Inconsistent Stimulus Timing & Timing Reports in Data .csv

I have an experiment that loops through a .xlsx file. Each loop has different duration and I provide those durations in the file as well. When I repeat the same experiment multiple times, I get different results on the timing of each stimulus. The entire experiment should last 3 minutes and 6 seconds but when I actually run it it seems to last anywhere between 3 minutes and 3 minutes 12 seconds. I checked the data files to see where the inconsistencies where occurring and found some even weirder issues with the data.

When I look at the data .csv files, the text.started and text.stopped columns either show that the start and stop times are the same or that the stop time is before the start time which doesn’t make any sense. I watched the experiment and the stimulus were definitely presented start and stop times with 0 duration or negative duration are not accurate representations of what actually happened. Any ideas about what could be going wrong with the data reporting?