User Tools

Site Tools



This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision Both sides next revision
start [2022/07/29 10:44] old revision restored (2022/03/31 04:37)
start [2022/07/29 10:44] old revision restored (2022/03/23 03:31)
Line 1: Line 1:
 ====== Generate Data from Graphs ====== ====== Generate Data from Graphs ======
 [[|WebPlotDigitizer]] [[|WebPlotDigitizer]]
Line 55: Line 56:
 The key is to find the number of subjects that would let us gain the statistical power for the clincial trial. The key is to find the number of subjects that would let us gain the statistical power for the clincial trial.
-=====Structure of Statistics=====+************************************************************** 
 +***                  Interview Questions                   *** 
 +Randomized two arms, active and placebo 
 +outcome variable y 
 +measure twice, once baseline, once post treatment measurement. 
 +Whether getting active improves y? 
 +You have two treatment groups and you have two measurements on each group. 
 +Best single answer: 
 +Two measurements on placebo and active, what are you comparing on the t test? t test is just a simple test on the two groups' mean difference, no model. t test can not add covariate, can not adjust anything 
 +1. get change from the baseline for the placebo 
 +2. get change from the baseline fort the active. 
 +Compare the mean difference of these two changes. 
 +2nd method just to compare post treatment values: 
 +Compare the mean differences between the two treatment results.  Do a t test on the baseline to see if comparable. 
 +Besides t test, we can use Anova, as with Anova we can use model, y(response, dependent variable)=treatment.  
 +PROC ANOVA DATA=datasetname; 
 +     CLASS factorvars (such as treatment); 
 +     MODEL responsevar (such as change) factorvars; * we can have baseline as covariate in the model 
 +If active has 10 observations, placebo has 5 observations, missing data, not balanced, using glm method handling missing data. 
 +Outcome is 0, 1, how to determine the active is helping? 
 +What if it's continuous variable, but not normally distributed, what would you do? 
 +If you are doing Anova, what would you do? Does getting active treatment improve on outcome y?  What is the model would be? 
 +What is the dependent variable in the ANOVA model? Left hand side is dependent variable 
 +In t test, the dependent variable could be the change from the baseline.  Explanatory variable is the treatment.  
 +The dependant variable is just the post treatment value or the change from the baseline, and explantory variable just is the treatment on the right hand side. 
 +Change from the baseline treatment, the result exactly is the t test. 
 +Question: Suppose outcome variable y 0, 1, or yes no, we got baseline and treatment, and we got some other variables, gender taking into account.  treatment has effect on y after adjusting for gender.  What would you analyze that? 
 +Think more on the statistical side, not the programming side. 
 +======Knowledge Structure of Biostatisticians====== 
 +  * theory: probability, inference, linear models, others 
 +  * applications: ANOVA, ANCOVA, t-test, F test, survival analysis, categorical analysisys 
 +  * experimental design 
 +  * Environments / FDA regulations 
 +  * Software : SAS, R
start.txt · Last modified: 2022/10/07 18:04 by