===================================================== == The dataset file is 'german_language_supsi.csv' == ===================================================== - The file has 452 rows. - The first row is a header. - The 451 remaining rows correspond to the answers to the test as provided by different students (one for each row). - The first 95 columns (headers Y1,Y2,...,Y95) correspond to the 95 questions in the test - The last 4 columns (headers X1,X2,X3,X4) correspond to the 4 skills to be assessed by the test - Values 0/1 for the questions correspond to wrong/right answers to the questions - Values 0/1/2/3 for the skills correspond to levels A1/A2/B1/B2 ================================================ == Info about the questions are in 'keys.csv' == ================================================ - The file has 96 rows. - The first two is a header. - The 95 remaining rows correspond to the 95 questions in the test - The first column is the name of the question (Y1,...,Y95) - The second column is the type of question (easy,medium,difficult) - The third column is the skill associated to the question (X1,X2,X3,X4) The levels of the students are obtained from the percentage P of correct answers computed separately for each skill. The conversion levels from P to the actual lavel of skill X are as follows: - if P <= .35 then X=0 - if .35 < P <= .55 then X=1 - if .55 < P <= .75 then X=2 - if P > .75 then X=3