Læseklar: Helping the children most at risk of reading difficulties become ready-to-read

The aim of the literacy program, Læseklar, is threefold: First, to make the students realize that words denote content (e.g. “lion” represents the animal) and has a form made up of a number of phonemes (distinct units of sound). Second, to help students grasp the connection between letters and sounds. Third, to show students how to decode simple words (i.e., to translate written words into sounds). Decoding skills are often thought to be a necessary step towards reading comprehension, and in the “simple view of reading”, efficient decoding is one of the two main determinants of reading comprehension (the other is language comprehension; e.g., Hoover & Tunmer, 2018).

The intended frequency and duration of the program is 3-4 times a week for about 8-10 weeks, over a total of 30-35 sessions where each session lasts about 10-15 minutes. Instruction is given by a tutor one-to-one (for the students with the greatest difficulties) or in small-groups of 2-3 students. We plan to use school personnel (teachers, special educators, pedagogues) as tutors. The experiment will use a waiting list control group design and the children assigned to the control condition will wait approximately 12-17 weeks before starting the program (8-10 weeks of training for the treatment group plus a few extra weeks due to testing and school holidays).

Læseklar contains three types of sessions. The first type is focused on phonemic awareness and phonics, and the second and third on decoding. In the first type of session, tutors use clay figurines representing 25 (originally) Danish Swedish language sounds and letters, a box with a compartment for each letter, and a laminated sheet with the fingerspelling alphabet. A session starts with the tutor selecting three figurines. The students’ task is to place the figurine in its designated compartment, or its “house”. For the figurine to be placed into its house, students need to figure out a code. The first part of the code is the sound of the first letter in the figurine’s name, for example, “ss” in “sun”, and the second part consists of fingerspelling the letter. When the students have figured out that the word “sun” begins with an “ss” sound, they should make the sound for a while and at the same time fingerspelling the letter. When this combination is accomplished, the figurine can move into its house. Once all three figurines have moved into their houses, the session ends. During the next session, the students start with the previously “housed” figurines, and then continue with three new ones.

The second type of session starts after the students have learned around 15 of the letters and letter sounds, which typically takes less than a month. The tutor uses images representing short words glued to the pages of a booklet on a white card. The tutor acts as a secretary for the students when they sound out the phonemes in the word, by writing the word on the frontside of the page with the picture at the backsidecard. Each word-picture card is put into a binder. The booklet binder is gradually filled until it contains about 25 words. Using the bookletbinder, the students can also practice by themselves and at home, as they are able to verify whether they read the word correctly by turning the card to see the image. If a student needs more practice, the old words are tied together into a book and the student can begin filling athe new bookletinder with new words. If a student becomes a reasonably confident reader, he/she can go on to read short books used in regular reading instruction.

The third type of session aims to improve decoding speed and fluency by practicing the decoding of short syllables with the help of flashcards. It is therefore not used with students who have not yet reached the level where they are able to decode simple words. For the students that have reached this level, the task is to decode as many syllables as possible during one minute. The tutor registers the result on a chart and then the same syllables are repeated two more times in the same session. This procedure usually gives rise to a clear improvement, which students can follow on the chart. Making improvements visible for the students usually motivates them to practice more.

The theory of action behind Læseklar is that by focusing on phonemic awareness and phonics students will be helped in understanding that the audio stream from the spoken language can be divided into a number of letter-sounds and thereby learn how to differentiate them from one another. The small group instruction allows for an individualized pace and support, and frequent feedback from the tutor. The multi-sensory parts aim to involve as many senses as possible in order to activate the brain’s learning mechanisms and strengthen memory capacity and retrieval. Placing the figurines into their houses aims to activate the kinaesthetic, visual, tactile, and auditory senses, while at the same time creating a storyline about the figurines moving into their house that may activate the episodic memory (the memory of autobiographical events; e.g., Tulving, 2002). The consistent use of fingerspelling may also activate several senses, and the students can use this tool as a memory aid when they proceed to regular text reading. Working with figurines may also to strengthen the students’ motivation, as they may enjoy helping the figurines find their houses more than just thinking about how the beginning of a word sounds.

References:
Hoover, W. A., & Tunmer, W. E. (2018). The simple view of reading: Three assessments of its adequacy. Remedial and Special Education, 39(5), 304–312.
Tulving, E. (2002). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 1-25.

2019-10-03

2020-06-30

We will use three tests of literacy. To measure decoding skills, we will use a subtest from a standardized dyslexia test widely used in Denmark (“Elbros ordlister” from “Ordblinderisikotesten”; Gellert & Elbro, 2016). To measure letter knowledge and phonological awareness we will use two subtests from the test battery “Læseevaluering på begyndertrinnet” (“Alle bogstaver” og “Forlyd og rimdel”; Borstrøm & Klint Petersen, 2004).

We will use validated measures to test the effects of Læseklar on three skills related to executive functioning: Attention/inhibitory control, working memory, and verbal memory. Attention and inhibitory control is measured with three scales from the visual version of the Test of Variables and Attention (TOVA). The first scale we will use from this battery is “Response Time Variability”, which is a measure of the test subject’s consistency in their speed of responding. The second scale is “Errors of Omission”, a measure of inattention, and the third and final scale “Errors of Commission” is a measure of impulsivity and inhibitory control (Leark et al., 2008). TOVA is a culture- and language-free computerized test, which has been normed and can be used for ages from 4 years and upwards (www.tovatest.com). We will assess working memory with the Working Memory Index from the Danish version of Wechsler Preschool and Primary Scale of Intelligence – Fourth Edition (WPPSI-IV). The memory index is comprised of the two subtests “Picture Memory” and “Zoo Locations”, which focus on visual and visual-spatial memory. The WPPSI-IV norms are based on a Scandinavian sample of 400 children and the age interval for WPPSI-IV is from 2:6 years to 7:7 years (https://www.pearsonassessment.dk/produkter/udvikling-og-begavelse/wppsi-iv). We will measure verbal memory with the “Paired Recall” subtest from the Danish version of the Test of Memory and Learning 2 (TOMAL-2). The TOMAL-2 have been normed in the US, and can be used from age 5 and upwards (www.hogrefe.dk/shop/tomal-2.html).

These eight measures are our primary outcome measures. We describe the order we will test the hypotheses corresponding to each measure in the analysis plan (i.e., we will use a fixed testing sequence). All testers conducting the post-tests corresponding to the primary outcome measures will be hired by the researchers and will be blind to treatment status.

References:
Borstrøm, I., & Klint Petersen, D. (2004). Læsetræning på begyndertrinnet. København: Alinea.
Gellert, A. S. & Elbro, C. (2016). Rapport om projekt vedrørende tidlig identifikation af elever i risiko for udvikling af alvorlige afkodningsvanskeligheder (herunder ordblindhed). Københavns Universitet: Center for Læseforskning.
Leark, R. A., Greenberg, L. M., Kindschi, C. L., Dupuy, T. R., & Hughes, S. J. (2008). TOVA professional manual. Los Alamitos, CA: TOVA Company.

As secondary and exploratory outcome measures, we will use questions from a survey of parents. We will use this survey to examine whether the intervention changes the way parents help their children.

The assignment of students to treatment and waiting list control groups will be done in four steps:

1. Staff in each school who know the students will select the group of students believed to be most in need of extra help. This group is pre-tested using different versions of the tests of letter knowledge (“Store bogstaver”) and phonological awareness (“Vokaler”) from Læseevaluering på begyndertrinnet and the TOVA (as we expect no student to know how to decode, we do not use the decoding test at pre-test).

2. Based on the results from the pre-tests, the schools, with input from the researchers, select the final group that they want to include in the intervention. If schools, due to e.g., resource constraints cannot work with all students believed to need extra help, the students with the lowest scores on the tests of letter knowledge and phonological awareness will be chosen.

3. Researchers stratify by school and possibly, upon the school’s request, also by class (e.g., due to scheduling reasons). The exact stratification procedure may therefore differ between schools. Within each school/class we will match students based on their pretest scores on the two reading subtests, letter knowledge and phonological awareness. The matching will be done by calculating the mean share of correct answers of the two tests combined and match students into pairs by the nearest neighbor principle. If there is an uneven number of students, we will match one triple.

4. Within each pair/triple (and separately for each school/class), researchers then randomly assign students to either receive the program in a first period (i.e., to the treatment group) or be put on a waiting list and receive the program in a second period. In the case of triples, we will decide by randomization whether one or two students in the triple will be assigned to the treatment group.

We will use observations and interviews with participating teachers to examine the implementation of the intervention, as well as the control group condition (i.e., the instruction received by the waiting list control group). To further examine the implementation and guard against mistakes where students in the control group receives instruction instead of the treatment group students, the tutors will receive personalized “diaries” for each treatment group student. The diaries should be filled out after each session.

We will instruct the participating schools not to change the type of instruction given to the control group and not to implement Læseklar during lessons with regular reading/literacy instruction. This set-up will increase the chances that the treatment group is contrasted with a control group that receives treatment-as-usual instruction and that the quality of reading instruction does not change for the control group. Otherwise, if e.g., the tutors are the more qualified teachers or if resources are allocated from the control group to the treatment group, then the quality of control group instruction may decrease.

Randomization will be performed by the researchers using the random number generator in Stata, and anonymous student-identification numbers.

Individual students

Yes

Clustering of treatment and the number of clusters is unclear at this stage of the trial (the intervention has not yet begun). The assignment of treatment will not be clustered, but if some schools choose to work with small groups instead of one-to-one tutoring, the implementation of treatment will be clustered for some students.

100-120 students in 13-16 schools.

50-60 students in each treatment arm.

Assuming individual implementation of treatment and using the expected number of students (110), standard assumptions on statistical significance (0.05) and power (0.8), and using a simple comparison of means (i.e., without covariates), we can find effect sizes slightly below 0.6. If we use a linear regression model with pre-tests and pair/triple fixed effects as in Bøg et al. (2019), who evaluated the Swedish version of the program, and a similar level of explained variance by covariates as in that paper (R2 is around 0.56), the minimum detectable effect (MDE) is around 0.4. This MDE is below the effect size found for phonological awareness in Bøg et al. (2019) and less than half of the effect sizes found for decoding and letter knowledge. There were no tests of executive functions in Bøg et al. (2019). References Bøg, M., Dietrichson, J., & Isaksson, A. A. (2019). A multi-sensory tutoring program for students at-risk of reading difficulties: Evidence from a randomized field experiment. IFAU Working Paper 2019:7, Uppsala: Institute for Evaluation of Labour Market and Education Policy.