Contrib: Gradebook2
  1. Contrib: Gradebook2
  2. GRBK-689

CCFIT: Scantron/Clicker Import scores must ask for max points possible and translate the scores into %'s for % gradebooks

    Details

    • Type: Bug Bug
    • Status: Closed Closed
    • Priority: Major Major
    • Resolution: Fixed
    • Affects Version/s: 1.8.0-a4
    • Fix Version/s: Sprint, 1.8.0-a8, 1.8.0
    • Component/s: GWT Client
    • Labels:
      None
    • QA Status:
      PASS
    • Previous Issue Keys:

      Description

      Scantron and Clicker Import scores are # right of X # of questions
      This is fine for points gradebooks, but % and Letter Grade gradebooks expect the imported value to be a % or Letter Grade.

      When importing into a % gradebook, then the import process needs to
      1) determine the maximum points (=number of questions) either by asking the user, or by allowing the user to provide the points in the import preview spreadsheet (default is currently 100 there) before the following calculation occurs,
      2) calculate the percentage grade by dividing the imported score by the input maximum points value.
      3) proceed with the import process for actual percentage values.

      When importing into a Letter Grades gradebook, then the import process needs to
      1) determine the maximum points (=number of questions) either by asking the user, or by allowing the user to provide the points in the import preview spreadsheet (default is currently 100 there) before the following calculation occurs,
      2) calculate the percentage grade by dividing the imported score by the input maximum points value.
      3) look up the letter grade that corresponds to the calculated percent score
      4) proceed with the import process for actual letter grade values.

        Issue Links

          Activity

          Hide
          Michael Wenk added a comment -
          I do not understand what is asked of here.
          Show
          Michael Wenk added a comment - I do not understand what is asked of here.
          Hide
          Michael Wenk added a comment -
          So after talking to Dave, it sounds like what we want to do is force the user to enter max points for all scantron gradebooks.

          Depending how whether we use this as part of GRBK-752, we could force the user to enter max points on all assignments when we don't know the # of points, basically all non structure imports including scantron, clicker, etc.

          So we need to know whether this is part of GRBK-752 or not.
          Show
          Michael Wenk added a comment - So after talking to Dave, it sounds like what we want to do is force the user to enter max points for all scantron gradebooks. Depending how whether we use this as part of GRBK-752 , we could force the user to enter max points on all assignments when we don't know the # of points, basically all non structure imports including scantron, clicker, etc. So we need to know whether this is part of GRBK-752 or not.
          Hide
          Joyce Johnstone added a comment -
          Decision from 3/1/11: this will not be needed and will be part of GRBK-752.
          Show
          Joyce Johnstone added a comment - Decision from 3/1/11: this will not be needed and will be part of GRBK-752 .
          Hide
          Thomas Amsler added a comment - - edited
          Cannot estimate at this point: We still need to know what to do with scantron/clicker scores that are imported into a letter grade gradebook.
          Show
          Thomas Amsler added a comment - - edited Cannot estimate at this point: We still need to know what to do with scantron/clicker scores that are imported into a letter grade gradebook.
          Hide
          Joyce Johnstone added a comment -
          3/9/11: Mike will send questions to GRBK-752 (KIrk, Steve) Team about this ticket.

          Show
          Joyce Johnstone added a comment - 3/9/11: Mike will send questions to GRBK-752 (KIrk, Steve) Team about this ticket.
          Hide
          Thomas Amsler added a comment -
          I think this is done. See GRBK-1056, GRBK-1104
          Show
          Thomas Amsler added a comment - I think this is done. See GRBK-1056 , GRBK-1104
          Hide
          Thomas Amsler added a comment -
          This is fixed for Scantron bot not clicker type imports.
          Show
          Thomas Amsler added a comment - This is fixed for Scantron bot not clicker type imports.
          Hide
          Fernando Socorro added a comment - - edited
          Attached file is pristine. I was able to import it into prod site 7574c648-7349-4d83-911e-dbe409bef4e4. It added item called Chem 8Fall_10 26 2009:Possible(3) [100] points, but also imported "First" and "Final" as items, too.
          Show
          Fernando Socorro added a comment - - edited Attached file is pristine. I was able to import it into prod site 7574c648-7349-4d83-911e-dbe409bef4e4. It added item called Chem 8Fall_10 26 2009:Possible(3) [100] points, but also imported "First" and "Final" as items, too.
          Hide
          Fernando Socorro added a comment - - edited
          1. Imported both fileapril13.xls & fileapril13.xls (GRBK-389) and both worked, but, like with the previous file, it included "First" and "PRS Total" as items. I overwrote the fhe first ones, so there is no dup of these two "items."

          2. Also, going back to Mike Wenk's point (2nd comment) that "we could force the user to enter max points on all assignments when we don't know the # of points," I think it is a valid point, especially with clicker files, since the points is in () and the user can make the assumption that the gb understands that format, which is incorrect. Plus, as I have argued plenty, the 100 pt default should be removed.

          Just sayin'.
          Show
          Fernando Socorro added a comment - - edited 1. Imported both fileapril13.xls & fileapril13.xls ( GRBK-389 ) and both worked, but, like with the previous file, it included "First" and "PRS Total" as items. I overwrote the fhe first ones, so there is no dup of these two "items." 2. Also, going back to Mike Wenk's point (2nd comment) that "we could force the user to enter max points on all assignments when we don't know the # of points," I think it is a valid point, especially with clicker files, since the points is in () and the user can make the assumption that the gb understands that format, which is incorrect. Plus, as I have argued plenty, the 100 pt default should be removed. Just sayin'.
          Hide
          Jon Gorrono added a comment -
          So in that file the 5 columns I see headers are:
          Name | First | Student ID| PRS Total | CHEM 8FALL_10 26 2009:Possible (3)

          What is the 'First' column?

          Unfortunate name, that 'First', since that could be an assigment name on any other import

          Also, what is the native format? CSV or XLS? or is there an option?

          Show
          Jon Gorrono added a comment - So in that file the 5 columns I see headers are: Name | First | Student ID| PRS Total | CHEM 8FALL_10 26 2009:Possible (3) What is the 'First' column? Unfortunate name, that 'First', since that could be an assigment name on any other import Also, what is the native format? CSV or XLS? or is there an option?
          Hide
          Jon Gorrono added a comment -
          Show
          Jon Gorrono added a comment - I get an error on prod for this file in https://smartsite.ucdavis.edu/xsl-portal/site/319258e7-c2ee-4eab-80c8-65365458c345
          Hide
          Jon Gorrono added a comment -
          The error in that site is due to the gb being in percentage mode... if I delete the first two columns it imports, or if I change to a points gb, it also imports

          But the important thing to notice is that successful imports are happending because it looks like a no-structure import and so PRS Total and First are treated as new assignments.

          This is a real mess, we have to remap all the decision trees for this process... but for now I will try to convince gb that the clicker file is a scantron, which it almost is, and after that the max points popup will work if importing into percentage gb's

          Sound reasonable?
          Show
          Jon Gorrono added a comment - The error in that site is due to the gb being in percentage mode... if I delete the first two columns it imports, or if I change to a points gb, it also imports But the important thing to notice is that successful imports are happending because it looks like a no-structure import and so PRS Total and First are treated as new assignments. This is a real mess, we have to remap all the decision trees for this process... but for now I will try to convince gb that the clicker file is a scantron, which it almost is, and after that the max points popup will work if importing into percentage gb's Sound reasonable?
          Hide
          Fernando Socorro added a comment -
          Name is the last name. First is the, well, you get it. The native file format is CSV, but for testing someone (cannot remember who) wanted both. Re the site you had issues with, I am only a participant, and do not have admin, so I cannot test unless my role is changed.
          Show
          Fernando Socorro added a comment - Name is the last name. First is the, well, you get it. The native file format is CSV, but for testing someone (cannot remember who) wanted both. Re the site you had issues with, I am only a participant, and do not have admin, so I cannot test unless my role is changed.
          Hide
          Jon Gorrono added a comment -
          - added injection of headers and tests for headers for clicker format
          - overloaded code for scantron handling to treat clicker files by adding tests for clicker formats
          - removed 'failed key' item id prefix for scantron/clickers after values have been adjusted in importpanel
          - added handling for numberformatexceptions & NPE's in same block in importpanel
          - created a more abstract delegate for methods to get Max and Min scores from file
          - changed the wording in properties to not refer to scantrons
          Show
          Jon Gorrono added a comment - - added injection of headers and tests for headers for clicker format - overloaded code for scantron handling to treat clicker files by adding tests for clicker formats - removed 'failed key' item id prefix for scantron/clickers after values have been adjusted in importpanel - added handling for numberformatexceptions & NPE's in same block in importpanel - created a more abstract delegate for methods to get Max and Min scores from file - changed the wording in properties to not refer to scantrons
          Hide
          Jon Gorrono added a comment -
          Show
          Jon Gorrono added a comment - codereview: http://codereview.appspot.com/5618043
          Hide
          Jon Gorrono added a comment -
          important note: this currently only handles CSV native format clicker files... so if someone saves one as an excel file, it will not work correctly..... discuss. :)
          Show
          Jon Gorrono added a comment - important note: this currently only handles CSV native format clicker files... so if someone saves one as an excel file, it will not work correctly..... discuss. :)
          Hide
          Jon Gorrono added a comment -
          done.

            - jexlcel searches are case-sensitive, updated injected column names to reflect proper case
            - some i18n
            - set the import file as a scantron if it is a clicker to overload scantron code
          Show
          Jon Gorrono added a comment - done.   - jexlcel searches are case-sensitive, updated injected column names to reflect proper case   - some i18n   - set the import file as a scantron if it is a clicker to overload scantron code
          Hide
          Peter Peterson added a comment -
          From what I read here it doesn't sound like this is fixed. CSV works, but XLS will fail, plus I see many open comments and no, "this is fixed, lets test it" comments. Request review and possible reopen.
          Show
          Peter Peterson added a comment - From what I read here it doesn't sound like this is fixed. CSV works, but XLS will fail, plus I see many open comments and no, "this is fixed, lets test it" comments. Request review and possible reopen.
          Hide
          Joyce Johnstone added a comment -
          Jon assigning back to you per Pete's comments. Please review.
          Show
          Joyce Johnstone added a comment - Jon assigning back to you per Pete's comments. Please review.
          Hide
          Jon Gorrono added a comment -
          if you open the subversion commits tab you'll see when changes have been made to code between comments... after the comment that excel files don't work, I committed a fix for it (on the 10th) and then I posted a comments that pretty categorically announces 'done' ... like my dad used to say ....'Are you waiting for a gold-engraved invitation?' ;)

          Did you find problems or is this triage?
          Show
          Jon Gorrono added a comment - if you open the subversion commits tab you'll see when changes have been made to code between comments... after the comment that excel files don't work, I committed a fix for it (on the 10th) and then I posted a comments that pretty categorically announces 'done' ... like my dad used to say ....'Are you waiting for a gold-engraved invitation?' ;) Did you find problems or is this triage?
          Hide
          Peter Peterson added a comment -
          Ok, after vetting, renaming the header and clarification this is ready for testing.
          NOTE: in the future we agreed we should all strive for simple and easy to understand "Subjects Headers" and if time permits "Comments".
          Show
          Peter Peterson added a comment - Ok, after vetting, renaming the header and clarification this is ready for testing. NOTE: in the future we agreed we should all strive for simple and easy to understand "Subjects Headers" and if time permits "Comments".
          Hide
          Peter Peterson added a comment -
          Assigned to Fernando for QAing
          Show
          Peter Peterson added a comment - Assigned to Fernando for QAing
          Hide
          Fernando Socorro added a comment -
          QA Pass
          The percent conversion issue
          1. The import wizard notices the scores in the file are not 0-100 shows the highest and lowest scores, and sets the highest score as the default "points" value for the item.
          2. The grader has the option of changing the points. After the points are set, the grades in the spreadsheet reflect the desired percentages correctly. Neat, btw.

          File options: file contains
          1. The grader gets the option of choosing "clicker" in the dropdown menu
          2. In the next frame, only the item name and scores appear.

          Should this be retested when we get the 2.8 on test?
          Show
          Fernando Socorro added a comment - QA Pass The percent conversion issue 1. The import wizard notices the scores in the file are not 0-100 shows the highest and lowest scores, and sets the highest score as the default "points" value for the item. 2. The grader has the option of changing the points. After the points are set, the grades in the spreadsheet reflect the desired percentages correctly. Neat, btw. File options: file contains 1. The grader gets the option of choosing "clicker" in the dropdown menu 2. In the next frame, only the item name and scores appear. Should this be retested when we get the 2.8 on test?
          Hide
          Fernando Socorro added a comment -
          Should this be retested on 2.8?
          Show
          Fernando Socorro added a comment - Should this be retested on 2.8?
          Hide
          Peter Peterson added a comment -
          You mean the CLE 2.8? If so then yes we can add this to the regression tests, but this should be closed.
          DAVID: please add to the regression plan
          Show
          Peter Peterson added a comment - You mean the CLE 2.8? If so then yes we can add this to the regression tests, but this should be closed. DAVID: please add to the regression plan
          Hide
          Peter Peterson added a comment -
          Closed per QA ressults from Fernando
          Show
          Peter Peterson added a comment - Closed per QA ressults from Fernando

            People

            • Assignee:
              Unassigned
              Reporter:
              Fernando Socorro
            • Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: