One advantage of agent-based models is the ability to incorporate a variety of data sources. TourSim is no exception! I’ve listed many of the datasets used in creating and verifying this model.
1. Tourist Accommodation, Activity, and Distance Preferences
These preferences are based on data from the Statistics Canada Canadian Travel Survey and International Travel Survey (1998-2004). These surveys give the type of accommodation that respondents stayed in, the types of activities they participated in during their trip, and how far they traveled on their trip. In the model, tourists choose from this range of accommodation and activity in a manner that reflects the distribution found in these surveys. For example, if in the Canadian Travel Survey 20% of tourists were found to stay in a hotel on their trip, in TourSim, approximately 20% of the tourist agents will choose to stay in a hotel.
Distance preferences work in a similar manner, but are used as a maximum distance that a tourist will be willing to travel. For example, if the distance between Sydney and Digby is further than the tourist distance preference, this trip will not be made, even if Digby has the preferred accommodation and activity. Instead, the tourist will look for another destination that is closer and has the preferred accommodation and activity characteristics.
2. Total Tourists Per Day
The number of tourists entering the model, this value is drawn from the Key Tourism Indicators data provided by the Nova Scotia Department of Tourism, Culture, and Heritage. This value changes monthly, with peak numbers of tourists entering Nova Scotia during the summer months. These values are different for Domestic, International, and American tourists, with Domestic tourists representing approximately 84%, American tourists 13%, and International tourists 3% of the total tourists visiting Nova Scotia. This breakdown in based averages from 2000-2006, from the Visitation by Origin report by the Nova Scotia Department of Tourism, Culture, and Heritage.
3. Tourist Port of Entry
The specific number and location of tourists entering Nova Scotia is taken from the Visitation to Nova Scotia by entry point reports prepared by the Nova Scotia Department of Tourism, Culture, and Heritage.
4. Tourism Economic Impact
This value is a combination of daily expenditure values from the Canadian Travel Survey/International Travel Survey, and Economic Impact by Region reports from the Nova Scotia Department of Tourism, Culture, and Heritage.
5. Destination Supply
The range of accommodation and activity supply options were gathered from the official Nova Scotia Tourism online database at www.novascotia.com. The model currently uses 35 different destinations, and while this does not include all locations in the province, it provides a representative look at many of the key tourism locations within Nova Scotia. At this point the total capacity of a location is not taken into account.
6. Tourist Length of Stay
This measures how many days a particular tourist stays in the model. This is based on information from the Nova Scotia Department of Tourism, Culture, and Heritage 2004 Visitor Exit Survey (.pdf).
7. Destination Visitation
The number of tourists visiting each destination is in many ways a factor of the range of accommodation and activity choices present at that location. Other influencing factors include the distance from origin to destination, and “awareness”. Awareness is used in the model to represent the fact that tourists often do not have perfect knowledge of the entire province. For example, major tourist destinations such as Halifax and Baddeck likely have more top-of-mind presence than smaller, emerging destinations, such as Canso. In TourSim, each time a tourist looks for a new destination, at a random rate, they may not consider less popular destinations as a way to represent this process. This awareness factor was used to calibrate (or tune) the model so that visitation to each town was approximately the same as what was seen in the 2004 Visitor Exit Survey.