Reply With QuoteCome on. Nobody knows this?
I have an assignment and I’m really confused. The purpose of the study is to investigate the type of training on reading skills of learners after holding their intrinsic motivation constant. Accordingly Independent variable is the type of training with three levels, Dependent variable is students’ reading skills, and Covariate is their intrinsic motivation. 50 students are involved in the study.
By the way, I was given the descriptive statistics tables, Levene’s test, and between subjects output. I don’t have the data. ANCOVA is significant (p<0.1). The levene’s test is significant, too, which means that homogeneity of variance assumption is violated. I have two questions:
1) I can’t check the assumption of normality as I don’t have the data. I only have the information below. By looking at the table below, can I say that normality assumption is violated too (due to unequal and small sample size)? And if violated (as homogeneity of variance was violated as well) what should I do?).
Descriptive statistics
Method A (N=10) (M= 63) (SD=8.04)
Method B (N=20) (M= 53) (SD=5.93)
Control Group (N=20) (M= 44) (SD=8.19)
After ANCOVA
Method A (N=10) (M= 61.95) (SE=2.29)
Method B (N=20) (M= 53.126) (SE=1.86)
Control Group (N=20) (M= 44.89) (SE= 1.16)
2) Can I assume that random assignment has been done? I have searched the internet and read lots of documents but I couldn’t find any information regarding random assignment with unequal sample size? I mean is it realistic to randomly assign ten subjects to one group, 20 subjects to the second group and 20 subjects to the last group?
If you could help me, or at least suggest a reference, I would be grateful. Thanks in advance.
Come on. Nobody knows this?
Posting Permissions
|
|