Hi,
This is a 4-day training.
Before and after the 4-day training, I got the participants' score of variable A：
- Pre-A;
- Post-A
During the 4 days, I got each day's training quality scores:
- (Day1) B1, C1;
- (Day2) B2, C2;
- (Day3) B3, C3;
- (Day4) B4, C4.
All the scores are questionnaire scores.
How to use each day's quality scores to predict the changes of A?
Thanks.
Some questions:
What is the difference between C and B?
In order to find an appropriate model, it is necessary to know your assumptions about the relations between C, B and A and the occurences of C and B among themselves, i.e., does B1 has an effect on B2 etc?
Hi, Sven Greving , thanks for your reply.
B and C are the subscales of a questionnaire.
Hypothesis: the increase of A scores can be predicted by the training quality, which is reflected by the B and C.
Generally speaking, B score and C score may increase along with the consecutive training from day 1 to day 4. But the B and C scores may fluctuate with the quality of training on a certain day.
But, is B1 exactly the same as B2 only on another day?
In general, I think Multilevel Modeling might be suitable here as daily changes within persons are ideal to be modeled as random effects.
Dear Yu zheng Wang,
You got Pre-A (say A0) measures with the measures of questions B and C (B0 & C0) --> your regression should be for A0 = B0 + C0
After 1st day of training, you have A1 with the measure of questions B1 & C1.
Your first-day regression then should be for A1 = B1 + C1.
To get the difference, between the A0 and A1, your regression should be as
A1 = A0 + B1 + C1. If the coefficients of variables B1 and C1 are significant, then your first-day training has significant improvement from the previous day.
You can have the tests in a similar way for other days too.
Regression for day 2: A2 = A1 + B2 + C2
Regression for day 3: A3 = A2 + B3 + C3
Regression for day 4: A4 = A3 + B4 + C4
Hope this will help you. Good luck with your studies.
Regards,
S. Senthilnathan.
B1 and B2 are measured by the same scale. I agree with Multilevel Modeling.
Sven Greving
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training