home | email  


Exam Entry Parser for Print-Base Courses

There were four tables that have been used with Exam Entry Tool. They were

ExamMasterInfo(ExamId, ExamTitle, RenderSurl, QuestionCnt, QuestionPres, TotalPoints, PassPercent, RandomPres, TakeAllow, SkipAllow, SkipAllow, TimeAllow, CaseSense, ReportResults, ReportGrade, ShowRatUncond, InstPerQtnGrp, HeadText, TailText, ProctorEnab);

ExamInstGrpInfo(ExamId, InstId, InstTitle, InstText);

ExamQuestionInfo(ExamId, QnAIndex, ExamQnAType, QtnText, RatSurl, DepSurl, InstId);

ExamAnswerInfo(ExamId, QnAIndex, QnAOrder, QnASubOrder, AnsText, CorrectAns);

Fill in each field in these tables based on each exam page. Some fields can be set up with default values for all exams.

In ExamMasterInfo table, all fields can be read in from a file which is created in advanced (that is we need to preprocess the exam files). In this file, we need to give ExamId, ExamTitle, RenderSurl, QuestionCnt, QuestionPres, TotalPoints, PassPercent, RandomPres, TakeAllow, SkipAllow, SkipAllow, TimeAllow, CaseSense, ReportResults, ReportGrade, ShowRatUncond, InstPerQtnGrp, HeadText, TailText, ProctorEnab. If all fields except for ExamId, ExamTitle, QuestionCnt, and QuestionPre are the same for all exams, we can create an interface to input those different fields and automatically generate those same fields by default. Another option (that is perfect) is to read ExamTitle from the file and generate others automatically.

In ExamInstGrpInfo table, we need to pick InstId, InstTitle, and InstText manually or generate them automatically. If we pick them manually, we can either put them in a file in advanced or create an interface to input them.

In ExamQuestionInfo table, ExamQnAType might need to be input manually since it is hard for a computer find which question falls in which exam type just based on the plain text in a file. Or we can create an interface to tell the computer from which question to which question are all the certain types (Fill in black, true/false, etc). Another option is that we manually add question type information at the beginning of each exam file so that program can read them first and assign the type to each question in the follow up. The last option is to read exam question type directly from exam file if each exam uses the same words for question type like Yes/No, etc. and all files keep the same format for this information, for instances, this information always is at the beginning of all the following questions with this type.

In ExamAnswerInfo table, there seem no problem to fill in all the fields in this table since all the field values can be obtain from the above or from the exam file by parsing.

This Parser was not very automatic. Any text format change from file to file caused problems with the database so the designers had to go back to each table and check each record one by one for exam question, answer, and instruction group. If we pull out each question, answer from the exam file and rewrite them into another file with specified format and later on we use the new file as input file to input data in the database, that might be safer, but that also needs more human being interfering. Anyway, the more preprocessing by man power, the more accurate and also the more time consuming.

Eventually, a new exam entry tool was developed. It had a more friendly and efficient user interface. The designers could also use a ‘exam page preview’ after each exam or question was entered into the database. The exam page preview tool became especially important for time, cost and accuracy.


Copyright 2004 - University of Nebraska Board of Regents - All Rights Reserved

home ] contact ]