We have a group of patients who underwent bariatric surgery, their BMI and creatinine clearance were measured before surgery and monthly for the following three months, in total we have 4 pairs of observations for both variables per individual with 4 repetitions in time.
What we want to do is to verify if both variables are correlated, however we cannot figure out how to do it in this particular scenario. Any suggestion?
In addition to being able to obtain the correlation coefficient for each time between BMI and creatinine clearance and Scatterplot Matrix,
You can also create a regression model between the data.
creatinine clearance = b0 + b1 BMI + b2 Time
Of course, this model does not have to be linear and can be nonlinear. The Curve Estimation tool in SPSS software is useful in this regard.
1. for time series cross-correlation 4 repeats is not not enough...
2. Just do ANOVA at start to see if differences are there. Then do anova at time two using first time as covariate. time 3 anova with time 2 as covariate...time 4 anova with time 3 as covariate.
3. TAKE GROUP MEANS AT EACH TIME AND THEN DO CORRELATION USING THOSE MEAN VALUES.
- Badges
- Science topic
- Similar topics
- Higher Education
- Universities