|
|
|
Forum Newbie
Group: Forum Members
Last Login: 2/17/2008 5:45:58 PM
Posts: 3,
Visits: 13
|
|
Hi,
I have just finished programming my first experiment. It is run on a lenovo C100 laptop, 512 RAM, Intel M processor 1.86 GHz..
It is mostly composed of simple key press to some pictures (up loaded through a list bmp file).
There is NO time limit.
When I run it quickly there is no problem, but when a first subject did it on her own time, more slowly (which took about 20 minutes for about 150 pics)-
I got the Clock.SystemTimeDrift warning by the end of the experiment and all data was lost.
I run it a few more times slower by myself (taking some time before responding to each picture) and in each case the data was lost.
I don't mind not getting accurate RT, I am interested in the responses.
I also tried installing Directx 9.0 as I read it somewhere here but it didn't helped also.
Pleas help
|
|
|
|
|
Forum MVP
Group: Administrators
Last Login: Yesterday @ 1:09:20 PM
Posts: 550,
Visits: 1,206
|
|
| Get the latest copy of the RefreshClockTest experiment from http://www.pstnet.com/products/e-prime/timing and run the 25 minute test. The edat (not the .out file) will have a number of additional information about the drift test on that machine. BTW, you can use E-Recovery to recover data files from a session that may have aborted. -Brandon
|
|
|
|
|
Forum Newbie
Group: Forum Members
Last Login: 2/17/2008 5:45:58 PM
Posts: 3,
Visits: 13
|
|
| Thank you Brandon. I managed to get the data I wanted- RESP, ACC etc- just not any time based data. But right now the time issue is not that critical. I did run the time test, tough I'm not sure it was the 25 minute one, and it turned out fine, which is strange in my opinion. I still find it weird that some experiment runs are getting the problem and some not, by for now I will focus on the results I'm getting. Thanks again. Shani
|
|
|
|
|
Forum MVP
Group: Administrators
Last Login: Yesterday @ 1:09:20 PM
Posts: 550,
Visits: 1,206
|
|
| The clock drift test is based on a comparison against your date/time clock which is known to be inaccurate, which is why the test can yield a number of false positives. Besides the likely case that the date/time clock is the issue, if your computer is on a network your a likely sync your date/time clock with time.windows.com or time.org, etc which can cause a false positive. Typically, if your machine has a bad timing chip, the results normally will be very drastic and will show up every time on a 25 minute run of Refresh Clock Test. If it is shown that the date/time clock is the issue, you can change the threshold value for the check which can cause it to essentially be disabled so you do not lose data on runs. -Brandon
|
|
|
|
