If a script causes an error, TC will either terminate the test run or just go on, depending on the "stop on error" option. In both cases, trying to find the cause of the error is very time consuming...
I have my TC setup with a 'Pause on Error' set to true so if an error does occur, it pauses the program exactly when the error occurred, then I can check my variables and the screen's status, and if need be, do a step by step from then on or correct the values on my screen in order to re-continue the testing. This works fine unless the error caused is not fixable at that point due to something not done properly before the error occurred.
A better solution would be that when an error occurs and the 'Pause on Error' is activated, then if you are able to go back a few steps, possibly even change the script (which currently you can't in Pause mode) and/or change the screens data as well (this option may not be reversible), and continue testing from that previous spot. That will be much better as you can re-adjust your code 'on the fly' and continue the testing process without the need to re-start the whole test again.
For now, a 'Pause on Error' should do the trick for you.