Wednesday, April 27, 2016

Student Mark Scanners

In the beginning all student mark data were entered manually. Then card readers came. Then optical scanners. Now student mark data are collected on the Internet. Not having an answer sheet or card changes the testing.

Now a student must answer each question as it is presented. There is a loss of control. There is no time to think about it and to return to make a choice. There are no balls in the game; only strikes and outs. It speeds up the game but is another way of reducing judgment.



The student with the quickest response wins. Time for use of higher levels of thinking is reduced. Online and paper tests are measuring different things and at different levels of thinking using forced-choice traditional scoring.





Wednesday, April 20, 2016

Multiple-Choice Rosetta Stone

The process of setting standardized test scores in the past has been as much politics and money as it has been effective assessment for student, teacher and test development. Power Up Plus contained features that managed cheating as well as compared classroom designed tests with Item response theory (IRT) tests used by state departments of education (where cheating could be as common as in the classroom).



A number of these features were of interest in auditing the Rasch IRT model that took a very liberal view of student mark data. It provided the basis for several states to obtain amazing annual yearly improvement results. 



Wednesday, April 13, 2016

Marking and the Effective Teacher

Marking, scoring, and grading are as simple for knowledge and judgment scoring as for traditional right mark scoring with one big difference: guessing is not required. If a student choose to mark all answers, the test automatically returned to traditional forced-choice testing when Power Up Plus was used.


Wednesday, April 6, 2016

Multiple-Choice Test Scoring Methods

Valid cut scores have been set from over 70% to 35% on standardized tests. The traditional DUMB scores of A, 90%; B, 80%; C, 70%, D, 60%, and F, 50% and below are easily achieved by adjusting question difficulty. This common practice is rarely questioned. The main interest is the number of right marks.




On the other hand, any set of questions can be used, scored, and cut scores set from a classroom friendly item analysis. Now student quality, as well as quantity, is of interest. The value of right marks is independent from the number of right marks (all 20 marks out of 30 is 100% right; 67% right count; no wrong marks. This student has a solid basis for further learning by whatever means of instruction). Neither the student nor the teacher needs to guess about what has yet to be learned.