I can think of a case where the rate of delinquency in returning books depends on how many book you have. Thus the more outstanding books the less likely to return them. Would that be considered MNAR?
Hey, can you please explain what do you mean by error rate different/same for different axes? How do I know which variable my missing value feature is sort of dependent on?
Hey there. In my study, six different raters evaluated the presentation skills of 70 students. All students were rated by each of the raters at four different times. I now have the following data: 6 raters x 70 students x 4 measurement times x 17 items of the questionnaire. Since the evaluation took place on the basis of videos, body language could be marked as "not assessable" in the items of the scale (if, for example, gestures could not be evaluated due to the camera setting). I rate these cases as missing values and according to your examples, I would assume that these missing values are to be assessed as MCAR, since the missing values are not related to the remaining questionnaire items. Is this assumption correct? If I want to check the MCAR, for example with SPSS, which variables do I include in the analysis? Do I analyse on the basis of all items (the remaining scales are voice, expression, confidence and engagement) of the questionnaire or do I analyse for each subscale individually? So for example all variables of the scale body language of all raters at all measuring points? Thank you very much!
I'm not quite following the realism in the MCAR example. Why would a value be missing? When someone gets a new library card it makes sense to set their value to 0 and then increment automatically when a book turns overdue. At what point would a value go missing due to manual error?
This is the best explanation! PERFECT and friendly to my inner child. Simple, straight forward, analogue! Thank you
The perfect explanation for the problem of missing data. Many thanks!
So insanely helpful. Thank you so, so much for providing such a clear explanation!!
Thank you so much for your video! Perfect examples!
Great explaination! Thank you!
best explanation ever, thank you!
Great explanation! thanks.
Very nice explanation of a complex topic. Looking at the rates makes a difference.
Glad you think so!
Nicely done.! Thanks for explaining it so well.
Much more clear than my lecture:) Referring USYD here opps
Best explanation. Thank you.
Wow you made this seem so easy, thanks so much
Great video!
AMAZING EXPLANATION. Thanks!
Glad it was helpful!
good job friend
Very clear thank you !
I can think of a case where the rate of delinquency in returning books depends on how many book you have. Thus the more outstanding books the less likely to return them. Would that be considered MNAR?
great video. thanks
Great explanation! Even for non native speaker
Hey, can you please explain what do you mean by error rate different/same for different axes? How do I know which variable my missing value feature is sort of dependent on?
every good vidoes! thank you !
Very helpful
I just don't understand why more books I truly have but less likely I tell you.
Amazing!
Hey there.
In my study, six different raters evaluated the presentation skills of 70 students. All students were rated by each of the raters at four different times. I now have the following data: 6 raters x 70 students x 4 measurement times x 17 items of the questionnaire.
Since the evaluation took place on the basis of videos, body language could be marked as "not assessable" in the items of the scale (if, for example, gestures could not be evaluated due to the camera setting). I rate these cases as missing values and according to your examples, I would assume that these missing values are to be assessed as MCAR, since the missing values are not related to the remaining questionnaire items.
Is this assumption correct?
If I want to check the MCAR, for example with SPSS, which variables do I include in the analysis? Do I analyse on the basis of all items (the remaining scales are voice, expression, confidence and engagement) of the questionnaire or do I analyse for each subscale individually? So for example all variables of the scale body language of all raters at all measuring points?
Thank you very much!
I'm not quite following the realism in the MCAR example. Why would a value be missing? When someone gets a new library card it makes sense to set their value to 0 and then increment automatically when a book turns overdue. At what point would a value go missing due to manual error?
Thank you.
Who is send here by Schober?:D