Assessment: Preliminary item analysis data for statistics and physical science test one

MS 150 Statistics

Test one this term was substantively similar to test one last term permitting a term-on-term item analysis of student performance (fall 2008 versus spring 2009). Overall there was no change in performance as measured by an item analysis. There were, however, changes in performance within specific topic areas. The number correct (corr) refers to the number of students getting that item correct. The fall term had 45 students taking test one, this term the course has 57 students who took test one.

I	s	q	Spring 2009	corr	perc		Fall 2008 number correct	perc	change
t01	2	1	level of measurement	33	0.58		25	0.56	0.02
t01	2	2	sample size	54	0.95		40	0.89	0.06
t01	2	3	minimum	57	1.00		43	0.96	0.04
t01	2	4	maximum	57	1.00		44	0.98	0.02
t01	2	5	range	56	0.98		40	0.89	0.09
t01	2	6	midrange	40	0.70		36	0.80	-0.10
t01	2	7	mode	53	0.93		41	0.91	0.02
t01	2	8	median	44	0.77		36	0.80	-0.03
t01	2	9	mean	45	0.79		33	0.73	0.06
t01	2	10	standard deviation	43	0.75		35	0.78	-0.02
t01	2	11	coefficient of variation	38	0.67		30	0.67	0.00
t01	2	12	width	40	0.70		36	0.80	-0.10
t01	2	13	frequency table	21	0.37		28	0.62	-0.25
t01	2	14	frequency histogram	20	0.35		26	0.58	-0.23
t01	2	15	shape of histogram	38	0.67		28	0.62	0.04
t01	2	16	inference	21	0.37		7	0.16	0.21
t01	2	17	z-score	20	0.35		13	0.29	0.06
t01	2	18	inference	29	0.51		17	0.38	0.13

Performance fell markedly and significantly on material from section 2.2 of the text, frequency tables and histograms. As a result, this material was reviewed in detail with the students. This material will be retested on the midterm.

The overall success rate of 69% for the spring term is identical to the success rate for the fall term and falls well within the historical range of 65% to 75%.

SC 130 Physical Science

Although the tests are not substantively similar term-on-term from fall 2008 to spring 2009, there was overlap in material for test one in the two terms. On overlapping material there was a modest overall gain of 6% as measured by item analysis. Thirty-three students took test one spring 2009, in fall 2008 twenty-eight students took test one.

q	Spring 2009	corr	perc		Fall 2008 correct	perc	perc
1	plot data	33	1.00
2	inference from graph	29	0.88
3	calculate velocity	16	0.48
4	infer acceleration	0	0.00
5	peripatetic	20	0.61		20	0.71	-0.11
6	list three fundamental qualities	32	0.97
7	list qualities required for motion	29	0.88		24	0.86	0.02
8	calculate volume	27	0.82		23	0.82	0
9	calculate density	20	0.61		16	0.57	0.03
10	inference from result	22	0.67		12	0.43	0.24
11	inference from graph	25	0.76		13	0.46	0.29
12	inference from graph	8	0.24		9	0.32	-0.08
13	write an equation from graph	17	0.52		16	0.57	-0.06
14	given x determine y from quadratic	6	0.18
15	calculate mean	27	0.82		11	0.39	0.43
16	identify outlier error	25	0.76		26	0.93	-0.17

An encouraging gain was seen in the ability of students to make an inference from material provided on the test. This term marks the first term that the course is working solely from the physical science text I worked on during 2008. The two percent gain in overall performance as measured by the item analysis from a 62% success rate to a 64% success rate is not significant and yet is encouraging.

The specific performance by question on test one provides guidance on areas of student weakness that will be addressed as the course moves forward. Of glaring note is that no student could make the inferential leap that a constant velocity implies a zero acceleration (question number four). This matter was reviewed when the test was returned to the students.