Table 2 Effect sizes, frequentist statistics and Bayes factors for each preregistered hypothesis

From: A multi-country test of brief reappraisal interventions on emotions during the COVID-19 pandemic

Row number

Hypothesis

B (s.e.m.)

Standard deviation of B by country/region

t statistic (d.f.)

Holm’s adjusted P value

Cohen’s d [95% CI]

log10(BF) [under robustness check]

Verbal interpretation132 of log10(BF)

2

Reappraisal interventions (versus control) would reduce negative emotions in response to the photos (hypothesis 1a).

0.513 (0.021)

0.129

23.973 (52.36)

<0.001

0.392 [0.360, 0.425]

29.41 [29.47]

log(BF) > 2 represents “extreme evidence in favour of HA”;

2 > log(BF) > 1.5 represents “very strong evidence in favour of HA”;

1.5 > log(BF) > 1 represents “strong evidence in favour of HA”;

1 > log(BF) > 0.5 represents “moderate evidence in favour of HA”;

0.5 > log(BF) > −0.5 represents “inconclusive evidence”;

−0.5 > log(BF) > −1 represents “moderate evidence in favour of H0”;

−1 > log(BF) > −1.5 represents “strong evidence in favour of H0”;

−1.5 > log(BF) > −2 represents “very strong evidence in favour of H0”;

−2 > log(BF) represents “extreme evidence in favour of H0”.

3

Reappraisal interventions (versus control) would reduce negative state emotions (hypothesis 1b).

0.185 (0.013)

0.064

14.401 (36.39)

<0.001

0.313 [0.270, 0.357]

15.61 [15.15]

4

Reappraisal interventions (versus control) would reduce negative emotions about the COVID-19 situation (hypothesis 1c).

0.241 (0.019)

0.082

12.570 (30.67)

<0.001

0.239 [0.201, 0.277]

13.26 [12.92]

5

Reappraisal interventions (versus control) would increase positive emotions in response to the photos (hypothesis 2a).

0.711 (0.025)

0.166

28.301 (59.18)

<0.001

0.590 [0.549, 0.631]

34.65 [34.80]

6

Reappraisal interventions (versus control) would increase positive state emotions (hypothesis 2b).

0.178 (0.012)

0.064

14.263 (42.69)

<0.001

0.326 [0.281, 0.372]

15.90 [15.42]

7

Reappraisal interventions (versus control) would increase positive emotions about the COVID-19 situation (hypothesis 2c).

0.263 (0.018)

0.070

14.809 (31.21)

<0.001

0.266 [0.230, 0.301]

15.48 [15.23]

8

Reconstrual would lead to greater decreases in negative emotional responses in response to the photos than repurposing (hypothesis 3a).

−0.056 (0.023)

0.107

−2.438 (33.48)

0.041

−0.043 [−0.078, −0.008]

0.25 [−0.47]

9

Reconstrual would lead to greater decreases in negative state emotions than repurposing (hypothesis 3b).

−0.005 (0.016)

0.069

−0.321 (29.67)

0.751

−0.008 [−0.063, 0.046]

−1.09 [−1.87]

10

Reconstrual would lead to greater decreases in negative emotions about the COVID-19 situation than repurposing (hypothesis 3c)

0.068 (0.022)

0.045

3.139 (30.61)

0.011

0.067 [0.024, 0.112]

1.02 [0.32]

11

Repurposing would lead to greater increases in positive emotions in response to the photos than reconstrual (hypothesis 4a).

0.137 (0.022)

0.113

6.176 (46.79)

<0.001

0.114 [0.077, 0.151]

5.37 [4.84]

12

Repurposing would lead to greater increases in positive state emotions than reconstrual (hypothesis 4b).

−0.006 (0.011)

Random slopes by country/region were not included for the model to converge

−0.526 (20,340)

0.599

−0.011 [−0.049, 0.028]

−1.39 [−2.00]

13

Repurposing would lead to greater increases in positive emotions about the COVID-19 situation than reconstrual (hypothesis 4c).

−0.047 (0.026)

0.109

−1.781 (37.46)

0.166

−0.047 [−0.100, 0.005]

−0.41 [−0.93]

  1. All 87 countries/regions were included in the preregistered analyses regardless of their sample sizes. The signs of B, t-statistic and Cohen’s d are adjusted such that positive (negative) values indicate being consistent (inconsistent) with the direction specified in a hypothesis. For hypotheses 1–2, B reflects the difference on the original five-point scales between the average of the means of the two control conditions and the average of the means of the two reappraisal intervention conditions. For hypotheses 3–4, B reflects the difference on the original five-point scales between the mean of the reconstrual condition and the mean of the repurposing condition. Degrees of freedom (d.f.) vary due to random slopes133. Cohen’s d is calculated as the raw mean difference divided by the square root of the pooled variance of all the random components. HA, alternative hypothesis; H0, null hypothesis.
Back to article page