Socio-Econ. Plan. Sk. Vol. 11, pp. 109410. Pergamon Press 1977. Printed in Great Britain zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK NOTE IN DEFENCE OF ATTITUDINAL RESEARCH: A TRANSPORT EXAMPLE D. A. HENSHER School of Economics and Financial Studies, Macquarie University, North Ryde, Sydney, N.S.W. 2113, Australia (Received 2 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC February 1976; revised 9 September 1976) This note discusses the use of indices of satisfaction levels in the context of changes in the magnitude of trip characteristics associated with an individual’s usual transport mode for the journey to work. This is partly designed to gain a general impression on the consistency of perceptual response to hypothetical changes in modal attributes, along the attitudinal dimension. Car and train users were asked to indicate how satisfied they would be with their usual method of travel if the overall trip time (and cost) were to change in the direction of an increase and a decrease. The central tendency results are summarised in Table 1. Five time variations and four cost variations were considered ranging from the marginal time increase and then decrease, to the 10 min time reduction or 50% cost increase. The cost change for the car journey was described in terms of a cost per mile, with a mean perceived car mileage cost of 4 cents, based on previous evidence, used as the base value. The train cost relates to the total trip fare. All the mean levels of satisfaction are consistent with the directional expectation. The car user is, at the base mean, more satisfied with his trip time and cost than is the train user. When door-to-door travel time increases by 1 min there is an expected small change in level of satisfaction with the largest difference in the 1 min decrease for train users (an increase in satisfaction by 8.4%).The 1.06% variation between the base level of satisfaction for the total sample and the level of satisfaction associated with the 1 min increase (together with the 3.2% variation for the 1 min Y&at is S, now equals 63.8, from ((67.1)(128) t (60)(106))/234 and S2 now equals 42.5, from ((37.1)(128)+(49.1)(106))/234. Hence, S, - S, now equals 21.3compared to 25 (= 71-46) before. Total sample size = 234, No. of car users (as main usual mode) = 128, No. of train travellers (as main usual mode) = 106.60 and 37.1are respectively, levels of satisfaction associated with car and train as alternative main mode. 67.1 and 49.1 are obtained from Table 1. decrease) illustrates the general perceptual non-response to such a small change in an attribute. The difference is more likely due to the inability of a person to distinguish 1 unit gradations on a satisfaction scale. When a 5 min change is assumed, the level of satisfaction changes by 18.6% for the train user (increase and decrease). For the reduction by 10 min, there is a 34.6% change in the level of train user satisfaction compared with the 7.6%for the car user. Hence, it seems plausible to suggest that train users, who are generally persons using the relatively lower perceived satisfaction mode, are more satisfied with any positive or negative time adjustment than are car users. Hence it also appears that the greater the level of present usage satisfaction, the lesser the percentage variation in satisfaction associated with a change in a trip attribute. This implication could bring into further question the appropriateness of the abstract mode approach. The results of the cost changes are more difficultto assess as a modal comparison because of the different costing procedures. If a straight comparison was made, the same results as for time adjustments are found with respect to percentage variations in the level of satisfaction. The percentage changes for the car user for the two cost increases of 1 and 2 cents per mile and the two cost decreases of 1 and 2 cents per mile are respectively 16.6%, 24.7%, 14.3% and 23.8%.The comparative train user cost results for the two cost increases of 25%and 50%of the present fare and the two cost decreases of 25%and 50%of the present fare are respectively 41%. 68%, 2% and 53.6%. The usefulness of this information can be illustrated by the effect of the 5 min time increase for both car and train users. The mean value of I,“(S,, - SZI)is 725 utility difference units before the 5 min increase. The 5 min increase has an effect on the level of satisfaction associated with the individual’spresent mode, which reduces the mean value of I,‘& - S,,) to 565.5. That is, there is a change in S,, due to a change in car user satisfaction (1 = car) and a change in SZ, due to a change in train user satisfaction (2 = train).t It is assumed that there are no cross-mode Table 1. Base and adjusted levels of satisfaction associated with the modal characteristics, time and cost (SD. = standard deviation) 125% (50% Mean Mean S.D. S.D. _ _ -I 74.8 89.1 (16.3) (14.3)