offering insights into the relative merits of the two alternative vehicles, it was hoped that such a test would provide more evidence regarding the usefulness of the expectancy-value model as a decision-making tool. It was found that passengers did develop a cognitive structure regarding the PRT vehicle after a short exposure to it, then there would be more justification for believing that consumers really do form attitudes on the basis of their beliefs about products and brands.

Four hypotheses were tested in this study:

1. There is a positive relationship between attitude and cognitive structure as a result of direct experience with the PRT vehicle.

2. There is a significant interaction between vehicle size and the load factor in the PRT vehicle. Specifically, the large vehicle is rated relatively higher than the small vehicle in the single passenger condition; and due to the crowding factor, the small vehicle is rated relatively higher than the large vehicle in the fully loaded condition.

3. Differences in attitude toward the PRT vehicle across experimental conditions are reflected in corresponding differences in the index of cognitive structure (EBiai).

4. Differences in attitude across experimental conditions are reflected in differences in single Bi and ai components of cognitive structure.

SUBJECTS

Subjects were 65 undergraduate students enrolled in introductory marketing, management, and psychology courses at the University of Illinois. Since some of the experimental conditions involved running subjects in small groups, effort was taken to prevent groups of friends from comprising an experimental group. It was felt that being in the vehicle with a close friend would tend to bias the Subject's responses. All Subjects were volunteers who agreed to participate after hearing a short description of the project. At no time was the study referred to as an experiment rather, it was called a "marketing research study."

APPARATUS

The main piece of apparatus used in this study was the mock-up of the proposed PRT vehicle. This mock-up had previously been constructed by Mr. Paul Eshelman, who assisted with this study. The vehicle was adjustable to two sizes, which comprised one factor under investigation here. The "small" vehicle was 5' high and 7' long with four seats facing each other. It could accommodate no standing passengers, so the fully loaded condition consisted of four seated passengers. The "large" vehicle was 7' high and 9' long, again with four seats. Since there was now room for standing passengers, the fully loaded condition consisted of four seated and four standing passengers. Thus, when fully loaded, the large vehicle was actually more "crowded" than the small vehicle since twice as many people were in the only slightly larger vehicle. While the mock-up was stationary, it was windowless; this latter feature made it possible to create a sensation of movement during the simulated "trip" in the vehicle. A small electric motor was placed against the side of the vehicle, producing sound and slight vibrations. This measure was not intended to be a deception, as it was obvious to Subjects that the mock-up was immobile, but pretests had shown that the use of the motor did tend to add a degree of realism to the "ride."

EXPERIMENTAL SITUATION

The Subject(s) entered a large room and sat down at a long table. After introducing himself, the Experimenter explained the procedure which would be followed in the study. This consisted of three parts:

1. A 2-minute slide presentation providing information about PRT systems in general.

2. A 2-minute "ride" in the PRT vehicle.

3. Completing a short questionnaire.

The Subjects then saw the slide presentation, which was accompanied by a tape-recorded message.

Immediately following the slide presentation the Subject was read some additional information regarding the tunnel network for which the PRT vehicle was being designed. Following the reading of the tunnel information, the Subject was led into another room which contained the mock-up of the PRT vehicle. He was instructed to take a seat in the vehicle and not to talk (if in a group condition) with the other "passengers." Using a stopwatch to time the "trip," the Subject was left in the vehicle for exactly two minutes.

Upon completion of the trip, the doors of the vehicle were opened and the Subject was ushered back into the main room. At this point, he completed a questionnaire which tapped his cognitive structure regarding the PRT vehicle as well as obtaining several measures of attitude toward the vehicle and the slide presentation. After the Subject had completed the questionnaire, the Experimenter explained the nature of the study and cautioned the Subject not to reveal the purpose of the study to his classmates who had not yet participated. After answering any further questions, the Experimenter thanked the Subject for his cooperation and terminated the experiment.

ANALYSIS AND RESULTS

Subjects in the small vehicle--single passenger and large vehicle--single passenger conditions were all run individually. Subjects in the small vehicle-fully loaded condition were run in groups of four, and Subjects in the large vehicle--fully loaded condition were run in groups of eight. Due to Subjects not showing up when scheduled, two graduate students served as "stooges" to ensure a fully loaded vehicle. Since the smaller vehicle allowed no standing passengers, the standing--seated factor could not be investigated across the two sizes of vehicles. The only meaningful analysis involving the standing passengers in the large vehicle--fully loaded condition would be a t-test of the difference between the means of that group and the seated group in that same condition.

Attitude toward the PRT vehicle was assessed in three ways:

1. A semantic differential consisting of 15 scales. A principal components analysis yielded three scales which loaded highly on the evaluative factor (good-bad; ugly-beautiful; pleasant-unpleasant). The scores on these three scales were summed to form one measure of Ao.

2. A simple 1-10 rating scale of affect toward the vehicle, ranging from "like not at all" to "like as much as possible."

3. The derived EBiai index. Thirteen attributes of the vehicle were measured which had been identified as salient in open-end pretests. All B were measured via Sheth-type bipolar scales ("evaluative beliefs"); e.g., "dimly lit"-"brightly lit." All a were measured on 7-point bipolar scales ranging from "good" to "bad;" e.g., "The lighting in the vehicle was..."

The correlation between EB a and the semantic differential measure was .70. The EBiai--rating scale correlation was .53. Thus it appears that the cognitive structure index did provide a good measure of attitude toward the PRT vehicle. The semantic differential--rating scale correlation was .67, so all three attitude measures were strongly interrelated.

In testing the load- size interaction hypothesis, a 2x2 factorial ANOVA was conducted on each of the attitude measures. All three showed a significant main effect for the size factor, but did not show a significant interaction effect, thus invalidating the second hypothesis. Table 1 summarizes the ANOVA results.

TABLE 1

MEANS

An unexpected finding was that the smaller vehicle was rated higher than the larger vehicle. Calculation of Hays' (1963) omega-square statistic shows the amount of variance explained by the size factor for each of the attitude measures. From observation of Table 1, it is clear that the cognitive index measures of attitude (EBiai) does behave similarly to the other attitude measures, thereby supporting the third hypothesis. Additionally, the ANOVA using EBiai as the dependent variable was able to account for more of the variance in passenger reactions to the vehicle than either of the other attitude measures, thus lending further support to the use of the expectancy-value approach in the assessment of consumer attitudes.

Of interest are the differences in specific elements of cognitive structure which lead to the differences in overall attitude toward the PRT vehicle. Of the 13 Bi and 13 ai, a total of eleven significant (beyond a-.05) main effects were obtained from 2x2 factorial ANOVA. Table 2 summarizes these findings. All the main effects for vehicle size were in the same direction as the effects for overall attitude--i.e., the small vehicle. While it is beyond the scope of this report to speculate as to the reasons underlying this pattern of results, it is interesting to note that ai components were significant in three cases where there was no significant effect for the corresponding Bi components. While this may be a non-predicted result, it is not too surprising. Cohen and Houston (1971) have demonstrated that Bi components can differ, depending only upon usage of a particular brand. There is no reason to suspect that ai components would not behave similarly in certain situations. If a valued object is seen as possessing certain desirable qualities, the consumer may actually re-evaluate upward his liking for or satisfaction with those qualities. However, the results presented here only suggest such a process. Additional experiments are required to test that hypothesis.

TABLE 2

CONCLUSION

The results of this study are promising, from both a theoretical and a practical point of view. The expectancy-value model proved to be quite useful in determining reactions to the PRT vehicle, and the potential seems great for further applications of the model in the area of public policy decisions.

Secondly, this study illustrates the age-old claim that "there is nothing as practical as a good theory." Not only did this study provide useful information to designers working on PRT vehicles, but it also contributed to the body of literature surrounding theoretical properties of the expectancy-value model

Obviously, there are limitations involved in this study--the results in no way indicate how potential passengers might respond to the total PRT system. Only reactions to the design of the vehicle were considered. The sample was small and non-random, thus weakening the generalizability of the results. From a theoretical standpoint, this study can best be viewed as exploratory in terms of the hypotheses which can be generated from the results.

The primary contribution of the study was to lend further support to the usefulness of the expectancy value attitude model: it showed that a cognitive structure is developed rather quickly through direct experience with an attitude object. Previous studies have tended to deal with long-established attitudes or attitudes newly formed on the basis of information.

REFERENCES

Cohen, J. & Houston, M. Some Alternatives to a Five-Point Likert Scale (Especially if You Have a Purpose in Mind). Paper presented at Attitude Workshop, University of Illinois, December, 1970.

Cohen, J. & Houston, M. Cognitive Consequences of Brand Loyalty. Journal of Marketing Research, 1972, 9, 97-99.

Fishbein, M. An Investigation of the Relationships Between Beliefs About an Object and the Attitude toward that Object. Human Relations, 1963, 16, 233-240.

Griffit, W. & Veitch, R. Hot and Crowded: Influences of Population Density and Temperature on Interpersonal Affective Behavior. Journal of Personality and Social Psychology, 1971, 17, 92-98.

Hays, W. Statistics for Psychologists. New York: Holt, 1963.

Little, K. Personal Space. Journal of Experimental Social Psychology, 1965, 1, 237-247.

Rosenberg, M. Cognitive Structure and Attitudinal Affect. Journal of Abnormal Social Psychology, 1956, 53, 367-372.

Sheth, J. An Investigation of Relationships Among Evaluative Beliefs, Affect, Behavioral Intention and Behavior. Unpublished working paper, University of Illinois, 1970.

Stokils, D.,Rall, M., Pinner, B., & Schopler, J. Physical, Social, and Personal Determinants of the Perception of Crowding. Unpublished working paper, University of North Carolina, 1972.

----------------------------------------