Hello. I have a question about combining time series datasets.

We are trying an analysis technique in which we try to effectively double our temporal resolution by offsetting a repeated presentation of stimuli. Due to limitations in our desired acquisition parameters, we are using a TR of 3000 msec. The scanning physicist suggested showing the stimuli twice, with the second time-series offset by 1500 msec. This way, we would sample at different timepoints of the response relative to the stimulus onset. Before analysis, we would combine the time series by intercalating the offset time-points, so we would be able to model responses with TRs of 1500 msec.

I am looking for the best way to combine time-series data in this fashion. Any suggestions would be appreciated.

thank you,
George