Stratified Initial Values and Change Scores

Abstract

A common question that is encountered in many studies is relating a change in a variable to the variable’s initial value. It is well known that regressing the change on the initial value creates a spurious association and misleads the unwary. One attempt to cope with this problem has been to stratify that data based on the initial value. We show that this approach does not solve the problem. Theoretical and simulation results for independent normal and independent exponential random variables are provided.

Key-Words

Stratification - Change Score - Initial Value

• REFERENCES

• 1 Blomqvist N. On the relation between change and initial value. J Amer Stat Assn 1987; 72: 746-9.
• 2 Hayes RJ. Methods for assessing whether change depends on initial value. Statistics in Medicine 1988; 07: 915-27.
• 3 Oldham PD. A Note on the analysis of repeated measurements of the same subjects. J Chron Diseases 1962; 15: 969-77.
• 4 Wilson TW, Cohen RD, Lachenbruch PA. et al. Is the relationship between supine blood pressure and postural changes in blood pressure artefactual?. Results from the Kuopio Ischemic Heart Disease Risk Factor Study. (to be publ).
• 5 Harris T, Lipsitz LA, Kleinman JC, Cornoni-Huntley J. Postural change in blood pressure associated with age and systolic blood pressure: The National Health and Nutrition Examination Survey II. Medical Sciences 1991; 46: M159-63.
• 6 Computing Resource Center. STATA Reference Manual (Release 3, 5th ed). Santa Monica CA: 1992