[If not already delivered] DSS acknowledges the traditional owners
of country throughout Australia, and their continuing connection to
land, sea and community. We pay our respects to them and their
cultures, and to elders both past and present.
Hi everyone In this session I will be telling you about a new way
to access the PIA research dataset - through the ABS Tablebuilder
tool. I'll provide a short background on the data that is
available, then show you just a few ways you can use the product
using some examples. The slides will be available afterwards for
your reference, so please don't feel that you need to remember
everything!
We think the Tablebuilder tool will be very useful to a wide number
of researchers and I hope that this presentation will encourage you
to make use of this data yourselves.
1
2
4
Users are expected to include academics and NGO’s looking for data
to use in their grant applications as well as teachers and students
in an educational context, and journalists preparing data for
reports. 1. We will be looking at how to create and download both
tables and graphs in this
session today.
5
1. Fewer Variables: There are 38 variables available in
TableBuilder but there are 60 variables available in the full PIA
data available in SURE
2. Less Detail: For example, date of birth is only shown by year in
TableBuilder whereas it is shown by month in SURE.
3. Fewer People: The 5% sample is representative of the full
dataset. Importantly, if you are reporting population estimates in
your grant application, you will need to multiply the figures in
the table by 20.
6
We will walk through how to use TableBuilder once you have access.
To apply for access to PIA data in TableBuilder, simply register
through the ABS TableBuilder website: www.abs.gov.au/registration
TableBuilder holds over 50 datasets other than PIA. There is an
open version of the Census data that you are able to go in and play
with as a ‘Guest’ without needing to register. This might be good
to do if you would like to get a feel for what you can do in
TableBuilder before getting access to the PIA dataset.
Once you have been granted access and have logged in, this is the
first screen you will see. 1. The datasets you have access to will
be listed here in the left column and this will
depend on the datasets that your organisation has been granted
access to. For this reason it is important to register using your
organisation email, not a personal email account like bigpond or
gmail. If you don’t see The PIA dataset (as shown here), simply
send in an email asking for access to this particular dataset.
Email:
[email protected] (found in the TableBuilder User
Guide)
2. The middle column will display the predefined tables and any
tables you save over time.
3. The right hand column provides background information to
TableBuilder and links to the User Guides. Please refer to these
for more information about how to use the features available in
TableBuilder. When you select a dataset on the left, the
information on the right will then reflect the background
information of that dataset.
4. You can also access the TableBuilder User Guide by this Help
button at the top right.
5. You can come back to these pages anytime by selecting the
Datasets tab.
8
Let’s look at how to use TableBuilder to answer a few questions.
First, let’s look at those people who are receiving Carer
Payment.
9
By expanding the Benefit Type variable you can see all benefit
types available. 1. Selecting the particular benefit you are
interested in 2. allows you to use the buttons across the top. We
will add Carer Payment as a
Filter. Using the filter means that only those records (or people)
who are on Carer Payment will be added to the table.
3. At all times when using TableBuilder you are able to select
either the ‘i’ icon at the right of the title or the ‘About this
data’ link at the bottom of the main screen. These will both take
you to the ABS Microdata page where you can read background
information about the PIA data in TableBuilder.
4. The number at the end of each variable describes the number of
categories in that variable. For example, you can see there are 21
categories under Benefit Type. These are the 21 benefits that are
listed in this dataset.
5. Every time you make a change to your table, you will be prompted
to select ‘Retrieve Data’. When you do this, the results will
populate in the table. You can continue building your table until
you are ready to retrieve the data.
10
As mentioned, the PIA data contains 56 separate quarters of
information across 14 years. 1. These quarters are listed in the
‘Benefit Quarter End Date’ variable. Because the
data contains 56 separate quarters, it is important to incorporate
‘Benefit Quarter End Date’ into the table somewhere. If you don’t,
the results displayed in the table will be a sum of the records for
all quarters across the 14 years. That is, many of these will be
duplicates as most people have received payment across more than
one quarter. Depending on your research question, you can use
‘Benefit Quarter End Date’ in the table in different ways: by Row,
Column, Filter or Wafer.
2. Adding any variable to the wafer will turn a 2D table into a 3D
cube. 3. By adding the quarters to the Wafer, you will only see one
quarter at a time, but
can easily switch between them using the drop down box. 4. In this
case, you can see that there were 9,405 recipients in the quarter
ending
June 2015.
11
To find out “How many Carers are in each age group?”, we can add
all the age groups to the table. 1. To do this, we can click, hold
and drag Date of Birth into the Row to have the
years listed vertically.
12
1. By selecting ‘Retrieve Data’ we see how many people who were
receiving Carer Payments in June 2015 were born in each year of our
data.
13
1. To see a graphic representation of this data you simply select
the Graph View from the tabs across the top.
Please note, I have removed the years that did not contain data
before displaying this graph.
14
1. In Graph View you can also switch between the quarters using the
Wafers drop box.
15
Let’s see how the age of Carer Payment recipients has changed
across time.
16
To switch back to change the table, you simply select ‘Table View’
from the tabs at the top. 1. To compare across dates, I have
removed the Benefit Quarter End Date from the
Wafer. 2. I can select just a few of the available quarters by
selecting the quarters I am
interested in and then 3. using the buttons at the top to add these
quarters into the Columns of the table. 4. This way I have a table
that presents the number of recipients by birth year
across these four quarters
17
1. By selecting Graph View we can see what the distribution of age
groups are for each of the four time periods.
In this display it is difficult to see the change in shape across
the four time periods because they are displayed on top of each
other.
18
1. You can hide parts of the graph by clicking on the categories at
the bottom
19
1. Alternatively you can display the results by Column rather than
Row. Remember, in our table we had the four time periods in the
Column and so now this is how the data is grouped in the
table.
2. Now you can see not only the growth in numbers but also the
change in shape across the time periods. In particular you can see
that there are more young carers in the 2014 time period than there
were in the 2001 time period.
3. That is the shape of the 2001 distribution is leaning to the
left (more older aged carers) whereas the shape of the 2014
distribution has a slight lean to the right (there are more younger
aged carers).
20
Let’s see how the age of the carers might be different across
geographic regions.
21
We might look at age groups by State. Because these are larger
regions they will have larger numbers and therefore less chance of
being supressed. Where you select very specific information the
table can be suppressed as one of the privacy measures built into
TableBuilder. That is, where there are low numbers and the
possibility of people being identified, the information in the
table is suppressed and zero values are returned.
1. In this case I have selected only NSW, Victoria and Queensland.
2. I will use the Column button at the top to assign them to the
columns in the
table.
22
1. Please note that the order the categories are displayed is
determined by the order that they are ticked.
2. In this case I selected Victoria before I selected Queensland 3.
And as such, Victoria is listed before Queensland in the
table.
23
1. If you want to be able to come back to this table in a different
session, you will need to Save the Table.
2. Also as a side note, if you design a table that has too many
cells, it will be converted to a ‘Large Table’. There is
information about this in the TableBuilder User Guide but put
simply this means that your table will be created in the background
and available for download, but you won’t be able to view the table
in TableBuilder.
24
1. You will see your saved tables in both the Datasets tab and in
the Saved Tables tab.
2. In the Saved Tables tab you can Open, Copy, Delete and Rename
your table.
25
Now let’s look at a different question. Say we would like to know
information about people who are 50 years or older and receiving
Newstart Allowance. What do we know about these recipients
regarding their level of education and gender? How is this
different for indigenous and non-indigenous recipients?
26
1. To create this table I have created a new category called “DOB
for Over 50” using Custom Data. Please see TableBuilder User Guide
for details on how to do this.
2. I added this new category as a filter along with Newstart
Allowance. I have all 56 quarters in the wafer because I want to
see how this changes over time.
3. I added those education categories I was interested in, in an
order that makes sense: from Less than Year 10 to Master’s
Degree.
4. I then added both Gender and Indigenous Status to column to see
Gender split by Indigenous Status.
5. When you have a table you are happy with, you can download the
results for later use. To do this, select the type of file format
you prefer and
6. then select ‘Go’.
27
1. When there are multiple levels or categories in the wafer of
your table (such as we have now with all 56 quarters), the excel
worksheet you download will contain the data for each wafer on a
separate tab.
2. You can identify the data that each tab refers to by the
category identified in the cell above the table.
3. You can also see the Filters that have been applied
28
1. The Graph View of this table shows the number of Newstart
recipients aged over 50 in the June 2015 quarter
2. for each education level 3. for indigenous status by
gender
29
1. You can present the data in a number of ways. This is a good
feature to play with. 2. When you have a graph you like, you can
download using the ‘Go’ button at the
top right.
30
A lot of this functionality will be things you can discover with
use. Please make use of the ABS TableBuilder User Guide and the PIA
background information available on the TableBuilder site.
At this stage, are there any questions?
31
32