. Author manuscript; available in PMC: 2021 Dec 15.

Published in final edited form as: J Neurol Sci. 2020 Nov 1;419:117205. doi: 10.1016/j.jns.2020.117205

Table 2.

Clinimetric properties of the AM item for various exclusions

	Item to Total Correlation	Cronbach’s Alpha if Item Omitted	IRT Discrimination	MAD
SITE RATINGS
Original Sample (n = 208 patients)	0.011	Increase	0.09	N/A

Sufficient Duration and Severity (n = 145 patients)	0.153	Decrease	0.341	N/A

Sites with Greater Than 10 Patients (n = 137 patients)	0.182	Decrease	0.409	N/A

Reverse Scoring Omitted (n = 127 patients)	0.251	Decrease	0.467	N/A

VIDEO RATINGS
Video Ratings with 5 Raters (n = 80 patients, 400 ratings)	0.231	No Change	0.718	0.91

Sufficient Duration and Severity (n = 60 patients, 300 ratings)	0.405	Decrease	1.163	0.75

Reverse Scoring Omitted (n = 60 patients, 273 ratings)	0.444	Decrease	1.29	0.68

Clinimetric properties of the AM item, before and after excluding patienls with insufficient severity and duration of AM trial and raters using reverse scoring (IRT = item response theory; MAD = mean absolute difference).