Millisecond Forums

Date and time in web data files

https://forums.millisecond.com/Topic13913.aspx

By solarte - 8/4/2014

Hi,

We recently found that one participant (identified by their unique subject ID) completed an online task twice on the same data and time.e.g. both entries on June 10th 2014 at 22:53:59. These are not duplicate data, because the content of the data is different; we're using Stroop task and the trials are randomly presented, and these trial order differ between the two data files.

We checked back on the Inquisit Web Data Files page, and found that the date and time displayed on the web page are not the same as those indicated on the data file. e.g. one is June 11th 2014 at 03:09:44 and the second entry is June 11th 2014 at 03:02:27. When I open the data file directly from the web page (by clicking on one entry), the date and time are still wrong (they appear as June 10th 2014 at 22:53:59 in BOTH of the data files).

Am I missing something here, or is Inquisit inconsistent with recording date and time properly?

If there's any suggestions or advice, I would very much appreciate it!!!
By seandr - 8/4/2014

Hi,
The date that's recorded in the data is in the local time of the participant's computer, whereas the dates reported in the file names and on the web page are in GMT. So, there may be some hours difference between the former and latter depending on the participant's time zone.
The dates reported in the file names and on the web page should be the same, however. If they are not, it you tell me which files they are I can take a look and see if I can figure out the cause of the discrepancy.
-Sean
By Dave - 8/4/2014

The date and time you find *inside* a given data file's date / time columns reflect the local time / timezone for the client machine the data was collected on.

Dates and times in the web interface reflect the *server's* time (in UTC), which includes the date / time making up part of the filename (IIRC).
By solarte - 8/4/2014

Thank you so much for the quick reply!
That clears up the inconsistent date/time...

So does that mean in my case, where there were duplicate identical date/time recorded, something must have gone wrong with the participant's computer?
By Dave - 8/4/2014

Yes, that's possible. We've seen some double-executions in the past, although (a) those instances were rare / didn't occur systematically and (b) we haven't heard of any for a while. See e.g. https://www.millisecond.com/forums/Topic12943.aspx for an example. Is your web experiment using the latest Inquisit web version or an older one? If the latter, you might want to consider using the latest one instead.

Hope this helps.
By seandr - 8/4/2014

Given that there was a 7 minute lag between the first and second data files, my guess is that the participant mistakenly ran the test twice. This could happen, for example, if they started the test, then started it again as the first was loading, then completed the first test, and then saw the second test running and completed that one as well.

There was a change made to 4.0.6 that prevents Inquisit from opening two sessions at once and that should prevent this specific scenario from happening. But of course, the subject could (out of confusion) always return to the start page and launch the test again after the first was complete.

-Sean