4235.0.55.001 - Microdata: Learning and Work, Australia, 2010-11 Quality Declaration 
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 28/05/2013  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All

Header picture USING TABLEBUILDER


For general information relating to TableBuilder and instructions on how to use features of the TableBuilder product, please refer to the User Manual: TableBuilder, 2012, (cat. no. 1406.0.55.005).

More specific information applicable to the Learning and Work Survey TableBuilder, which should enable users to understand, interpret and tabulate the data, is outlined below.

COUNTING UNITS AND WEIGHTS

Weighting is the process of adjusting results from a sample survey to infer results for the total population. To do this, a 'weight' is allocated to each person. The weight is the value that indicates how many population units are represented by the sample unit.

Selecting a weight to use in a table can be done through 'Summation options'. There are two options when using the Learning and Work Survey TableBuilder file: 'Persons' and 'Qualifications'. It is critical the correct weight is selected when specifying tables to ensure the data generated is appropriate. The following image shows the available Summation Options:

Picture: screen shot of weights available on the file


A weight in bold, such as in the image above, indicates the weight being used in the table. The default weight when producing a table using the Learning and Work Survey TableBuilder is the Persons weight (in bold in the image above). This weight is usually automatically applied to any table being generated.

If you are estimating the number of persons with certain characteristics (e.g. 'Number of non–school qualifications completed') the weight listed under the category heading 'Socio–demographic and Learning and Work Data Items' must be used. If this weight is not in bold:
1. Click on the blue triangle 'twistie' () next to the 'Summation Options' line
2. Click on the 'Socio–demographic and Learning and Work Data Items' 'twistie'
3. Click on the 'Person Weights' 'twistie'
4. Click on the 'Sum' tick box
5. Add the person weights to your table by clicking on 'add to row' or 'add to column'
6. Ensure the 'Sum' tick box under 'Qualifications' is blank

To estimate the number of qualifications (e.g. the number of non–school qualifications completed in 2010 or later) the weight listed under 'Qualification Level Weights' must be used. The same process as above can be followed, ticking the 'Sum' tick box under the Qualification Level Weights 'twistie' instead.

Qualification level data items are weighted according to the characteristics of the person who undertook the qualification, and therefore the weights for each qualification are the same as the weight for the person. For example, if a person in the sample has a weight of 600 and that person has completed three non–school qualifications then the person represents 600 people in the total population and 1,800 qualifications.


SELECTING DATA ITEMS FOR CROSS–TABULATION

It is critical to understand the tabulation results obtained when cross–tabulating data items from the different levels (see File Structure) and when using different summation options. The Learning and Work Survey file has two levels each with several sub levels:

Picture: levels and groups available on the file

The Socio-demographic and Learning and Work Level contains a range of data items detailing the characteristics of the respondent including some education variables. The Qualifications Level contains data items about each of the qualifications that a respondent has obtained. The file is hierarchical with each respondent record potentially having multiple qualification records.

Cross-tabulating Socio-demographic and Learning and Work Level X Socio-demographic and Learning and Work Level data items

Cross-tabulating data from the Socio-demographic and Learning and Work Level with other data items from the same level will produce data about people. For example, cross-tabulating the geographic variable 'State or territory of usual residence' by the 'Level of most recent non-school qualification' produces a table showing the number of people in each region by the most recent qualification they have obtained (see below).

Table: an example table showing number of persons in state or territory of usual residence by level of most recent non-school qualification


*Person weight is the default weight selection; however, it is essential to check that that the Person weight is selected from the 'Summation Options' for cross-tabulations where all variables are from the Socio-demographic and Learning and Work Level.


Cross-tabulating Qualification Level X Qualification Level data items

Cross-tabulating data from the Qualification Level with other data items from the same level will produce data about qualifications when using the Qualification Weight. For example, cross-tabulating 'Level of non-school qualification' by 'Whether completed qualification' in Australia produces a table showing the number of qualifications completed in Australia. If a respondent has several qualifications, each of those qualifications is included in the table (see below). By using the Qualification weight in the 'Summation Options', each of a respondent's qualifications is included in the table.

Table: an example table showing level of non-school qualifications by whether completed qualification in Australia. The table uses the qualification weight.


If the same cross-tabulation of 'Level of non-school qualification' by 'Whether completed qualification in Australia' is generated but using the Person weight instead of the Qualification weight, the following table is produced:

Table: an example table shows level of non-school qualification by whether completed qualification in Australia. The table uses the person weight.


When using the Person weight, a respondent with several qualifications may have some qualifications excluded from the table. This occurs because the same combinations of responses can only be counted once in a table when the Person weight is applied. To illustrate, a person has the following five qualifications (each qualification appears as a separate record on the file):

QualificationWhether completed qualification in Australia
1Bachelor degreeCompleted in Australia
2Certificate IIICompleted in Australia
3Bachelor degreeCompleted in Australia
4Post graduate degreeDid not complete in Australia
5Bachelor degreeDid not complete in Australia

In this example, only Qualifications 1, 2, 4 and 5 will be counted in the tabulation. Qualification 3 will be excluded because the qualification and whether it has been completed in Australia is the same as Qualification 1. All the other combinations of qualifications and 'Whether completed qualification in Australia' are unique.

The need for data where the same combinations of responses are only counted once is likely to be limited so as a general rule, Qualification weight should be selected from the 'Summation Options' for cross-tabulations where all variables are from the Qualification Level.

Using Qualification Flags

To assist with analysis, several variables have been created to help isolate specific qualifications. The following shows the available Qualification Flags:

Picture: screen shot of qualification flags available on the file

By using a Qualification Flag, only one qualification for each respondent is included in a table. Selecting either the Person weight or the Qualification weight when using a Qualification Flag will produce the same result.

Cross-tabulating Socio-demographic and Learning and Work Level X Qualification Level data items

Cross-tabulating data items from the Socio-demographic and Learning and Work Level with data items from the Qualification Level can produce data about people or qualifications depending on the weight being used. Caution should be used when Cross-tabulating a Qualification Level data item while using a Person weight as a person with multiple qualifications may have the same qualifications counted only once in a table (for more detail see above: Cross-tabulating Qualification Level X Qualification Level data items).

Using a Qualification Flag may be worthwhile when cross-tabulating Socio-demographic and Learning and Work Level with Qualification Level data items as only one selected qualification will be included in the tabulation.

Example 1: Cross–tabulating qualification level data items by person level data items using the person weight – When using a qualification flag

When using a Qualification Flag (e.g. 'Second most recent qualification') in a table that cross–tabulates a qualification level data item by a person level data item, either the Person or the Qualification weight can be used and the same output will be generated. Restricting the table to a single qualification for each person (in this example the second most recent qualification) in effect turns this into a person level data item, as TableBuilder only needs to read one row of data from the qualification level for each person. The following table shows the results from such a tabulation where 'level of non–school qualification' is a data item from the qualification level and 'sex' is a data item from the person level.

Table: an example table showing level of non-school qualificaiton by most recent qualification flag by sex. The table uses qualification weight.


Example 2: Cross–tabulating qualification level data items by person level data items using the person weight – When NOT using a qualification flag

When a Qualification Flag is not used, TableBuilder will read each row of data from the qualification level for each person. In this case, TableBuilder effectively calculates the tabulation as a 'multi–response' table (i.e. the same person can be counted more than once), but it counts the same categories of information about different qualifications only once. It treats them as 'one or more occurrences' of that category. For example, if a respondent completed three qualifications, and for two of these qualifications the main field of study was 'highly relevant' to their current job where as the third qualification was 'not at all relevant' to their current job, then the person would be counted once for 'highly relevant' and once for 'not at all relevant'.

Therefore, in these particular types of tabulations, components of the table will not add to the total number of persons (as persons can be counted more than once), but the total will be the correct count of persons as TableBuilder calculates the total in such a way that each person is only counted once. An example table is shown below:

Table: an example table showing whether main field of study is relevant to current job if not working in field by sex.

In summary, qualification level data items can be cross–tabulated with person level data items with or without Qualification Flags. Qualification Flags should be included in tables when a user wants information only about one particular qualification (e.g. the highest qualification or the most recent qualification), but should not be used in tables looking at all qualifications.


TABLE POPULATIONS

Table populations or units of measure can be found by looking at the 'Counting' subheading (see example below).

Table: The table shows weekly personal income from all sources - deciles by number of non-school qualifications completed.


ADJUSTMENT OF CELL VALUES

The TableBuilder dataset has random adjustment of cell values applied to avoid the release of identifiable data. All cells in a table are adjusted to prevent any identifiable data being exposed. For this dataset 'additivity' has not been applied, that is, when the interior cells are randomly adjusted they have not been set to add up to the totals. As a result, randomly adjusted individual cells will be consistent across tables, but the totals in any table will not be the sum of the individual cell values.

ZERO VALUE CELLS

Tables generated from sample surveys will sometimes contain cells with zero values because no respondents that satisfied the parameters of the cell were in the survey. This is despite there being people in the population with those characteristics. That is, the cell may have had a value above zero if all persons in scope of the survey had been enumerated. This is an example of sampling variability which occurs with all sample surveys. Relative Standard Errors cannot be generated for zero cells.

MULTI–RESPONSE DATA ITEMS

A number of the survey's data items allow respondents to report more than one response. These are referred to as 'multi–response data items'. An example of such a data item is pictured below. For this data item respondents can report all of their sources of personal income.

Picture: Screen shot of all sources of personal income data item

When a multi–response data item is tabulated, a person is counted against each response they have provided (e.g. a person who responds 'employee income' and 'unincorporated income' and 'government pensions and allowances' will be counted once in each of these three categories).

As a result, each person in the appropriate population is counted at least once, and some persons are counted multiple times. Therefore, the total for a multi–response data item will be less than or equal to the sum of its components. Multi–response data items can be identified by the initials 'MR' in the data item list, which can be accessed from the Downloads page. In the example below, the sum of the components is 25,256,200, where as the total population is 17,735,300.

Table: the table shows all sources of personal income by persons.

NOT APPLICABLE CATEGORIES

Most data items include a 'not applicable' category. The 'not applicable' category comprises those respondents who were not asked a particular question(s) and hence are not applicable to the population to which the data item refers. The classification value of the 'not applicable' category, where relevant, is shown in the data item list (see the Data Item List in the Downloads tab).