Search This Blog

Wednesday, 19 April 2017

Checking Missing Values

Missing values is when your data set contain some missing values where there will be participants did not answer some items in the questionnaire or did not complete the trial in an experiment.

1. When you enter the data, leave the missing values as a blank cells


2. Go to SPSS and fill all the empty cells by click "Transform and Recode into same variable".

3. Move all the variables into the "Numeric Variables" and click on "Old and New Values".


4. On the left select "System-or-user-missing" and at the right "New Values" enter the number will not otherwise occur in the data. Click "Add" and "Continue". Click on "OK".


5. All the blank cells will replaced with the value that you entered in previous step.


6. After the step, SPSS will not include these number in any calculations you must complete one final step. Click on "Variable View".

7. At the eighth column, click on "Missing" at the first cell under the column. Click on the blue box that appears in the cell.


8. Select the "Discrete Missing Values" and enter in the box the number you choose like previous step 4. Click "OK".



9. Repeat back step 7 and step 8 for every row in the variable view in your data. You also can copy and paste into all cells below.



Conclusion : If you computing the data please consider the impact of missing values. it might be more suitable to calculate the mean instead based on number of the participants. Alternatively, some questionnaire will be manuals advise replacing missing values with the participants mean score before calculating the total.


By: Nur Fariza

References:

- Book
Pallant, J. (2013). SPSS survival manual: A step by step guide to data analysis using SPSS (5th ed.) Maidenhead: Open University Press/ McGraw-Hill.

- Website
https://www.spss-tutorials.com/spss-missing-values-tutorial/
http://stats.idre.ucla.edu/spss/modules/missing-data/
https://www.ibm.com/support/knowledgecenter/pl/SSLVMB_22.0.0/com.ibm.spss.statistics.help/spss/mva/idh_miss.htm

No comments:

Post a Comment