Martin Gillies, Australian Electoral Commission (AEC)
|
Martin brought along a simple gender distribution by Commonwealth Electoral Division at the Close of Rolls for the 2007 Federal election. He has samples of the results table from the AEC submission to the Joint Standing Committee on Electoral Matters, a SuperWEB recreation of the source table and a graph indicating how elector numbers grew week by week during 2007. |
Source data
|
|
Sample SuperWEB table
|
Sample SuperWEB table as a report
|
|
Sample SuperWEB chart |
Phil Casburn, Australian Taxation Office (ATO)
| Phil presented a couple of tables from the latest edition of the Taxation Statistics 2006-07 published on the ATO web site. |
Sample SuperCROSS table
|
Sue Stockbridge, Catholic Education Office (CEO)
|
Mapping school grades against national testing – a simple table uncovers many questions. |
Sample SuperWEB table
|
Rod Battye, Tourism Research Australia (TRA)
Rod brought along some tables that illustrate how they used SuperCROSS to figure out why an increase in the number of international backpackers was not translating into a commensurate increase in occupancy rates at backpacker hostels.
|
Tourism Research Australia We had feedback from the backpacker association that, although the number of international visitors were increasing, they were not seeing this in their occupancy rates, the number of nights spent in backpacker accommodation was not growing accordingly. To make matters worse, they had seen an increase in the number of backpacker establishments in anticipation of the growth. We started looking through our international figures in SuperCROSS, within minutes it was clear backpackers were simply moving away from the standard types of backpacker/hostel style accommodation establishments. Upon further investigation we found backpackers were staying in groups of 5 or 6, staying in 4 or 5 star resort style apartments (higher standard of accommodation), with access to swimming pools, spas, tennis courts, etc., rather than go it alone in conventional backpacker establishments. |
Sample SuperCROSS tables
The backpacker industry is now better equipped to manage size and number of establishments in that part of the industry, so fewer business go bust. Everyone has a better understanding of what is happening in that part of the market. |
Stephen MacDermott, Australian Bureau of Statistics (ABS)
Stephen presented an example graph and table taken from the ABS’ 2006 Time Use Survey. It displays the proportion of the population engaged in various activities over the course of a 24 hour period. From a SuperSTAR perspective it demonstrates the impact of database design on ease of analysis.
|
The ABS submission uses data from the 2006 Time Use Survey. This survey ran over a 10 month period to account for seasonal variations in time use. Respondents were asked to complete time use diaries on 2 selected days in which they recorded information about each episode of time use such as: primary activity; secondary activity; start time; end time; and location. For more information, see 4153.0 – How Australians Use Their Time, 2006. The collected data was processed and stored in a SuperMART. With correct database design SuperCROSS can easily produce a wide variety of tables including: average time; ranged time; and time distribution. The table submitted by the ABS for the ExciTable competition at SSUG Canberra 2009 is a time distribution table. It shows the percentage of people aged 15 or over sleeping, eating, working, and watching TV at 15 minute intervals throughout the day. |
Sample SuperCROSS table ABS Time Use Survey 2006 |
|
Some of the patterns are not surprising - you can see that most people (over 50%) are in bed by 11pm and up by 7am. Others are more interesting - the data for eating show distinct peaks corresponding to the 3 main meals but the peak for breakfast is much smaller than the other two. Some SSUG participants took this to mean that people were skipping breakfast but this is not necessarily the case. The area under the curve represents the total time spent by all people on the given activity so a smaller area can mean fewer people but it can also mean less time per person. So the smaller peak for breakfast could simply mean that breakfast is a shorter meal - additional tabulation could easily resolve this ambiguity. The curve for work reflects a standard 9 to 5 working day with flex time and a dip for lunch. It tops out at just over 30% but this does not necessarily mean that only 30% of people work - rather that only 30% are working at any one point in time with others on leave or off shift. Finally the TV curve shows a steady rise over the course of the day with a peak at around 9:30. The "peakiness" of the distribution is surprising - most people expected more of a plateau. One possibility is that this could reflect a shortcoming of the collection methodology - there is nothing to prevent respondents from under-reporting low-status activities, replacing them with high-status activities they would normally perform if not for some accident of fate. |
Sample SuperCROSS chart ABS Time Use Survey 2006
|
