Measuring and Forecasting Demand - Tourism: Principles, Practices, Philosophies - page 295

Cruise travel is forecasted to continue to be the fastest-growing sector of the travel industry. This forecast
is based on survey research, past trends, demographics, customer satisfaction, marketing efforts, and
new ships joining the eet. PhotobyAndyNewman;courtesyofCarnivalCruiseLines.

exchange rates, and distance or journey time. These relationships are usually identi ed using multiple
regression, as already discussed.
Computer simulation

models include a complex set of equations that will usually combine both
the trend-line extrapolation methods and the regression-technique models into a more comprehen-
sive systems simulation. Relationships between many variables are speci ed through interrelated
equations. Simulation models rely on historical data for input and model calibration. Once a model
gives reasonably accurate distributions for past years, it can be used to predict probable future
distributions.
Simulation models require specially trained personnel with a high degree of technical expertise to
set up original model and data processing programs. Knowledge of time-series, cross-sectional, and
causal relationships and change processes is required. Also, powerful computing resources and high
data precision are necessary. These are serious problems that have to be faced by any tourism
organization that might consider using this approach. Simulation forecasting is best suited for a
problem that is complex with known and quanti able relationships and some feedback effects. It is
also suitable for long forecast horizons.

Executive Judgment (Delphi) Method

Mathematical and statistical models are most useful and often produce accurate results. However, the
combined experience of tourism executives is also valuable. The Delphi method, in essence, consists of
a systematic survey of such experts. A series of questions is asked, and then the results, as a consensus,
are reached. An example of the executive judgment method is the United Nations World Tourism
Organization (UNWTO) Panel of Tourism Experts, where 250-plus experts contribute information on
tourism trends.
Mathematical statistical tools cannot incorporate the in uences of variables not explicitly
included in the model. For example, under multiple regression, income and travel prices were the
only two variables used to predict demand. However, other factors, such as the political situation,
fuel situation, changes in taste, amounts of leisure, and the effectiveness of promotion campaigns,