Skip to main content

Table 3 Probability that the number of successes (performance differences between post- and pre-assessment > critical differences) is due to chance alone given a probability of success for each individual comparison of p = 0.01 and the total number of comparisons performed in each group.

From: Specific effects of EEG based neurofeedback training on memory functions in post-stroke victims

(A) Probability that chance alone is responsible for cognitive improvements after training and total number of comparisons performed in each group

 

Alertness

Divided Attention

Cogn. Flexibility

Inhibition

WM

STM

LTM

SMR patients

0.38 (22)

0.65 (44)

0.42 (33)

0.86 (33)

0.17 (22)

1.64 × 10−5 (66)*

2.66 × 10−15 (99)*

UA patients

0.72 (12)

0.21 (24)

0.27 (18)

0.55 (18)

0.03 (12)

1.25 × 10−4 (36)*

1.11 × 10−12 (54)*

TAU patients

0.42 (14)

0.95 (28)

 

0.64 (21)

0.04 (14)

0.06 (28)

0.02 (35)

SMR CG

0.97 (32)

0.89 (64)

0.87 (48)

0.99 (48)

0.003 (32)*

3.50 × 10−8 (96)*

3.55 × 10−15 (144)*

UA CG

0.20 (48)

0.05 (96)

0.18 (72)

0.86 (72)

4.35 × 10−5 (48)*

3.12 × 10−11 (144)*

<1 × 10−35 (216)*

(B) Probability that chance alone is responsible for cognitive declines after training and total number of comparisons performed in each group

SMR patients

0.66 (22)

0.27 (44)

1.00 (33)

0.23 (33)

0.90 (22)

0.002 (66)

0.94 (99)

UA patients

0.34 (12)

0.92 (24)

0.85 (18)

0.85 (18)

0.72 (12)

0.06 (36)

0.04 (54)

TAU patients

0.77 (14)

0.79 (28)

 

0.89 (21)

0.77 (14)

0.14 (28)

0.69 (35)

SMR CG

0.84 (32)

0.96 (64)

1.00 (48)

0.87 (48)

0.40 (32)

0.05 (96)

0.37 (144)

UA CG

0.35 (48)

0.36 (96)

0.99 (72)

0.99 (72)

0.53 (48)

0.0005 (144)*

<1 × 10−35 (216)*

  1. Probability that chance alone is responsible for improvements (A) and declines (B) in cognitive performance after training, presented separately for stroke patients and healthy controls for the different cognitive constructs (short-term STM, long-term LTM, and working memory WM). Uncorrected p-values, which were significant after correction for multiple comparisons employing false discovery rates [48], are marked with asterisks.