Inferring a shifted model from data
Key knowledgeIdentify possible model families from graphical or tabulated data features.
Build the idea
A vertical shift can hide the constant ratio of an exponential pattern. Consecutive differences may reveal the decay ratio before the long-run level is known. Circular data require a midline, amplitude, period and phase. A successful fit is a candidate model; meaningful interpretation and comparison with later observations are still necessary.
Try the sample task
At t=0,1,2,3, a measured quantity is 18,12,9,7.5. Fit y=c+A rt by first comparing successive differences. State the long-run value and predict y at t=4.
The full worksheet includes the worked solution and remaining task set.
What this activity covers
Focused on identifying shifted-exponential and circular families from idealised data, not a full regression analysis.