Spring 2006. Kyle Forinash.
For further information, Contact: Kyle Forinash
Abstract:
A standardized pre and post test (the FCI) has been given to all introductory physics classes at IUS since 1999 (in addition to a substantial amount of graded quizzes, tests, group problem work, lab work and other assessment tools). The gain scores (percentage increase between the pre and post test) on content knowledge eliminate factors such as the initial knowledge level of the students or the popularity of the instructor and thus clearly assesses the teaching effectiveness of an individual instructor using a given pedagogical method. Nationally this test has been used to establish definitively that interactive teaching methods in physics are superior to a traditional lecture approach and that some of these interactive methods achieve positive results which are independent of the instructor (references on request).
The Force Concept Inventory test (FCI) was developed in the early 1990s to diagnose the conceptual understanding and critical thinking of students in introductory level physics classes. The test has been revised several times, validated by numerous groups and used extensively in the past few years at many different schools. A 6000 student meta-analysis was done by R. Hake at IU Bloomington in 1998 validating the test. The test is very difficult; a score of 87% is considered 'mastery' (PhD level understanding), 60% is considered the threshold for understanding Newtonian mechanics. National averages run from 25% to 70% for the pre test and 40% to 85% for the post test depending on the school.
Results of this and other pre and post tests are sometimes reported as the fraction of maximum possible gain realized given by G = (S f -S i )/(100-S i ) where S f and S i are the pre and post test scores as percentages. The national average G score is 0.25 for traditionally taught (lecture only) courses, regardless of the initial knowledge base of the students . Courses taught using newer, more interactive methods (collaborative learning, interactive lecture- demonstration methods, group problem solving, etc.) typically get higher gain scores (0.30< G< 0.68; gain scores above 0.70 are extremely rare), regardless of the initial knowledge base of the students . Gain scores are independent of the initial student scores but instead depend on the method being used to teach the class. Gain scores at IUS have ranged from 0.25 to 0.60 depending on the instructor and the method of instruction.