Recoding Data Into Different Variable

Saturday, December 18, 2010 JS Hanniffy 0 Comments

Here is my favourite task of data analysis and computing, is to recode data into different variable. The purpose of recoding data into different variable is to change your old data into new (different) variable but keep its original variable. 

Go to transform then click recode into different variable (see the table below).

Table 1

And then a  new window will appear (see the table below). In a new window, it will show all the variables, choose one of variables you'd like to recode into different variable. Let's take a variable of satisfaction with life; move this variable by using arrow into the side form; type a new name for the variable e.g. Newsat; as well as the label e.g. Newsatlife; all typings should be done without a space. And then click Old and New values after this step is done.

Table 2

Presumably, you want to change the variable of Satisfaction with life into two groups so what you need to do is go to Range and then fill the the empty box with appropriate values (look to your variable view of data). For example, in my data the satisfaction with life has its value up to five values , namely: 1. Very satisfied, 2. Satisfied, 3. Ambivalent, 4. Dissatisfied, 5. Very dissatisfied. So to devide these values into groups we need grouping these values into two values. Let's sort it the values 1 and 2 into group 1 and the values 3, 4, 5 into group 2. 

Thus, the range in the Old Value will be 1 through 2 and the New Value will be 1 and so on. Click add in the form of old ---> new and then click continue. You can also add a missing value.




Table 3


And the output will show its result like the table below.

Table 4




DBLN, 15.32-181210

Blogger since 2008, writing with a fresh perspective.

Twitter Feeds