From: SourceForge.net <no...@so...> - 2010-01-22 13:00:23
|
Bugs item #1027980, was opened at 2004-09-14 17:07 Message generated for change (Comment added) made by liedekef You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=108956&aid=1027980&group_id=8956 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Admin Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Nobody/Anonymous (nobody) Summary: Zero value in choice_id on required field Initial Comment: My survey has a large number of 'response_single' replies which are marked as 'required'. The survey handler treats this perfectly - i.e. respondents are blocked from passing onto next question before the previous has been answered. When reviewing the database I notice that approx 40 out of 6000 rows in the 'response_single' table have a zero-value (0) in it for the field 'choice_id'. Since all the underlying questions are "required" I cannot understand how that should be possible. I have tried very hard to provoke this error without luck. Mostly the surveys are run with permission to save and resume but even when I have turned that off I see the problem occassionally. I enclose a dump of the 'response_single' table. Thanks, Sren Voigt ---------------------------------------------------------------------- >Comment By: Franky Van Liedekerke (liedekef) Date: 2010-01-22 14:00 Message: way out of date and should be fixed in any 2.x version ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2004-09-16 13:58 Message: Logged In: NO Well you are right for all of the cases exept one. When a respondent exits before the survey is completed, that creates zero responses which is obviously quite sensible. And admitted - that does bring down the number of problem responses to one. Still this one bugs me. My sections / pages have two (r_single) questions and a (r_text) question each. Relative to the r_single-table a respondent should never have more than two zero responses. >From the file I include the respondent_id 124 has four zero responses. response_id question_id choice_id 124 2666 0 124 2665 0 124 2669 6505 124 2670 6510 124 2673 6513 124 2674 6518 124 2677 6523 124 2678 6528 124 2681 6533 124 2682 6538 124 2685 6545 124 2686 6550 124 2689 6555 124 2690 6560 124 2694 6570 124 2693 6565 124 2698 6578 124 2697 6573 124 2702 6590 124 2701 6585 124 2706 6598 124 2705 6593 124 2710 6608 124 2709 6603 124 2714 6620 124 2713 6615 124 2717 6625 124 2718 6630 124 2722 6640 124 2721 6635 124 2726 6650 124 2725 6645 124 2730 6658 124 2729 6653 124 2734 6670 124 2733 6664 124 2738 6680 124 2737 6675 124 2741 6683 124 2742 6688 124 2746 6698 124 2745 6693 124 2750 6708 124 2749 6703 124 2754 6718 124 2753 6713 124 2758 6728 124 2757 6723 124 2762 6739 124 2761 6734 124 2767 0 124 2765 0 In the case of respondent_id 124 she has provided zero- responses for section 1 but has perfectly valid responses until she exits before the last page. I cannot understand how this respondent was allowed to go beyond page 1 without providing answers. And yes - I have tested that phpESP really does require input on this first page. I have tried to exit and come back but seemingly phpESP treats this correctly - I am coming back on the correct first page. ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2004-09-14 17:43 Message: Logged In: NO Are you sure that the responses w/ zeros were completed? select id from response where complete = 'Y'; -James ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=108956&aid=1027980&group_id=8956 |