Arkansas site leader Gay Stewart and future teachers discuss the design of an assessment.

 

Contact Us

University of Arkansas Project Report 2007

Assessment

Here we present a brief summary of the assessment statistics for the calculus-based physics sequence, where most of our future teachers are recruited. Additional evaluation data on these and the other PhysTEC courses can be found under the links in Part II, Section 3, Course Reform.

In all cases, the normalized (Hake) gain g is calculated as the average over the gains for individual students. The gain for each student is the percentage of the improvement it was possible to make that was made:

g= 100 × (Post-test score - Pre-test score)/(total possible - Pre-test score)

In the tables below, N is the number of students that had both a pre- and post-test for calculation of the normalized (Hake) gain.

Calculus-based Introductory Mechanics, FCI‡ (30 possible)
Semester	N	Pre-test average	Post-test average	Average normalized gain, g	Standard error of g
Fall, 2001	133	11.6	19.1	42.5	27.3
Spring, 2002	105	7.4	19.6	54.0	21.7
Fall, 2002	138	7.8	19.7	53.2	24.5
Spring, 2003	145	12.2	18.9	40.7	25.7
Fall, 2003	179	12.6	19.3	41.1	26.5
Spring, 2004	157	12.7	20.1	43.7	26.9
Fall, 2004	162	11.7	18.9	41.8	26.0
Spring, 2005	196	12.7	24.0	63.6	21.9
Fall, 2005	163	12.8	21.1	46.7	30.2
Spring, 2006	170	10.7	19.5	45.5	19.2
Fall, 2006	186	12.4	21.8	56.0	24.5
Spring, 2007	191	14.0	22.5	56.0	24.0

Materials have been progressively modified to introduce more interactive engagement in the discussion (lecture) sections and to improve the quality of the supporting materials, to ease adoption by non-PhysTEC faculty, or other institutions. (In fall 2005 and spring 2006 the course was taught by a new visiting professor.)

‡David Hestenes, Malcolm Wells, and Gregg Swackhammer, “Force Concept Inventory,” The Physics Teacher, Vol. 30, 147 (1992)

 

Calculus-based Introductory E&M and optics, CSEM† (32 possible)
Semester	Spr-2002 is NSF-CCD baseline	N	Pre-test average	Post-test average	Average normalized gain, g	Standard error of g
Spring, 2002		106	9	17	34	19
Fall, 2002	Introduction of the first set of PhysTEC revisions	74	9	19	42	21
Spring, 2003		90	9	18	42	22
Fall, 2003		106	9	19	43	19
Spring, 2004		112	9	20	45	16
Fall, 2004	Major new revision-- added more math and inquiry	94	9	21	50	20
Spring, 2005		129	9	20	49	20
Fall, 2005		100	9	20	48	18
Spring, 2006		126	10	21	52	20
Fall, 2006**	New software support	97	5	17	45	18
Spring, 2007		129	10	20	50	22

**Fall 2006 was a particularly weak class, coming in with a statistically significantly lower pretest average. Although their absolute gain in conceptual understanding was as large as for any semester, the lower starting point gave them a lower normalized Hake gain.

†NSF Sponsored Conceptual Survey in Electricity in Magnetism, a product of the Two Year College Physics Workshop Project, ÓTYC Physics Workshop Project (1998). Curtis J. Hieggelke and Thomas O'Kuma, co-directors. http:/www.tycphysics.org. Note how the CSEM scores increased over the duration of the PhysTEC project.