|Page tools: Print Page Print All|
APPENDIX - BREAK IN TIME SERIES
Between the March and June quarters 2003, the ABS undertook rigorous coverage checking for caravan parks to improve the Survey of Tourist Accommodation (STA) frame. This coverage checking resulted in 74 additional caravan parks being identified for the first time in the June quarter 2003. While these establishments were identified in this quarter, they actually commenced operations prior to this time period.
These additional 74 caravan parks were added to the STA frame in the June quarter 2003. The addition of these caravan parks added $2.9 million (1.7%) to the total takings from accommodation for the June quarter, 158,557 site nights occupied (1.4%) and 5,267 sites, cabins, flats, units and villas available (2.2%).
The addition of the 74 caravan parks resulted in a break in time series between the March and June quarters 2003. The impact of the break in time series is explained below.
THE IMPACT OF THE BREAK
The extent of the impact varies by geographic region. The table below shows the number of establishments which were added by state/territory. As illustrated in the table below, the major additions occurred in New South Wales, Victoria and Queensland. No additional caravan parks were added to the Australian Capital Territory.
The effect of the break in series on key data items is shown in Table 2.
Table 2 shows that New South Wales had $0.9 million added to total takings for caravan parks. This was 1.5% of the total takings for New South Wales. Overall, the additional 74 caravan parks added $2.9 million (1.7%) to total caravan park takings for Australia.
Table 4 shows how the 74 caravan parks which were added to June quarter 2003 were distributed at the Tourism Region level. No Tourism Region had more than four caravan parks added.
Where releasable, Tables 5 and 6 present factors which can be used to take into account the break in series. As most Tourism Regions had less than three caravan parks added, the majority of factors are suppressed due to confidentiality. All SLAs have had less than three caravan parks added, therefore factors at this level are confidential.
SITE OCCUPANCY RATE
The caravan park site occupancy rate is calculated by the ABS after data has been collected from providers. This data item is derived by dividing one collected item into another (site nights occupied divided by site nights available).
Both the numerator and denominator used to calculate this item have been affected by the additional 74 caravan parks to a similar extent. Therefore the addition of the 74 caravan parks has a minimal effect on the calculation of the site occupancy rate. Thus a break in series is not likely to have occurred for the caravan park site occupancy rate.
The addition of the 74 caravan parks increased the level of the series between the March and June quarters 2003. When calculating percentage changes between the March and June quarters 2003 care needs to be taken to ensure the effect of the break in series is taken into account.
The factors do not need to be applied to caravan park data for 2000. As the number of caravan parks has fluctuated between 2000 and 2003, these data are not strictly comparable. Care should be taken when calculating percentage changes between 2000 and 2003.
Tables 5 and 6 contain factors for use when calculating percentage changes for caravan parks between the March and June quarters 2003. These factors were calculated by dividing June quarter 2003 data including the 74 caravan parks, by the June quarter 2003 data excluding the additional caravan parks. These factors adjust the level of the series for the March quarter 2003 to take account of the missing 74 caravan parks.
The method for calculating the percentage change is shown by the example below. Takings from accommodation data can be found in Table 2 of this publication. Factors can be found in Table 6 of the Appendix. These figures are summarised in Table 3.
Multiply the March quarter 2003 figure by the appropriate factor to obtain an adjusted March quarter 2003 figure
$189,000 x 1.01705 = $192,222
Calculate the percentage change in the usual manner except in this case use the adjusted March quarter 2003 figure
($175,947/$192,222 - 1 ) x 100 = -8.5%
This same methodology applies for calculations of percentage changes for all affected caravan park data items between the March and June quarters 2003.
SHORT-TERM AND LONG-TERM CARAVAN PARKS
Factors for short-term and long-term caravan parks have not been included because the vast majority are confidential. Users should contact The Transport and Tourism Business Statistics Centre for further information.
The additional 74 caravan parks have been added into each month of the June quarter 2003 (ie April, May, June). Therefore the break in series has occurred between the months of March and April 2003. When calculating percentage changes, the appropriate quarterly factor should be used to adjust monthly data in respect of the break in series.
Users should contact the Transport and Tourism Business Statistics Centre on 07 3222 6185 if they have any queries on the break in series.
The tables in the Excel Spreadsheet contain information for use when calculating percentage changes:
These documents will be presented in a new window.