Q: What’s the easiest way to specify covariates in study files?

A: When there are a number of subjects being analyzed using the same basic model, certain attributes may vary among the subjects, such as weight. An exogenous input, for example, might be specified as the amount of the bolus divided by the weight of the subject. Assuming that the magnitude of the bolus is constant, the least efficient method of specifying the input would be to actually calculate the value of bolus / weight for each subject and to enter this value into the Exogenous Input window in each respective study file.

A slightly better method would be to define the input by the equation, ex1 = 1.4e6 / weight, in the Exogenous Input window, and then write an equation for each subject, such as weight = 43.6, in the Equations window. While this is an improvement, it still stores subject information in both the Model and Data.

The best method is to define the input by equation in the Exogenous Input window, such as ex1 = 1.4e6 / weight, but to use the data constant feature in SAAM II for the weight. As explained in the Manual and Help, a data constant has the same effect as declaring one value for a data element at all experiment times. For this example, a separate line would be added to the data like:

This is also equivalent to using the equation, weight = 43.6, in the model. The advantage to this method is that it allows the model to remain the same for each subject and relegates the differences to the data. This makes entering the specific information for each subject faster and more efficient.

Note that the amount of the bolus could also be stored as a data constant if it varies among the subjects.

<< Back to Support Page