Use the Define Recode window to create and modify recodes:
Element
Description
Recode Name
The name of the recode.
Leave this unchanged to edit the field or recode you opened.
Change it to create a new recode field.
Field Values
The list of available values for this field.
If you have a large list of values, right-click and select Find to search for a value.
Recode Values
The list of values currently include in the recode. Only values that appear in this list will display in the table.
Use Names
Use Codes
Use Both
Use these options to specify whether to display values using their names, codes, or a combination of both.
Changing this setting will not change any values already added to the Recode Values list. You can therefore use this option to include a combination of names, codes and combinations of both in your recode if necessary.
>> Singly >>
Click to add the selected Field Values to the recode as individual items.
>> Group >>
Click to add the selected Field Values to the recode as a group.
Make sure you set the Group Name first before adding the group.
>> Subitems >>
For hierarchical fields, click to add
<< Remove <<
Click to remove the selected Recode Values from the recode.
Group Name
When adding a group to the recode, set this to the group name and then click >> Group >>
Total Name
When creating a total for the recode, select the values you want to include in the total, set this to the total name, and then click Add Total.
Select all
Click to select all the Recode Values. You can use this when creating totals and when you need to remove all the values from the recode.
Add Total
Click to add a total to the recode based on the selected Recode Values.
Footnote
Enter text here to display as a footnote whenever the recode is used in a table.
Define Recode Window - Summation Options
When you are recoding the summation options, the Define Recode window has some additional options that only apply to summation options:
Element
Description
Weight By
You can only use weights from the same record type level as each selected measure. The default weight displayed is loaded during installation, these function can be configured.
Select one of the following:
Weighting
Description
Unweighted Sum
Lets you add unweighted measures with no functions.
Replicate Weights
Lets you add functions such as Jack-Knife. If a replicate weight is selected and a non replicate function is added for a measure, the first weight in the replicate weight collection is used in the function.
Main Weight
Functions
If Unweighted Sum is selected, you can select from the following functions. Click >> Add >> to add the function to the currently selected value in the Recode Values list.
Function
Description
Sum
The total of all the values.
Mean
The total of all the values divided by the number of values.
Standard Deviation
How spread out a distribution is, defined as the square root of the variance.
Variance
How spread out a distribution is. This function is closely related to the Standard Deviation. The variance is defined as the average squared deviation of each number from its mean.
First Quartile
Cuts off the lowest 25% of data.
Last Quartile
The last quartile cuts off highest 25%.
First Decile (10%)
Cuts off the lowest 10%.
Second Decile (20%)
Cuts off the lowest 20%.
Third Decile (30%)
Cuts off the lowest 30%.
Fourth Decile (40%)
Cuts off the lowest 40%.
Median (50%)
Cuts the data set in half (the median is the middle of a distribution).
Sixth Decile (60%)
Cuts off the lowest 60%.
Seventh Decile (70%)
Cuts off the lowest 70%.
Eighth Decile (80%)
Cuts off the lowest 80%.
Last Decile (90%)
Cuts off the lowest 90%.
Gini
A measure of inequality of a distribution, defined as a ratio with values between 0 and 1. The numerator is the area between the Lorenz curve of the distribution and the uniform (perfect) distribution line; the denominator is the area under the uniform distribution line.
Largest (3)
Maximum and next 3 largest values from contributing unit records.
Smallest (3)
Minimum and next 3 smallest values from contributing unit records.
Count Distinct
A count of the number of distinct results.
If Replicate Weights or Main Weight is selected, you can select from the following functions.
Function
Description
Weighted Sum
Rate
Covariance
The extent to which two random variances vary together (co-vary) can be measured by their covariance.
Unweighted Sum
Confidentiality - At Level
The record type to apply the confidentiality (concealment) rule to.
The levels available are defined in the database hierarchy. When the application calculates the number of contributors for a cell it has to know the level to count on. Therefore the At Level option is only for concealment rules, not rounding rules. The level should be as high or higher than the highest level of the measure in the Recode Values list or an error displays when the tabulation process is run.
You can also set the level to Default, which means that each measure is summed at its own level. For example, for a database containing two tables, Person and Household:
With the level set to Default: contributors to, or percentages for, a measure from the Person table will be counted at the Person level and contributors to or percentages for a measure from the Household table will be counted at the Household level. If there are two people in one household, using a measure from the Person level will count two contributors, but using a measure from the Household level would count only one.
With the level set to Household: contributors to or percentages for a measure from either table will be counted at the Household level. If there are two people in one household, and you are using a measure from the Person table, using the Person level will count two contributors, but using the Household level would count only one.
Confidentiality - Rule
A rounding rule (confidentiality or concealment rule) for the selected measure in the Recode Values list. The options are as follows:
Rule
Description
Effect of Applying Rule
None
No rule is applied to the selected field (when None is selected, no footer displays).
None
0-3 Rounding
Randomly changes any values of 1 and 2 in the table to either 0 or 3.
0 is not changed
1 is changed to 0 with probability 2/3
1 is changed to 3 with probability 1/3
2 is changed to 0 with probability 1/3
2 is changed to 3 with probability 2/3
Values of 3 and above are not changed
Random Rounding
All values/numbers in the table are rounded up or down to base 3.
1-4 Rounding
Randomly changes any values in the table between 1 and 4 inclusive to a random number between 1 and 4 with equal probability distribution.
0 is not changed
1 is changed to 1, 2, 3 or 4 with equal probability
2 is changed to 1, 2, 3 or 4 with equal probability
3 is changed to 1, 2, 3 or 4 with equal probability
4 is changed to 1, 2, 3 or 4 with equal probability
Values of 5 and above are not changed.
Graduated Rounding
This rule is similar to random rounding in that all values are rounded. However, the amount by which values are rounded depends on the size of the value.
0 is not changed
Values from 1 to 18 are randomly rounded to one of the two nearest multiples of 3
Values of 19 randomly round to 18 or 20
Values from 20 to 100 randomly round to one of the two nearest multiples of 5
Values greater than 100 randomly round to one of the two nearest multiples of 10.
Confidentiality rules may also be configured at the server level, using the Data Control confidentiality plugin. See Confidentiality Plugin Overview for more information.
