Mathematics intervention 2017

TMTM is a mathematics intervention programme for marginal groups in Danish Elementary Schools. The programme seeks to improve the mathematical skills/competences and motivation of students in the 2. and 8 grade through a dialogue-based approach to mathematical teaching.
These are the key elements in TMTM:
- A four day TMTM teacher course that aims to give the teachers knowledge and practical experiences with the TMTM intervention model.
- A pre-test of the students to identity the 20 % low achieving students and 20 % high achieving students.
- A 12 weeks intervention period in which the teachers teach the identified students 4 times a week.
- A post-test of the students.
In some of the treatment arms the intervention also involves:
- A tree day Supervision course for matematikvejledere (math teachers who guides and supervise the other math teachers at the school).
- Supervision of the teachers by the matematikvejleder during the intervention period.

The TMTM intervention is structured with three treatment arms for grade 2 and three treatment arms for grade 8. The treatment arms are displayed below:

Grade 2, low achieving students:
1. 1 teacher to 1 student. The teacher is supervised
2. 1 teacher to a group of students
3. 1 teacher to a group of students. The teacher is supervised

Grade 2, high achieving students:
1. 1 teacher to a group of students
2. 1 teacher to a group of students. The teacher is supervised

Grade 8, low achieving students:
1. 1 teacher to 1 student. The teacher is supervised
2. 1 teacher to a group of students
3. 1 teacher to a group of students. The teacher is supervised

2017-09-01

2018-06-30

Primary outcome:
TMTM17 pre and post tests. Test is designed for the intervention

Secondary outcomes:
Grade 2:
National tests in mathematics, grade 3
National tests in Danish, grade 2
National school children well-being survey, January-March each year

Grade 8:
9th grade school leaving exam results
Subsequent educational choice

In the first-step randomization, schools will be assigned to one of the three treatment arms by randomization. Randomization will be stratified by the number of children in grade 2. This randomization was performed in June 2017.

In the 2nd step randomization, low-achieving students will be randomized within schools such that 2/3 of students receive the intervention, and the remaining 1/3 are controls. For the high-achieving students, the ratios are reversed; i.e. 1/3 will be randomly allocated to the treatment group).

Uniformly distributed random numbers are drawn using STATA’s runiform() function, with seed set at 647877.

The seed number was determined at a meeting held at Metropol on April 27, 2017, where each participant had to write a number 0-9 on a piece of paper. These were folded, put on the table, and drawn in random order, resulting in the seed number above. This seed was used for randomizing in grade 2 as well as grade 8, and for first and 2nd step randomization. Random numbers were drawn using STATA’s runiform() function.

The schools are sorted by number of children in grade 2 in descending order, and split into strata of six schools.

The first-step randomization was performed on May 30, 2017 by Michael Rosholm.

Schools and individuals

No

62 schools in grade 2
56 schools in grade 8

765 low achieving students in grade 2 765 high achieving students in grade 2 652 low achieving students in grade 8

Grade 2: 21 schools and 255 children in each treatment arm, split into treatment and control group within each school
Grade 8: 19 schools and 217 children in each treatment arm, split into treatment and control group within each school

In grade 2, with approximately 255 low-achieving students, a power of 80%, significance level of 95%, and 95%, and, and pre-test scores explaining 50% of the variation in post-test scores, we have Minimum Detectable Effect sizes (MDEs) of 0.25 with a 50/50 split. For grade 8, with 217 students per treatment arm, we have MDEs of 0.27 with the same assumptions.