提升公共運輸使用率之遊戲化激勵機制設計與評估

Abstract

我國交通運輸目前面臨公共運輸使用率難以進一步提升以及尖峰供給不足的兩大問題。為解決這些問題，以往的策略是透過票價補貼、轉乘優惠、新闢路線、興建場站、加開班次等方式增加使用率以及提高供給，這些策略相當昂貴且效果不顯著。參考國外成功的遊戲化實驗案例，本研究旨在檢驗各種遊戲化激勵機制對運具轉移和尖峰轉移之效果。 為此，本研究邀請50名參與者進行為期三週之遊戲化實驗，以評估遊戲化激勵機制之效果。每週設計不同的遊戲化激勵機制，透過社交媒體鼓勵參與者之間的交流和競爭，進而搭乘大眾運輸和離峰出發來賺取積分。為了調查遊戲化激勵機制對他們看法和意圖的影響，在每個實驗前與實驗後進行問卷調查。在實驗中，由於「選擇公共運輸」以及「選擇離峰出發」兩種選擇行為分為「不選擇」、「低機率選擇」、「中機率選擇」、「高機率選擇」四種程度，本研究運用雙變量次序普羅比（Bivariate Ordered Probit, BOP）同時模化運具選擇以及出發時間選擇行為。 估計結果顯示，這兩種選擇行為不相關，即兩種選擇行為被視為獨立行為，並可以用不同的解釋變數單獨估計。影響大眾運輸選擇變數中，性別、同行人數為重要變數，而影響離峰出發選擇中以職業、收入及同行人數為顯著變數。有趣的是，沒有金錢獎勵的遊戲化機制也被發現有效。此外，在運具轉移和尖峰轉移中，介入機制對時間轉移較為有效。

Difficulties in further increasing the usage rate and insufficient supply at peak hours are two major problems of Taiwan public transportation (PT) systems nowadays. To solve these problems, government has already implemented various strategies to maximize PT usage rate and to relieve PT supply shortage, including fare subsidy, transfer discount, new route opening, new station construction and denser frequency. These strategies are very costly with and insignificant effects. With successful gamification trials in promoting usage of public transportation in many cities, this study aims to examine the effects of various designed gamified incentives to shift mode and departure time decisions towards public transportation and off-peak, respectively. To do so, this study designs a three-week gamification experiment with 50 invited participants to evaluate the effect on gamification incentives. Each of weeks exercises a specific incentive scheme to encourage communication and competition among participants through social media and to earn their credits by using PT and departing at off-peak hours. To investigate the effect of gamification incentives on their perception and intention, questionnaire surveys were conducted before and after each of weeks. Since two credited choices: to use public transportation and to depart at off-peak are responded at four degrees: no, low, medium, and high probability, this paper uses of bivariate ordered probit (BOP) to simultaneously model mode choice and departure time choice behaviors. The estimation results show that these two choice behaviors are not correlated, that is, each of two choice behaviors would be identified as two independent behaviors and can be estimated separately with different explanatory variables. For the choice of PT, gender and number of accompanies and are identified as key factors. As to the decision of off-peak departure time, occupation, income, commuting distance and number of accompanies are significantly tested. Interestingly, the gamification scheme without monetary reward is also found effective. Additionally, gamification incentives are found to be more effective for departure time shift than mode shift.