Table 2.

The impact of different task modifications on mean-level model-based estimates and its reliability.

	N	MBI Score		Split-half [95%CI]^a	t/F
	N	M(SD)	Range	Split-half [95%CI]^a	t/F
Overall Task					n/a
	5005	0.26 (0.33)	[−0.49–2.26]	0.65 [0.63−0.67]
Transition Ratio (Between-subject)^b					8.40^***
80:20	2138	0.30 (0.34)	[−0.47–2.03]	0.67 [0.64−0.69]
70:30	2138	0.22 (0.31)	[−0.45–2.16]	0.63 [0.61−0.66]
Difficulty/Order (Within-subject)					20.16^***
Easy/1^st Block	5005	0.30 (0.32)	[−0.49–1.90]	0.71 [0.71−0.76]
Medium/2^nd Block	5005	0.22 (0.29)	[−0.53–1.92]	0.68 [0.66−0.70]
Reward Drift Set (Between-subject)					−5.33^***
Drift A/1^st Block	2395	0.28 (0.32)	[−0.46–1.84]	0.74 [0.71−0.76]
Drift B/1^st Block	2610	0.33 (0.33)	[−0.51–1.93]	0.69 [0.66−0.71]

^aData was split using odds-evens method, reporting Pearson R.

^bSubsample of total dataset to achieve two age, gender and education-matched samples.

***p < 0.001.

MBI Model-based Index, M Mean, SD Standard Deviation, CI Confidence Interval.