We appreciate that the observed variation in IVC may relate to differences in standing height. The illustration to the right shows the bell-shaped distribution of the independent variable, and tilted by 90 degrees, similarly of the dependent variable; the height of the bell corresponds to the number of observations. Clearly much of the variability in Y (the IVC) can be explained by differences in X (standing height). This does not account for all variability, however, otherwise all points would be on a straight line.
We previously addressed systematic and random measurement errors. Random errors in both IVC and length explain part of the scatter about the regression line. The remaining part is due to other ‘errors’, such as differences in body build, or in the properties of chest, lung parenchyma and airways, so that people of the same length do not have the same IVC. Also, for the same standing height males have a larger IVC than females. Similarly there are ethnic differences: Caucasians have a larger IVC for the same length than black people. These all represent sources of variability between subjects.
In the case of repeated measurements on the same person we do not get the same result. This may be due to subject co-operation, or the measurement influencing the respiratory system and hence the result of the next measurement. In addition the time of the day, or the season of the year, prior physical exercise, etc. may affect the results. In patients the disease process, or the use of drugs, may influence the IVC. These are all sources of biological variability within subjects.
See also: Sources of variability between subjects.