Table 4.

Multiple Logistic Regression Using All Speech Features Jointly to Predict MCI Status.

	β	Wald	Exp(β)	95% Confidence Interval for Exp(β)
				Lower Bound	Upper Bound
Intercept	81.45	1.33
Total words	0.00	0.00	1.00	0.98	1.02
Filler words	6.99	4.28	1089.78	1.45	821255.69
Empty words	−3.38	1.60	0.03	0.00	6.37
Definite articles	4.88	4.01	131.37	1.11	15572.28
Indefinite articles	3.43	1.95	30.92	0.25	3841.58
Pronouns	−0.01	0.00	0.99	0.13	7.28
Nouns	−2.70	4.30	0.07	0.01	0.86
Verbs	2.44	4.64	11.42	1.25	104.65
Determiners	−2.83	1.45	0.06	0.00	5.88
Content words	0.11	0.01	1.11	0.19	6.64
Lexical frequency	−0.78	0.41	0.46	0.04	5.05
Type-token ratio	−48.52	0.90	8.49 x 10−22	3.00 x 10−65	2.40 x 10+22
Honoré’s statistic	−0.55	1.78	0.58	0.26	1.30
Brunet’s index	−5.17	1.81	0.01	3.03 x 10−06	10.65
Speech rate	1.32	1.28	3.74	0.38	36.66
Filler rate	−45.83	2.53	1.25 x 10−20	3.68 x 10−45	42149.03