Hi, I'm doing my master dissertation. My research aims to investigate how individual factors (intrinsic motivation, paradox mindset, task significance, resource available, proactive personality) in total five independent variables impact employee creativity mediated by unlearning(the process employees abandon old outdated knowledge and skills to learn new ones).
When I was trying to do the modification indices in AMOS, it said the an error occurred while attempting to fit the model that modification indices cannot be computed with incomplete data.
Have you opted for estimating the means and intercepts? In case of missing data, most likely AMOS cannot function unless you opt for it (in the "analysis properties" dialog box).
Hope it helps!
Regards,
Rajiv
A "trick" that I've used before, which I picked up from a professor in grad school, is to create a correlation matrix and then import that into Amos as the dataset. Then, run your same model but using the correlation matrix, though only using these results to get the modification indices - then, you can apply those results to make modifications to your main model which uses the original dataset with the missing data. Note that Amos requires correlation matrices to be formatted in a certain way - you need the N, for example.
MISSING DATA makes the data set imperfect or defective. Any attempt to "fix", no matter how sophisticated it may sound, is nothing more than an act of dishonesty because the person is "making up data" that does not exist. The row that has missing data should be removed from the column. This may shorten your data segment (sample size) but it does not raise any ethical issues.
DATA IMPUTATION or fixing missing data begs the question of data integrity. What is the reliability of the data set when some of the data points in the set is missing and is "made up" by "fixing" or by "imputation" or whatever name you may call this faking of data that does not exist?
- Badges
- Science topic
- Similar topics
- Medical Science Nursing
- Dissertations