Participants
Written invitations and information flyers were sent to 26 elementary schools in and around a medium-sized city in Lower Saxony, Germany. In the end, nine classroom teachers from five schools participated in the study. To prevent the sharing of intervention content between the teachers (or the children) from the two groups, all classes in one school were either assigned to the IG or the ACG. This resulted in five classes of the third grade for the IG, whereas two third grade classes and two fourth grade classes made up the ACG. At T1 data was collected from N = 140 children. Among them, n = 81 were in the IG and n = 59 were in the ACG. A few children had missing data at one or two measurement points. As a result, the sample sizes for the T1, T2, and T3 data collection were N = 140, N = 137, and N = 136, respectively. Missing data and dropouts were primarily due illness, relocation to another city, or refusal to participate. Participation rates for T2 and T3 data collections were 98% and 97% of the T1 sample.
The sociodemographic characteristics of the sample at T1 can be found in Table 1. Children in the ACG (M = 8.76 years, SD = 0.80) were significantly older than in the IG (M = 8.31 years, SD = 0.56, t(138) = 3.95, p < .001) because half of the ACG children were in fourth grade classrooms, whereas all IG children attended third grade classrooms. Neither the proportion of boys and girls nor parents’ educational attainment nor family migration nor the proportion of children from bilingual families nor the proportion of children with special educational needs (Learning, n = 4; Physical Development, n = 1; Emotional Development, n = 3, Chronic Illness, n = 1 with one child having two different needs) differed between the ACG and the IG. ACG children reported more often that they engaged in breathing exercises on their own than IG children.
Table 1 The sociodemographic background of the children in the IG and the ACG at T1 with percentages in parentheses
|
IG N (%)
|
ACG N (%)
|
Total N (%)
|
Test statistic a
|
Gender
|
|
|
|
X²(1) = 1.73, p = .188
|
Female
|
35 (43)
|
33 (56)
|
68 (49)
|
|
Male
|
46 (57)
|
26 (44)
|
72 (51)
|
|
Parents‘ education
|
|
|
|
Χ²(1) = 1.32, p = .251
|
No vocational qualification
|
14 (17)
|
16 (27)
|
30 (21)
|
|
Vocational qualification
|
63 (78)
|
41 (70)
|
104 (74)
|
|
Missing
|
4 (5)
|
2 (3)
|
6 (4)
|
|
Immigrant family
|
|
|
|
Χ²(1) = 1.76, p = .185
|
No
|
65 (80)
|
42 (71)
|
107 (76)
|
|
Yes
|
13 (16)
|
16 (27)
|
29 (21)
|
|
Missing
|
3 (4)
|
1 (2)
|
4 (3)
|
|
Bilingualism
|
|
|
|
Χ²(1) = 1.98, p = .160
|
No
|
69 (85)
|
43 (74)
|
112 (81)
|
|
Yes
|
12 (15)
|
15 (26)
|
27 (19)
|
|
Special educational needs
|
|
|
|
Fisher's exact test p = .721
|
No
|
77 (95)
|
55 (93)
|
132 (94)
|
|
Yes
|
4 (5)
|
4 (7)
|
8 (6)
|
|
Prior experiences with breathing exercises
|
|
|
|
Mantel-Haenszel Χ²(1) = 4.21, p = .040
|
Never
|
65 (80)
|
36 (61)
|
101 (72)
|
|
Once a month
|
5 (6)
|
8 (14)
|
13 (9)
|
|
Once a week
|
5 (6)
|
9 (15)
|
14 (10)
|
|
2-4 times a week
|
3 (4)
|
5 (9)
|
8 (6)
|
|
5-7 times a week
|
2 (3)
|
1 (2)
|
3 (2)
|
|
Missing
|
1 (1)
|
0 (0)
|
1 (1)
|
|
Note. IG = Intervention group; ACG = Active control group; N = sample size; SD = standard deviation. a Unless otherwise stated the test is a Pearson's Chi-square test
|
Table 1 compares the IG and the ACG in terms of gender, parents’ education, immigrant family, bilingualism, special educational needs, and prior experience and finds no statistical differences between the two groups. There was only one significant difference: ACG children tended to have more experience with breathing exercises than IG children.
Procedure
After receiving approval from the Ethikbeirat of Leuphana University on May 6, 2020, T1 data collection took place in September 2020 after informed consent was obtained from all subjects and their legal guardians. T2 data were collected in December 2020, just a few days before the second COVID-19 pandemic-related lockdown in Germany. The T3 assessment took place in May 2021. Because of the pandemic, children’s attendance at school between T2 and T3 was limited. In order to limit infections, most classes were split up, with each group attending school for two or three days per week. Data were collected with the help of tablet computers, with a test manager in front of the class who read the items and instructions. Children could read along silently. The working memory updating task was taken in small groups. Trained undergraduate research assistants supported the children who encountered technical difficulties with the tablets. All methods were performed in accordance with the relevant guidelines and regulations. More details can be found in von Salisch and Voltmer15.
Data collection took three to four teaching hours at each measurement point, with the usual breaks in between. Children could participate in the RCT only when they provided a written consent from a parent or guardian. Their participation was voluntary. At each measurement point, all children received a small gift after finishing the data collection. Teachers completed rating scales for each child of their class and paper-pencil questionnaires about themselves. Parents completed a brief questionnaire regarding their family situation, the languages spoken at home, and their educational and professional backgrounds. At each measurement point, children took the objective test of arithmetic in their regular math lesson under the supervision of their math teacher.
In June 2020, the nine teachers of the IG received the BBI curriculum and participated in a short mindfulness training. During the 15 hours of the training, they were introduced to the concept of mindfulness, practiced key mindfulness exercises, and received guidance on leading the breathing exercises for the children in their classes. Teachers' implementation of breathing exercises with the children was supervised by a certified MBSR teacher with extensive experience in teaching mindfulness in primary schools.
The BBI was conducted between T1 and T2 for approximately ten weeks in the IG classrooms. Teachers were asked to lead the Breathing Breaks up to three times per day, every school day. Due to the nationwide pandemic-related lockdown from December 2020 to March 2021, most children were taught remotely, which made it impossible to continue the Breathing Breaks in school. During the intervention period, ACG teachers were asked to engage their students in coloring mandalas up to three times per day, every school day.
Measures
Arithmetic performance. The German Lernverlaufsdiagnostik-Mathematik für zweite bis vierte Klassen (LVD-M 2-4) [Curriculum-based Measurement-Mathematics for Second to Fourth Grades] by Strathmann and Klauer33 was used to assess level and development of elementary mathematics calculation skills. The test consists of a total of 24 items - six for each arithmetic operation (addition, subtraction, multiplication, division). In LVD-M 2-4, each child receives an individualized set of items which is randomly generated by a software which comes with the test. In second grade, mental arithmetic involving addition, subtraction, multiplication, and division within a number range up to 100 is assessed. In third grade, mental arithmetic in all four mathematics operations in addition to written arithmetic tasks involving addition and subtraction within a number range up to 1000 are tested. In fourth grade, students are asked to solve mental arithmetic problems in addition and subtraction, as well as written arithmetic in all four mathematics operations up to 10,000. Correct answers are summed up and the maximum score is 24. German norm data are available so that raw scores could be translated into T-scores.
Because of the first Covid-19 related lockdown and associated delays in learning in the preceding school year, teachers expressed concern that many of their students were not able to meet grade-level expectations and would be frustrated by a difficult test. Therefore, at T1 the test versions of the previous grade were used, e.g., third graders took the test for second graders and fourth graders that for third graders. T-scores from the end of second grade were used to evaluate third graders’ performance at T1, whereas T-scores from the end of third grade were utilized for fourth graders’ performance at T1. At T2 and T3, however, grade-appropriate arithmetic tests and T-scores were employed. At T2 (December) T-scores for the middle of the school year were used. T-values for the middle of the school year were also chosen for T3 because the usual growth in learning could not be assumed due to the second Covid-19 related lockdown. Criterion validity of the LVD-M 2-4 was established by high correlations with another standardized mathematics test (r = .53 to r = .80) and the math grade (r = - .54 to r = - .77) in the norming sample32.
Teacher ratings of mathematics. At all three measurement points, teachers rated children's overall mathematics performance with one item (“How do you rate the student's level of achievement in mathematics?”) on a scale ranging from 1 to 5 (strongly below average [1], below average [2], average [3], above average [4], strongly above average [5]).
Working Memory Updating (WMU). Children’s WMU was measured through the backward digit span of the app version of the Eichstätter Messung des Arbeitsgedächtnisses (EI-MAG; Eichstatt Measurement of Working Memory34). In this objective test, the digits 1 to 9 were presented via headphones in intervals of 1.5 seconds. Afterwards, a 3x3 matrix of Arabic numerals appeared on the tablet screen with the digits 1 to 9. Children were instructed to enter the digits of the previously heard sequence in reverse order into the keyboard of the tablet. The number of correct answers (accuracy) was tallied. In the adaptively designed EI-MAG, the number of digits to be remembered increased which raised the level of difficulty. Children were asked to reproduce two out of the three series of the same length in each block correctly before they could proceed to the next level. In the first block, two digits were to be remembered. In the most difficult block, a span of eight digits had to be reproduced. The test ended when children had made three mistakes within one block. The EI-MAG started with a practice phase. All children who achieved zero correct answers during the practice phase were excluded from the analyses because it was assumed that they had not understood the instruction.
Socioeconomic Status. Children’s socioeconomic status (SES) was operationalized by their parents’ highest educational attainment which was assessed on a scale of 1-7 (no degree [1], secondary school degree (lower track) [2], secondary school degree (middle track) [3], high school diploma [4], apprenticeship [5], technical college [6], university degree [7]) which they had self-reported. Subsequently, a dichotomous scale was derived, which categorized the participants into two groups: those without a vocational qualification [1] and those with a vocational qualification [2].