SPSS for DATA ANALYSIS - Part 6

SPSS for DATA ANALYSIS – PART 6

In this section of the series of SPSS for DATA ANALYSIS, we shall be looking into the most required techniques while analyzing qualitative data – Variable Transformation. Variable transformation belongs to preprocessing part of data analysis and it involves the method of transformaing a data type into another to enable some analyses that can only be performed on a particular data type. For instance, Ade is a Data Analyst and he receives a qualitative data (say responses of participants to a survey questionnaire with options: Strongly Agree, Agree, Neutral, Disagree and Strongly Disagree) from a Client that requests for the following analyses on the data: Alpha Cronbach analysis, KMO analysis, Barlett’s analysis, AVE analysis, content analysis, correlation analysis, covariance analysis, ANOVA, T test and so forth. To perform most of these analyses, the data has to be a numeric data type.

Let’s transform the “Marital Status” variable of the data that we are using for this series into numeric data type.

Vamos comecar! (Lets get started!)… You just learned a sentence in Portuguese (Lol)!

Figure 39 shows the variable that we are interested in transforming into numeric data type which is a string data type. To transform the variable follow the arrows in the subsequent figures.

In Figure 40, click the bar “Transform” then option “Recode into Different Variables..” The option “Recode into Different Variables…” is chosen because we want the transformed the variable in another column and not to overwrite the original as that is what the option “Recode into Same Variables” will do.

In Figure 41, select (3) the variable that is to be transformed and put into the box (4) for the analysis to be performed.

In Figure 42, the name of the new (transformed) variable is to be put in the box (5) then you may add label in the box (6). Figure 43 shows the name of the new variable and label assigned.

The next step is to assign the new data (numeric) to the old data (string). Click on (7) and Figure 44 displays to input the “Old value” and “New value.” In the old value, the marital status (say Single) is to be entered then in the new value a number (say 1) is used to recode the old value. After the two data are entered, then click “Add” (10) to store the data in the box as shown in the Figure 45. Therefore, we have Single = 1.

When the same process is followed for the categories of data in the “Marital Status” variable, then the result should appear as shown in the Figure 46.

After the whole data has been transformed as shown in Figure 46 then click on the “Continue” and a new variable must have been added to the data in the SPSS.

A new variable “Marital Status Tr” is added to the data but SPSS added a two decimal of which we only the whole numbers without decimal point. So, to convert to whole number just click on the “Variable View” then under the “Decimals” property click on the “2” and replace with “0.”

Figure 50 shows the transformed variable as a whole number. So, whenever the variable “Marital status” is to be used for analysis that reaquires a numeric data type then the variable “Marital_Status_Tr” is used instead. It must be noted that nothing has changed about the original data, it was only transformed into another data type.

Conclusion

In this section of the series, data transformation has been dealt with indepth. This technique is useful in almost all data analysis that will be conducted. It could be from numeric to string or vice-versa. The bottomline is that in data analytics transforming variable is a frequent analysis that comes up. NOTE: when a data is transformed into another data type, nothing has changed about the data only its data type has been altered. It is recommended that “Recode into Different Variables” option is selected so as to preserve the original data.