The Effects of Reusing Written Test Items: A Study Using the Rasch Model

Table 4

Results of reliability analysis.

Day	Reliability
	𝑛^∗p	𝑛i	𝑅	𝑅^∗	𝑅30
1-cw20	95	19 (4/-/15)	0.27^1	0.36	0.67^1
2-cw21	183	14 (4/5/5)	0.39^1	0.29	0.55^1
3-cw22	67	14 (4/5/5)	0.55^1	0.57	0.80^1
4-cw23	188	15 (4/7/4)	0.32^1	0.28	0.51^1
Total	535	33	0.39^2	—	0.80^2

cw: calendar week 2009; 𝑛^∗_p: number of persons taking test as a pretest; 𝑛_i: number of items in analysis (published/reused/new); 𝑅: reliability; 𝑅^∗: Cronbach’s alpha/day without previously published items; 𝑅₃₀: reliability adjusted for test length (30 items).
¹𝛼: Cronbach’s alpha (is listed complementary for illustrative reasons).
²rel_ps: reliability of person separation.