Abstract
Many authors have commented on the difficulties caused by the inability to measure dietary intake variables accurately. Due to this inaccuracy many dietary studies are now analyzed by errors-in-variables methods. Previous work has shown that in such a setting the necessary sample size of a study to establish the relationship between intake of a nutrient and the occurrance of a disease is highly sensitive to two separate correlations: (i) the correlation between the true intake and the reported intake of the nutrient; and (ii) the intraindividual correlation between two reported intakes measured close together in time. In this paper we develop and apply methods of estimating and making inferences about these correlations. Our methods are applied to data from the Women’s Health Trial Vanguard Study.
| Original language | English |
|---|---|
| Pages (from-to) | 2327-2340 |
| Number of pages | 14 |
| Journal | Communications in Statistics - Theory and Methods |
| Volume | 22 |
| Issue number | 8 |
| DOIs | |
| State | Published - 1 Jan 1993 |
| Externally published | Yes |
Bibliographical note
Funding Information:This work is based in part on the first author's Ph.D. dissertation at Texas A&M University. His research was supported by grants to his major professor, R. J. Carroll, from the National Institute of General Medical Sciences and the National Cancer Institute.
Funding
This work is based in part on the first author's Ph.D. dissertation at Texas A&M University. His research was supported by grants to his major professor, R. J. Carroll, from the National Institute of General Medical Sciences and the National Cancer Institute.
| Funders | Funder number |
|---|---|
| National Cancer Institute | |
| National Institute of General Medical Sciences |