The Box–Jenkins method can be used to select the parameters of an ARIMA (“uh-REE-muh”) model for data indexed by this variable. For 10 points each:
[10e] Name this variable plotted on the x-axis of namesake “series” of data. Stochastic processes like Brownian motion and stock market returns are often indexed by this variable.
ANSWER: time [prompt on t]
[10h] The Box–Jenkins method uses a variant of correlation denoted by this adjective, which first regresses off data at shorter lags. Similarly, this adjective denotes a regression diagnostics method that plots the response against the added variable, after regressing off all other variables.
ANSWER: partial [accept partial correlation or partial autocorrelation function; accept partial regression]
[10m] An early step of the Box–Jenkins method is to transform or apply differencing to better attain this property. A time series with this property has the same joint probability at any point in time.
ANSWER: stationary [or stationarity]
<Rutgers A, Other Science>