Related Experiment Videos
Measurement error in human dental mensuration
J A Kieser1, H T Groeneveld, J McKee
1Department of Oral Pathology, University of Witwatersrand, Johannesburg, South Africa.
Annals of Human Biology
|November 1, 1990
Summary
Human odontometric data reliability is low due to significant interobserver error. Mesiodistal measurements showed more imprecision and bias than buccolingual measurements, indicating a need for standardized methods.
Area of Science:
- Forensic Anthropology
- Dental Morphology
- Biometrics
Background:
- Odontometric data are crucial for human identification and anthropological studies.
- Assessing the reliability of these measurements is essential for accurate analysis.
Purpose of the Study:
- To evaluate the reliability of human odontometric data.
- To determine the interobserver error in mesiodistal and buccolingual tooth dimensions.
Main Methods:
- Sixty teeth were measured by three observers using consistent definitions.
- Measurements were repeated after two months.
- Statistical analyses included Pearsonian correlation, mean absolute error, Bland-Altman procedures, Student t-tests, and ANOVA.
Main Results:
- Odontometric data exhibited a high interobserver error component.
- Mesiodistal dimensions demonstrated greater imprecision and bias compared to buccolingual measurements.
- ANOVA results indicated that bias stemmed from interobserver error, not measurement timing.
Conclusions:
- Human odontometric data possess significant interobserver variability.
- Standardization of measurement techniques is recommended to improve data reliability.
- Buccolingual measurements appear more reliable than mesiodistal ones.