Long post alert,
Statistical Question:
we are conducting a non-randomized prospective case control study, to test the effect of a surgical complication on the functional outcome of the subjects (both groups had the same surgery, while the control group did not develop the tested complication), we identified some confounder that might have effect of the outcome too (age, sex, level of activity,…etc). We tried to control for these confounders at the level of the study design by matching and restriction.
Now, we are at the stage of statistical analysis; we used the SPSS software to conduct one-way analysis of covariance (ANCOVA) to test the effect of the confounders. However, I needed also to calculate the mean difference and the 95% confidence interval of the mean difference, so I conducted the independent sample T test to calculate these values when the effect of the confounders was found to be insignificant by the ANCOVA.
Is this the correct way or I cannot do the independent sample t – test once I have done the ANCOVA?