Data Screening, Cleaning and How to Replace Missing Values in SPSS

The session guides on how to check respondent misconduct using MS Excel. Further, This session discusses in detail missing data and how to replace missing values using the Imputation techniques (Series Mean and Linear Interpolation Method) in SPSS
For Notes Visit
researchwithfawad.com/index.p...
#SPSS #SPSSTutorial #SPSSDataAnalysis #MissingData #SeriesMean #LinearInterpolationMethod #dataanalysis #datacleaning

[@Junaidkhan-ce4yd]

Thank you so much for such a nice and clear explanation

[@researchwithfawad]

Pleasure. I am glad you liked it.

[@Alhamzah_F_Abbas]

Great explanation. well done

[@researchwithfawad]

Glad you liked it!

[@talhamansoor7108]

That's great

[@researchwithfawad]

Thanks. I am glad you liked it

[@gentesipam3895]

thank you so much

[@researchwithfawad]

You're welcome!

[@480sachin]

EXCELLENT SIR

[@researchwithfawad]

Thanks. I am glad you liked it.

[@HaPham-fq1xt]

Thank you so much for the great explanation. However, you mentioned that the data should be removed if a standard deviation is under .25. Is that any reference for it?

[@researchwithfawad]

Thanks. I am glad you liked it. You may refer to
Collier, J. E. (2020). Applied structural equation modeling using AMOS: Basic to advanced techniques. Routledge.

[@alexanderstevens2077]

Also, would you recommend examing missing values on a case-by-case basis for missing items that may have been unintentionally skipped by a respondend? It seems to me that, for variables missing > 10% doing this would allow the researcher to determine if, due to some factor, the respondent unintentionally skipped it and directly fill in the best possible value

[@researchwithfawad]

Yes, you can perform that imputation and there is also support in the literature for it.

[@alexanderstevens2077]

Can you please include a video on using Expectation Maximization for data imputation?

[@researchwithfawad]

Thanks. Good idea. Hopefully soon.

[@alexanderstevens2077]

Unengaged respondents are those that answer all the same values or does so in patterns such as 1111, 3333, 2222, 4444, and so on. How do you detect those?

[@researchwithfawad]

Thanks for your comment. One way to do it is to take Standard Deviation of individual constructs for each respondents. Hope this helps.

[@reisterantiporta9767]

Hello! Is it okay if I categorize Likert scale responses as ordinal in SPSS? Or should I categorize it as scale?

[@researchwithfawad]

The categorization in SPSS doesnt affect the results. You can put them as ordinal or scale.

[@amalwijenayaka410]

Dear Prof: you mentioned that the data should be removed if a standard deviation is under .25. Is there any reference? We have no access to read this book. (Applied structural equation modeling using AMOS: Basic to advanced techniques) That is why?

[@researchwithfawad]

Thanks for your interest. That book is the reference. You can quote the book.

[@mansour6629]

Thanks for the great video!
Could Mode be as well used to replace missing data?

[@researchwithfawad]

Are you trying to replace the demographics? They should not be replaced.

[@mansour6629]

@@researchwithfawad No, sir. There are some missing values in likert scale data. Is Mode an appropriate technique to deal with them?

[@researchwithfawad]

No.

[@mansour6629]

@ResearchWithFawad Thanks!
Could you please provide a good reference for full guidance on data screening?

[@researchwithfawad]

You may refer to
Collier, J. (2020). Applied structural equation modeling using AMOS: Basic to advanced techniques. Routledge.

[@saminazaidiacademy8219]

How to put multivariables into one variable for data analysis..like 5 items are showing results of one dimensions so how can i analyze

[@researchwithfawad]

Thanks for your interest.
You will need to take the sum of the individual items. Let say, i have Organizational Commitment measured using 4 items COM1, COM2, COM3, COM4
If you have it in SPSS,
Go to Transform -> Compute Variable
In the Target Variable Enter the Name of the New Variable that is to be created based on taking the average, let say COMM.
In the numeric expression type in
Mean(COM1, COM2, COM3, COM4)
Press OK. The new variable is created at the end of the Data View and is also visit in the variable view. You have now composite score for each respondent that you can use in regression.

[@saminazaidiacademy8219]

@@researchwithfawad thanks for your timely reply🙏
I have 5 items in.each COM1 ,COM2 ,COM3...Then how to compute them....to make it one COMM

[@areegtooba8260]

If we added the new column due to imputation. Now may we delete the previous one which was with missing value? How we will use this new one in applying any test?

[@researchwithfawad]

Use he newly formed variable. You can keep the old one but not use it for further analysis