Nonstationarity
When the data-generating process changes over time
Nonstationarity means the underlying process that produces data changes over time. When conditions shift, older evidence may no longer apply without adjustment: an effect can disappear, invert, or become conditional simply because the world is different.
This matters for synthesis because time-driven change can masquerade as contradiction. Knowledge synthesis therefore separates nonstationarity from true incompatibility by tracking when and where claims were intended to apply.