Geography Reference
In-Depth Information
Fig. 10 Data completeness
and accuracy by survey
means
Figure
10
shows that the self-administered diary approach can achieve better
completeness for recording temporal events of tourist movement; in addition,
given the time to record the travel events on the day, the diary method captures
more accurate spatial-temporal information of the itineraries. The differences are
not that great to make either approach redundant, and indeed analysis has shown
that the two sets of data provide a similar set of estimates in terms of tabulations of
data at key tourist sites (Forer and Simmons
2011
).
This section systematically assessed the spatiotemporal data quality of WCTFS.
The quality issues were diagnosed in terms of gaps from uncompleted questions,
and errors from records that were not consistent with plausible behavior. Three
types of gaps were examined in the dataset: missing arrival time, missing duration
and possible missing prior stop. The one type of error used is derived from
inconsistency between reported arrival times and modeled expectations.
A number of findings have been determined through this assessment. First of
all, although the case study data shows a certain level of data confidence in terms
of the overall spatiotemporal data accuracy, there are as expected missing temporal
events in the movement survey data. There are also a significant number of
reported arrival times in the dataset that have notable levels of discrepancy from
modeled estimates of arrival time. These are likely to reflect poor time keeping or
overlooking a prior stop.
Preliminary exploration of incompleteness and inconsistency has demonstrated
a number of latent patterns in the dataset. Several parameters have been used to
identify the possible impacts of specific factors such as age, gender, activity at stop
and time of capture. These provide evidence of particular factors at work, gen-
erally significant but in all cases confounded by each other. A useful finding is the
recorded data accuracy of spatiotemporal events is palpably influenced by the
activities tourists carried out at the stops.
The locations and functions of places on the West Coast attract tourists taking
different activities and staying for different durations. Consequently, the accuracy
level of captured spatiotemporal data varies from one stop to another. The main
