species competitive interactions), but even relatively few interacting system components can lead to complex dynamics.

• Though ultimately everything is hitched to everything else, significant effects are not automatically transferred through a connected system of interacting components—locality can matter.

• Sequences of interactions can determine outcomes—program order matters.

Data and Measurement

• Only a few basic data types arise (numeric, ordinal, categorical), but these will often be interconnected and expanded (e.g., as vectors or arrays).

• Consistency of the units with which one measures a system is important.

• A variety of statistical methods exist to characterize single data sets and to make comparisons between data sets. Using such methods with discernment takes practice.

Stochasticity

• In a stochastic process, individual outcomes cannot be predicted with certainty. Rather, these outcomes are determined randomly according to a probability distribution that arises from the underlying mechanisms of the process. Probabilities for measurements that are continuous (height, weight, etc.), and those that are discrete (sex, cell type) arise in many biological contexts.

• Risk can be identified and estimated.

• There are ways to determine if an experimental result is significant.

• There are instances when stochasticity is significant and averages are not sufficient.

Visualizing

• There are diverse methods to display data.

• Simple line and bar graphs are often not sufficient.

• Nonlinear transformations can yield new insights.

Algorithms

• These are rules that determine the types of interactions in a system, how decisions are made, and the time course of system response.

• These can be thought of as a sequence of actions similar to a com-