3. SEM | SPSS AMOS Lecture Series - Data Screening and Imputation using SPSS

แชร์
ฝัง
  • เผยแพร่เมื่อ 1 ก.พ. 2025

ความคิดเห็น • 28

  • @rupashrimurthy8309
    @rupashrimurthy8309 2 ปีที่แล้ว +1

    Extremely helpful video on data screening. Appreciate your effort. It has helped me understand the process of data screening completely. Thank you!

  • @atulcharasiya2809
    @atulcharasiya2809 11 หลายเดือนก่อน

    Using this series for my undergrad thesis study
    Seems really apt, probably better than online paid courses too (though I haven't gone through them rigorously)
    Love your approach and steps!!
    Thank you!!

    • @researchwithfawad
      @researchwithfawad  11 หลายเดือนก่อน +1

      Thank you. I am glad you liked it.

  • @ibrahimabdullatifibrahim9942
    @ibrahimabdullatifibrahim9942 ปีที่แล้ว +1

    Thanks, sir for great effort
    For which Likert scale we can delete these cases with SD less than 0.25. Likert scale may be on 3, 4,5,6,7,9 points.
    Greetings

    • @researchwithfawad
      @researchwithfawad  ปีที่แล้ว

      Thanks for Watching. Yes Please.

    • @selimy5553
      @selimy5553 ปีที่แล้ว

      Respected Sir, thanks for this very informative video; can you please tell us should we delete cases with SD less than 0.25 even if we have a 5-point likert scale? Bcs, there is not information about what to do if we have 5 point likert scale, or 9 or 3 etc. in Collier's book.
      Is there a formula to calculate SD cut off point for the purpose of cleaning cases for 5 point likert scale? Thanks a lot🙏@@researchwithfawad

  • @talhamansoor7108
    @talhamansoor7108 3 ปีที่แล้ว +1

    Great knowledge

  • @isratjahanlinda9278
    @isratjahanlinda9278 11 วันที่ผ่านมา

    Where can I find a dataset please to practice your explained sessions? Please help.

    • @researchwithfawad
      @researchwithfawad  10 วันที่ผ่านมา

      Please visit www.researchwithfawad.com

  • @FooodConfusion
    @FooodConfusion 2 ปีที่แล้ว +1

    Dr Fawad, do you have any reference that states that we have to delete the response whose STDEVP is less than.025, any paper or book that states this?

    • @researchwithfawad
      @researchwithfawad  2 ปีที่แล้ว +3

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

  • @Alhamzah_F_Abbas
    @Alhamzah_F_Abbas 2 ปีที่แล้ว

    Thank you, prof. can we use the same process when we use smartPLS?

    • @researchwithfawad
      @researchwithfawad  2 ปีที่แล้ว +1

      Pleasure. No.

    • @Alhamzah_F_Abbas
      @Alhamzah_F_Abbas 2 ปีที่แล้ว

      @@researchwithfawad in this case ,how we gonna deal with missing data in PLS?

    • @researchwithfawad
      @researchwithfawad  2 ปีที่แล้ว +1

      There are mean replacement options that you can choose from when running the analysis

  • @sethulakshmig2999
    @sethulakshmig2999 2 ปีที่แล้ว +1

    Really useful

  • @FooodConfusion
    @FooodConfusion 2 ปีที่แล้ว +1

    Dear Dr. Fawad, great job, you are doing fantastic job that will last forever.
    Kindly clarify one question of mine based on your expertiese:
    There are four Latent constructs, each containing 6 items/questions. Where in my upcoming study, I want to take all those 24 items (6 questions each) and test one higher order construct as an independant Variable...kindly confirm in this case, if any of my item is deleted because of poor loading out all those 24 items, will it cause any trouble? In simple words there were four different constructs representing one higher element, but I want to treat all these items under one higher order construct.
    Thanks

    • @researchwithfawad
      @researchwithfawad  2 ปีที่แล้ว

      Thanks. I am glad you liked it.
      No deleting n item from LOC will not cause ajy trouble. LoC will have to be validated. After LOCs are validated, validate HOCs as well.

  • @zahrabakhtiary7345
    @zahrabakhtiary7345 10 หลายเดือนก่อน

    If the data distribution is not normal, are we allowed to manipulate the data? and how? (all data are in likert spectrum) Thank you in advance. 😇🥰

    • @researchwithfawad
      @researchwithfawad  10 หลายเดือนก่อน

      Thanks for watching. No, we are not allowed to manipulate the data.

  • @mrchilly28
    @mrchilly28 ปีที่แล้ว

    Hi Dr Fawad, thank you for these lectures, I've been using them to teach myself SEM for my PhD. For my work I need to use the references that found up to 20-30% of missing data can be remedied with imputation. You cite Hair et al., 2009, and Eekhout et al., 2013, however I'm unable to find the full references anywhere. Would you be able to send me these as full references please?
    Thank you

    • @researchwithfawad
      @researchwithfawad  ปีที่แล้ว

      Thanks. I am glad you liked them. Please find below the full references
      Eekhout, Iris, Henrica De Vet, Jos Twisk, Jaap P.L. Brand, Michiel de Boer, and Martijn W. Heymans. (2013), “Missing Data in Multi-Item Instrument Were Best Handled by Multiple Imputation at the Item Score Level”, Journal of Clinical Epidemiology, 67 (3), 40-55.
      Hair, Joseph F., William C. Black, Barry J. Babin, and Rolph E. Anderson. (2009), Multivariate Data Analysis (7th ed.). Upper Saddle River, NJ: Prentice Hall.

    • @mrchilly28
      @mrchilly28 ปีที่แล้ว +1

      @@researchwithfawad Thank you so much

  • @yinanlee5473
    @yinanlee5473 3 ปีที่แล้ว

    Dear Sir, I have seen all your videos about PLS and also AMOS, so helpful for me. May I ask your a questions about the comments I have received from a reviewer. In the model of my study, there are two formative constructs which made me to run PLS. I've conducted EFA for exploring the dimensionality. The reviewer suggested me to do also CFA after EFA, and then run the model with Smart-PLS. I'm confused about if it will work like that ..... Thank you very much for your help since now.
    Best regards

    • @researchwithfawad
      @researchwithfawad  3 ปีที่แล้ว

      If your items are based on established scale you do not need to perform EFA. As for CFA, Smart-pls only deals with part of it and can assess reliability and validity. Whereas Model fit indices are not yet established for PLS.
      The indicators for reflective constructs ar required to be correlated, whereas the indicators for formative constructs are not required to be correlated. As such, formative constructs should not be expected to properly factor in an EFA