After key-in your data yang beratus tu kenkawan pun rasa cam loya loya nak muntah gitu (ini untuk case cam I yang tak mampu nak upah org key -in ye).This is what I diagnosed as "Syndrom Mabuk Data", the next step to think about is how to screen the data. Just want to share apa yang nak buat during this screening process. I think how much to do and how serious you are in dealing with data screening bergantung pada your s/vsor school of thought. Ada org tak le buat banyak benor and ada gak yang tak report in their thesis. But, as usual I am not that "lucky" one. heeeee.....
Outliers: 1) univariate - z score
2) bivariate- scatterplots
3) multivariate- Mahalanobis D
Normality: 1) skewness & kurtosis
2) Kolmogorov-Smirnov
3) examine linearity of graph MahalanobisVs Chi square values
Linearity: 1) Scatterplots
Homoscedasticity: 1) Scatterplots
Homogeneity of variance: 1)univariate-lavene test
2) multivariate- Box's M test
Bila dah buat ni baru le kita bole cakap yang data kita dah meet basic assumptions and our data tu suitable for further multivariate analysis.
Selamat "cuci mencuci" kenkawan....
