Discreet Choice Experiment on car sales in Norway

The data is from an online survey administered to a representative population of former and prospective car buyers in Norway. This dataset contains the discrete choice experiment (DCE) data collected as part of the survey, stored in NLogit format.

Codebook:

Option - 1 (car one); 2 (car two); 3 (neither)
Choice - respondent choice
Treated - 0 (non-monetary framing); 1 (monetary framing)
Safety - % of max Euro NCAP rating
Energy - litres per 10km
Capacity - litres of boot capacity
Cost - car price in NOK
ID - respondent identifier

This research set out to examine the role that monetary running cost information can play in terms of highlighting the fuel efficiency of new vehicles. Specifically, this study involved the distribution of a split sample (control/treatment) discrete choice experiment to a representative sample of the Norwegian car buying population, via an online survey undertaken in late 2017. This survey was distributed to over 1000 individuals representing a cross section of the Norwegian population in all regions of the country.

Prior to the distribution of the survey, a series of focus groups identified safety rating and luggage space as the most important attributes to include in the experiment, in addition to the research parameters of interest: purchase price and energy efficiency. Attribute levels were selected to reflect those currently present in the Norwegian automobile market, see Table 1. A fractional factorial design, utilising the JMP software package, generated 32 unique choice pairs. To prevent respondent fatigue, these pairs were split across four survey blocks, so that each respondent faced only eight choices. These eight choices were presented in either the control or treatment format, with each respondent only receiving choices in a single format to avoid any framing contamination effects. Therefore, there were eight versions of the survey in total, four control and four treatment blocks.

In the control version of the experiment, the attributes were displayed in a simplified version of how they are currently displayed on new cars in Norway. In the treatment version, the energy consumption variable was augmented with a monthly running cost estimate, displayed in terms of Norwegian Kroner (NOK). Both the treatment and control images also contained a graphic with the vehicle’s environmental rating (A-G), as mandated under current EU and Norwegian legislation. The rating is based on CO2-emssions, which is proportional to fuel consumption when fuel type is constant. In this study, all vehicles considered used gasoline.

The findings from our analysis of the data suggest that with the addition of running cost estimates, individuals’ WTP for more efficient vehicles can be significantly increased, in the case of this research by up to 28%.

本数据集源自一项在线调查，该调查针对挪威的既往及潜在汽车购买者群体进行。数据集包含作为调查一部分收集的离散选择实验（DCE）数据，并以NLogit格式存储。

### 数据代码簿：

- 选项 - 1（汽车一）；2（汽车二）；3（两者皆不选择）
- 选择 - 受访者的选择
- 处理 - 0（非货币化框架）；1（货币化框架）
- 安全性 - 最大欧元NCAP评分的百分比
- 能源 - 每10公里的油耗
- 容量 - 车尾箱容积（升）
- 成本 - 汽车价格（挪威克朗）
- ID - 受访者标识符

本研究旨在探讨货币化运营成本信息在凸显新车燃油效率方面的作用。具体而言，本研究涉及向挪威汽车购买者群体的代表性样本发放（控制/处理）离散选择实验的分割样本，该调查于2017年末在线进行。该调查发放给超过1000名代表挪威各地区人口的个体。在调查发放之前，一系列焦点小组确定了安全评级和行李空间为实验中应包含的最重要属性，此外还包括了研究关注的参数：购买价格和能源效率。属性水平的选择旨在反映挪威汽车市场中现有的水平，详见表1。使用JMP软件包生成的分数因子设计生成了32个独特的选择对。为了防止受访者疲劳，这些对被分为四个调查块，这样每个受访者只需面对八个选择。这八个选择以控制或处理格式呈现，每个受访者只接受一种格式的选择，以避免任何框架污染效应。因此，总共有八个调查版本，四个控制版和四个处理版。

在实验的控制版本中，属性以当前在挪威新车上显示的简化版本呈现。在处理版本中，能源消耗变量被增加了一个月运营成本估算，以挪威克朗（NOK）的形式显示。处理和控制图像还包含了一个图形，显示了车辆的环保评级（A-G），这是根据当前欧盟和挪威法律规定的。评级基于二氧化碳排放量，当燃料类型恒定时，与燃料消耗成正比。在本研究中，所有考虑的车辆均使用汽油。

我们分析数据的结果表明，随着运营成本估算的加入，个人对更高效车辆的支付意愿（WTP）可以显著增加，在本研究案例中，可增加高达28%。