I've been working on a high school project attempting to determine whether or not there exists a relationship (and if it exists, information on the strength and duration of the relationship) between stock market data and election polls (both in time series format, n=250). While I'm moderately familiar on how to fit an ARIMA Model to a univariate time series using the Box-Jenkins approach, I've only encountered confusion when attempting to fit a model incorporating two variables.

Another response to a question similar to mine recommended using a transfer function, though I don't know if that would be applicable in my case as the goal is more to reveal information about a potential relationship, not necessarily to derive an exact formula linking the two series to be used for prediction. Implementing a bivariate ARIMA model in R appears to be as easy as simply adding an additional series, yet there seems to be surprisingly little literature on this or how to interpret such a model.


closed as off-topic by Firebug, Peter Flom Oct 23 '17 at 11:20

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or performing routine operations within a statistical computing platform. If the latter, you could try the support links we maintain." – Peter Flom
If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ Maybe you could try the var model (vector auto-regression model), it can deal with multiple variables. $\endgroup$ – wolfe Mar 24 '17 at 14:04
  • $\begingroup$ Possible duplicate of Fitting a multivariate ARIMA model with R $\endgroup$ – Firebug Oct 23 '17 at 10:56

You can start by reviewing a very basic ( and largely presumptive ) tutorial here https://onlinecourses.science.psu.edu/stat510/node/75/ . Review How to include control variables in an Intervention analysis with ARIMA? where I contributed to the discussion. The issue is to develop the relationship in a robust manner where anomalies do not distort.


Not the answer you're looking for? Browse other questions tagged or ask your own question.