Table 1.

Differences between the three treatment groups when performing the post hoc analysis using a multilevel model on the forecast interval widths

	Control vs. one-briefing	Control vs. two-briefing	One-briefing vs. two-briefing
Including outliers	P = 0.002 (df = 88, t = 3.16)	P < 0.001 (df = 86, t = 5.16)	P = 0.045 (df = 86, t = 2.03)
Excluding outliers	P < 0.001 (df = 88, t = 3.41)	P < 0.001 (df = 86, t = 5.57)	P = 0.030 (df = 86, t = 2.20)

df, degrees of freedom.