I have a time series data for shipment with following variables:

  1. Year: 2008, 2009, 2010, 2011, 2012, 2013
  2. Month: jan, feb, ..., dec
  3. Number of ordering days
  4. Shipment Volume

I want to know the effect of the number of ordering days on the shipment volume.

Can this can be done using regression method? If so, should I consider month and year as categorical variables?

  • $\begingroup$ You might want to look at the US Census Bureau's X13 software census.gov/srd/www/x13as $\endgroup$
    – david25272
    Commented Feb 5, 2014 at 2:19
  • $\begingroup$ You want to fit a time series model, which will take the correlation over months into account. $\endgroup$ Commented Mar 10, 2016 at 5:01

2 Answers 2


You could take a time series approach to your analysis. Have the data as times series with monthly frequency. This will allow you to look at the variation in shipment volume due to the cycles and the actual effect of number of days.

  • 2
    $\begingroup$ Thank you so much for your answer...Can explain it more...It would be great, $\endgroup$
    – Arushi
    Commented Oct 8, 2013 at 6:39

Sounds like an interesting problem. You could take a look at this free online book that I think is really good.

I think "example: births" on page 78 might give you some ideas for how to analyze your data.

Using different types of independent variable codings for (in the link's notation $\mathbf{x}$) in R is well described by this page


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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