4402.0.55.001 - Microdata: Childhood Education and Care, Australia, June 2014 
Quality Declaration
Quality Declaration ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 06/07/2015
Page tools:
 Print Page
 Print All
| ||
|
USING TABLEBUILDER CONTINUOUS DATA ITEMS The TableBuilder file contains a number of continuous data items that are available for selection from 'Summation Options' in the 'Customise Table' pane. Continuous data items are those data items which can have a response value at any point along a continuum. Continuous data items include cost of care and number of hours care used. Some continuous data items are available as categorical data items with values grouped into categories. For example below, Cost of care is available as ‘Cost of care in a usual week after the Child Care Benefit and Child Care Rebate’ and ‘Cost of care in a usual week after the Child Care Benefit and Child Care Rebate - categories’ with cost categorised into $25 ranges. Some continuous data items are allocated special codes for certain responses (e.g. 9999 = 'Not applicable'). When creating ranges in TableBuilder for such continuous items, special codes will automatically be excluded. Therefore the total will show only 'valid responses' rather than all responses (including special codes). For example: The following shows the tabulation of the data item 'Age child first commenced long day care (in months)'. The continuous values of the data item are contained in the 'A valid response was recorded' row. 'Could not be determined' includes children for which the age they commenced long day care is not know. 'Not applicable' includes children not currently attending long day care. To show the actual continuous values in a table, a range must be created. Here is the same table with a range applied for the continuous values for the data item 'Age child first commenced long day care (in months)' (Age first commence LDC). Note that the numbers of children for the 'Could not be determined' and 'Not applicable' categories no longer contribute to the table. Any special codes for continuous data items are listed in the Data Item List. USING REPEATING DATASETS The Income Unit and Child levels are counting units, whereas the Income Unit Care and Child Care levels are repeating datasets. The repeating datasets in the CEaCS are a set of data with a counting unit which may be repeated for a child or an income unit. The 'one to many' relationships, described in File Structure section, between the Income Unit level and the Income Unit Care level, and the Child level and the Child Care level, shows the connection between counting units and repeating datasets, i.e. an event or episode is repeated so that multiple records with the same set of data exist for the same child (or income unit). For example, a child may have used more than one instance of child care such as (i) a long day care centre, (ii) family day care and (iii) grandparents. Consequently, three records would be present on the Child Care level for this child, representing a repeating dataset, with each record containing information for a common set of data items, e.g. Number of days of care used, Number of hours of care used, cost of the care and so on. Also, the child will have summary records in addition to the individual care records, described below. In this example, although the three records all relate to a single child, any totals from the Child Care level are a count of child care arrangements. Summary Records and Data Items In addition to the general or base records present in the repeating datasets (i.e. on the Income Unit Care and Child Care levels) that, for example, provide details about each instance of child care, there are also 'summary' records that provide aggregate information for selected groupings of the types of care. For example, summary records are available for groupings of All formal care, All informal care, All care and preschool and All care excluding preschool. In the example of a child who attended long day care, family day care and also received care from a grandparent, there are three base records on the Child Care level because they attended three separate instances of child care. For each record, the data item for the cost for the type of care was reported as $38, $10 and $5 respectively. Therefore, the summary record for this child for the total cost of formal care (i.e. long day care and family day care) is recorded as $48 ($38 + $10). Similarly, the summary record for this child for the total cost of all care (i.e. all three types of care) is recorded as $53. The following data items comprise the classifications that enable the data for these summary records to be tabulated: Income Unit Care level - Type of care used by the family. Child Care level - All types of care. COUNTING UNITS AND WEIGHTS The Summation Options section in the Customise Table panel contains the counting units/weights that are available. It is critical that the correct weight (or summation option) is used when specifying tables. The names of the weights (summation options) for each level are highlighted below. The default summation option in this TableBuilder is the Income Unit level weight (Weights). The default summation option will be automatically added to a table when the table is being specified - so care needs to be taken that this is the correct weight required for the particular tabulation. If the default weight is not the required weight then select the correct weight from the Summation Options list, because what is counted in a table depends on the weight or counting unit selected. It is critical to use the correct weight (or summation option) for the data items you include in the table. Using the incorrect one will produce incorrect estimates. The weight corresponding to the lowest level data item in a table should be used. For example, if a Child care level data item is cross-tabulated with a Child level data item, the Child care level weight should be used. Similarly, if an Income unit care level data item is cross-tabulated with an Income unit level data item, the Income unit care level weight should be used. In general, the Child weight is used if child estimates are required and the Income Unit weight is used if estimates of income units are required. Child Care level weights should be used when producing tables with data items from the Child Care Level. Income Unit Care level weights should be used when producing tables with data items from the Income Unit Care level. USING FLAG ITEMS To enable easier table specification and to ensure that the correct populations, and hence the correct data, are being tabulated, a number of 'flags' have been included in the TableBuilder that should be used at all times when extracting data. Usual or Last week To differentiate between child care used in the 'last week prior to the survey' and care 'usually' used, data items (or flags) are available to restrict the population in a table for this purpose. These flags are on the Income Unit Care and Child Care levels. It is imperative that these usual or last week care flags are used when any data items from the Child Care level or the Income Unit Care level are used, regardless of whether the care level data items are used alone or with other Income Unit or Child level data items. If these flags are not used for Child Care or Income Unit Care data items, the data will be incorrect. The flags and their categories are: Income Unit Care level
1. At least one child in the family used care usually
1. At least one child in the family used care last week Child Care level
1. Care used usually
1. Care used last week Labour force scope flag In households where all adults were out on scope or coverage of the LFS, no information was obtained for the 2014 CEaCS. However, as long as at least one parent in the household was in scope for the LFS, information about children aged 0–12 years and some information about their parents were collected and included in the 2014 CEaCS. There is a labour force scope flag (Labour force flag - indicating whether parent did not respond to LFS) to indicate whether the income unit is out on scope. This flag (present on the Income Unit level) indicates if one parent in a family was out on scope or coverage. Limited employment and demographic data are available for these families. Information about the working arrangements used by parent/guardians to help care for their child was not available for parent/guardians who were out on scope or coverage of the labour force for any reason. CROSS–TABULATING DATA ITEMS FROM DIFFERENT LEVELS General Information Cross-tabulating data items from different levels will produce different results depending on the weights that are used. It is therefore important to understand how TableBuilder works and what is being counted. Estimates will always be produced in TableBuilder - but care must be taken to ensure that they are logical and correct. In general, when cross-tabulating data items from different levels using the weight from the lower level, the results are the sum of the counting units or weights at the lower level. For example, if cross-tabulating 'State or territory of usual residence' from the Income Unit Level (i.e. the higher level) by 'Sex of selected child' from the Child Level (i.e. the lower level) while using the lower (Child) level weight, then the results will be the sum of children - as this is the counting unit on the Child Level. In simple terms, TableBuilder effectively copies the data from the higher level to each record on the lower level and then sums the relevant records at the lower level. In this example, the table will produce estimates of the number of male children and female children by State/Territory. When cross-tabulating data items from different levels using the weight from the higher level things are more complicated and care needs to be taken. Under this scenario, fields at the lower level effectively become multi-response fields at the higher level and, in addition, each category in a data item is tabulated as 'one or more occurrences with the particular characteristic(s)'. For example, if there are two male children in the same Income Unit then the results are tabulated, and should be interpreted, as an estimate of the 'number of Income Units with one or more male children' - in this case the Income Unit is only counted once. If there is one male and one female child in the same Income Unit then the Income Unit is counted twice (i.e. multi-response) with the male child included in the estimate of the 'number of Income Units with one or more male children' and the female child from the same Income Unit included in the estimate of the 'number of Income Units with one or more female children'. It is important to note that the totals in these types of tables indicate the actual estimate of Income Units (i.e. each Income Unit counted only once) while the component cells, which contain the multi-response concept, will be greater than the total. It is therefore critical that the construction of a table, when cross-tabulating data items from different levels using the weight from the higher level, is conceptually well considered. For some data items the data tabulated may not be particularly useful or logical. For this reason the CEaCS TableBuilder contains particular data items that enable easier and more logical specification of cross-tabulations. These data items comprise filtering or index data items, flags and population data items. The main filtering or index data items are 'Type of care used by the family' on the Income Unit Care Level and 'All types of care' on the Child Care Level. Also, on both the Income Unit Care and the Child Care Levels, there are flags that indicate whether care was 'usually' used or used 'in the last week' prior to the survey. And finally, some population data items (such as 'Children aged 0–8 years') are included on the Child Level. These data items should be used in almost all tables as they restrict or filter particular populations and determine the correct types of care that have been used. More importantly, they eliminate many of the complexities when cross-tabulating data items from different levels. The following examples show how these data items are used. Example 1: Cross–tabulating Child level data items by Income Unit level data items using different weights. The table below uses the data items 'State or territory of usual residence' from the Income Unit level and 'Whether child attends school' from the Child level. The Child level weight is used as it is the lowest level. The result is a count of the number of children attending school by state. Using the same table as above, but replacing the Child level weight with the Income Unit level weight, the result is a count of the number of Income Units (i.e. Families) with at least one child attending school. If a family has a child attending school and a child aged less than 4 years this family will be counted twice. Therefore, 'Whether child attends school' is treated as a multi-response data item when the Income Unit weight is used. Example 2: Cross–tabulating Child level data items by Child Care level data items using Child Care level weight. Most cross tabulations done within the CEaCS publication involved a data item called 'All types of care' found on the Child Care level. Understanding how this data item works is instrumental as it can result in miscounts if used incorrectly. Below are the output categories for this data item:
01. Before and/or after school care 02. Long day care centre 03. Family day care 04. Occasional care centre 06. Grandparent 07. Brother/sister 08. Non-resident parent 09. Other relative 10. Other person 21. All care and preschool 22. All formal care 23. All informal care 24. Preschool 25. All care excluding preschool 31. Used formal, used informal, used preschool 32. Used formal, used informal, no preschool 33. Used formal, no informal, used preschool 34. Used formal, no informal, no preschool 35. No formal, used informal, used preschool 36. No formal, used informal, no preschool 37. No formal, no informal, used preschool 41. Used formal care only (disregarding preschool) 42. Used informal care only (disregarding preschool) 43. Used both formal and informal care (disregarding preschool) 52. Family day care and Occasional care 61. Brother/sister and Other relative 62. Brother/sister, non resident parent and other relative 63. Brother/sister, non resident parent, other relative and other person Categories 21, 22, 23 and 25 must be used when calculating totals for formal care, informal care and all care. This is because a child must only appear once when calculating totals. You cannot sum formal care types or informal care types as the total will not represent the sum value for each child, but the number of instances of each type of care, resulting in double counting. This is particularly important when cross-tabulating the 'all types of care' with another variable which may differ between different types of care for the same child (e.g. cost of care or days attended). For example, consider creating a table cross-tabulating the type of care by cost of child care last week in $25 ranges. If a child attended family day care at a cost of $50 and occasional care at a cost of $25, that child would appear once in the 'All formal care' row with a value of $75. This is the result that will be produced by using the category 21 to get a total of 'all formal care'. If categories 01, 02, 03 and 04 were summed together to create a total for 'formal care', this child would appear twice in the 'total formal care' row, once with a value of $25 and once at $50. They would not appear with their true value of $75 for cost for care. Similarly, if a child attended long day care 2 days last week and occasional care 1 day, the child would appear once in the 'All formal care' row with a value of 3 days. Again, if categories 01, 02, 03 and 04 were summed together to create a total for 'formal care', this child would appear twice in the 'total formal care' row, once with a value of 1 day and once with 2 days. This does not reflect the actual value, of 3 days, they attended formal care last week. The table below is the number of children by the number of days care used last week by formal care types. It shows that the sum of the formal care types is not equal to the 'All formal care' category. For '2 days', the sum of the components is 264.3, whereas the total of 'All formal care' is 261.3. Categories 31 to 37 must also be used when calculating totals. For example, to determine how many children attend both family day care, grandparent care but don't attend preschool, use the 'Used formal, used informal, no preschool' category which is category 32. This is because if you group together these items (categories 03 and 06) to generate a total, then a child will appear twice in this total, resulting in double counting. Using category number 32 only counts the child once. Three new categories have been added in 2014, these are '41 - Used formal care only (disregarding preschool)', '42 - Used informal care only (disregarding preschool)' and '43 - Used both formal and informal care (disregarding preschool)'. These differ from 34, 36 and 32' respectively. Category 41 combines 33 and 34 but ignores any preschool cost and hours attended. Similarly, category 42 combines 35 and 36, and category 43 combines 31 and 32. Categories 41, 42 and 43 should be used where preschool attendance is not of importance and is to be removed from the denominator of tables. The 'All types of care' data item, includes responses for both attendance last week and usually. To generate frequencies for care usually used by the child, select the data item called 'Flag to indicate whether care used usually' and select 'Care used usually' and use this as a filter to restrict the table population. This same logic applies for tables generating frequencies for care arrangements used 'last week'. Example 3: Cross–tabulating Income Unit level data items by Child level data items by Child Care level data items using Child Care level weight. The table below is a count of the number of instances of care for children aged 0–8 years who usually use care by state. It uses the data items, 'State or territory of usual residence' from the Income Unit level, 'Children aged 0–8 years' from the Child level and 'All type of care' and 'Care used usually' from the Child Care level. The weight used is the Child Care level weight as the lowest level used is Child Care level. In this example, as it is a count of the instances of care, 'Care used usually' or 'Care used last week' must be used otherwise some double counting will result. A total can't be calculated for 'All types of care' as this data item is treated as a multi-response data item. Totals are available in the summary records for the data item 'All types of care' (i.e. All care and preschool, All formal care, All informal care and All care excluding preschool). Example 4: Cross–tabulating Income Unit level data items by Income Unit Care level data items using Income Unit Care weight. The table below is a count of the number of instances of care for families with at least one child usually using care by state/territory. It uses the data items, 'State or territory of usual residence' from the Income Unit level and 'Type of care used by family' and 'At least one child in the family used care usually' from the Income Unit Care level. The weight used is the Income Unit Care level weight as the lowest level used is Income Unit Care level. In this example, as it is a count of the instances of care, 'At least one child in the family used care usually' or 'At least one child in the family used care last week' must be used otherwise some double counting will result. ADJUSTMENT OF CELL VALUES To minimise the risk of identifying individuals in aggregate statistics, a technique is used to randomly adjust cell values. This technique is called perturbation. Perturbation involves small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of identifiable statistics while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics. After perturbation, a given published cell value will be consistent across all tables. However, adding up cell values to derive a total will not necessarily give the same result as published totals. The introduction of perturbation in publications ensures that these statistics are consistent with statistics released via services such as TableBuilder. ZERO VALUE CELLS Tables generated from sample surveys will sometimes contain cells with zero values because no respondents that satisfy 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. Whilst the tables may include cells with zero values, the ABS does not publish such zero estimates in Childhood Education and Care, Australia, June 2014 (cat. no. 4402.0) and recommends that TableBuilder clients do not use these data either. Document Selection These documents will be presented in a new window.
|
|