Design and procedure of the preschool study
In their last preschool year (age 5–6), the children received either phonological awareness (phonology group, N = 106) or morphological awareness (morphology group, N = 127) training or were assigned to a control group (N = 36) that participated in ordinary preschool activities.
The duration of the training sessions in the experimental groups was approximately 30 min per week for 17 weeks. Occasionally, the training sessions were divided into 2 shorter periods of 15 min each and were conducted on 2 different days.
The sampling procedure
Only Norwegian-speaking children and children with no serious developmental disabilities were included in the study. Teachers from each of the 25 preschool groups were randomly assigned to the two experimental groups. It was a challenge to find preschool teachers for a control group because the head of the municipal preschool system wanted all of the preschools to participate when we asked for permission for the study. However, before the preschool teachers could be accepted for participation, they needed to confirm that they could attend all of the training sessions and follow-ups after the intervention. For various reasons, for example supplementary training courses that had to be fulfilled during the intervention period, 5 preschool teachers were not available when the training for the intervention study occurred; thus, their groups, totalling 36 children, were selected to represent the control group. When the preschool teachers in the control group were absent, substitute teachers replaced them. All preschool teachers were qualified and had been in the community for at least 4 years and had attended the same training courses. The experimental group teachers attended several phonological awareness courses or morphological awareness courses (depending on their group assignment) before the intervention began and four courses during the intervention. The preschool teachers in the phonological group were introduced to phonological elements and phonological awareness activities, and those in the morphological group were introduced to morphological elements that are common in the vocabulary of 6-year-old children and to morphological awareness activities. Each of these meetings or courses introduced activities for the next 3–4 weeks of school, and meetings or courses were held on a monthly basis until the intervention period ended. The training was conducted on a whole-class basis or in groups of 6–12 children. For the largest groups, teacher assistants were present. These assistants primarily helped the children understand the tasks presented or helped to ensure that the children were attentive.
The preschool intervention
Phonological awareness training
The phonological awareness training programme was similar to that used by Lundberg, Frost, and Petersen (1988) and included some of the phonological activities used by Bradley and Bryant (1983). The children participated in play-like activities involving syllable and sound blending and matching words that rhymed or started with the same sound. The most demanding activities in the phonological awareness programme were carefully introduced half way through the intervention period and involved identifying phonemes (What sound do you hear at the beginning of/sun/?) and sound deletion and manipulation activities (What remains if you delete/s/at the beginning of/sun/?). Sound identification and deletion activities involved, to start with, only words starting with continuant consonants so to make the sounds easier to identify.
Morphological awareness training
The morphological awareness training was similar in structure and activities to the phonological training, except that morphemes were the word segments targeted.
The morphological training involved teaching children to identify suffixes and prefixes in words and to recognise component words in compound words. The most demanding activities in the programme involved manipulating morphemes. One activity with compounds required the child to identify the two words in “skoeske” (shoe box), then delete the first word to leave only “eske” (box), and then delete “eske” to leave only “sko” (shoe). As the last step in this activity, the children were asked to change the sequence of the morphemes to create a new word “eskesko” (box shoe) and to indicate whether this was a real word. In this case, the resulting “eskesko” is not a Norwegian word. For other training items, such morpheme sequence changes would create a real Norwegian word.
A similar procedure was used for prefixes and suffixes. Inflections with the plural form –er (gutt–gutter/boy–boys), the past-tense endings –t and –te (hoppe–hoppet/jump–jumped and smile–smilte/smile–smiled), the comparative and superlative forms of adjectives and adverbs (blå–blåere–blåest/blue–bluer–bluest and rask–raskere–raskest/fast–faster–fastest) and derivations (happy/unhappy) were used in the training. For example, the children would be asked about the word “lykkelig” (happy), and if they knew the meaning of the word, they were then asked to add “u” (un-) in front and determine the meaning of the new word. They were then asked about the meaning of “ulykkelig” (unhappy), how it differed from “lykkelig” in form and meaning and whether they knew other words that could be changed in the same way to yield the opposite meaning. The same procedure was used for verb endings (regular verbs only) and regular nouns to help the children identify changes in verb endings and in the plural form of regular nouns. For example, the children made figures to use in the training. One of these figures they called “pei” (a nonword with an adequate Norwegian orthographic structure). This “pei” (a drawn figure) was placed on a flannel board, and the children placed the name “pei” beside it. Because the “pei” was “trist” (sad), we also added the word “sad” next to it and talked about what sad means; we also mentioned the synonym “unhappy” and the antonym “happy” on this occasion. Another “pei” was then added, and the children were asked whether it is correct to say “pei” when there were two of these figures. The dialogue with the children ended with an agreement that there were two “pei-er” (peis) and that the written word, like the spoken word, needed an ending (the inflection –er was added to “pei” on the flannel board) (see also Author). Comparatives and superlatives were also added in this manner. The second “pei” appeared to be even sadder than the first one; the second “pei” was “tristere” (sadder). Actually, he was the “tristest” (saddest) of the two. For the comparatives and superlatives, the children were also shown that the written words were changed in accordance with changes in the spoken words.
Exposure to print in the experimental groups and the control group
In Norway, there is no systematic teaching of letters and no reading instruction before children enter school. The Norwegian preschool is not a part of the school system. There is no reading instruction, even in the last year before school entrance. Children have books read to them and activities are used to stimulate their language development but a large part of the time is devoted to play activities.
However, there is some teaching of letters and written words, especially with children’s names and play writing. Often, words are written on posters or children’s drawings that have been placed on the walls, providing some exposure to print during the preschool stage.
In both the phonological and morphological awareness programmes, the children were exposed to print for some activities (to the same extent in both programmes according to our instructions and according to what we observed when visiting and observing teaching), but most of the phonological or morphological training sessions contained little print exposure. For example, the children in the phonological programme listened to words and observed in print that words with the same rime have endings that look alike. The children also listened to the sounds at the beginning of words such as “sol” (sun), “seil” (sail), and “sekk” (sack); when shown the written words, the children saw that the first sound in each word was also written in the same way, with the letter “s”. The children in the morphological programme observed in print how two words could make a new, longer word and, as presented above, they saw how prefixes and suffixes were added to and deleted from words. Instead of focusing on the initial sounds of words, they focused on the suffixes and instead of focusing on the rime, they focused on the affixes. The control group had some exposure to print during their ordinary activities but they received no deliberate teaching concerning the phonological or morphological form of words.
Participants in the follow-up study
Of the 269 children in the original study, 115 (22 from phonological group, 63 from the morphological group and 30 from the control group) were included in the follow-up to be tested in Grade 6 using a word reading test and 3 text reading tests. The children attended 18 different schools after preschool, but only 9 classes in 8 different schools were followed up in Grade 6. There were no differences between the schools who remained in the study and those that left as far as national reading results are concerned. The loss of children at the follow-up and the lack of balance between the group sizes raise the possibility that the outcome of this study will be biased. However, Little’s Missing Completely at Random (MCAR) test of all the variables indicated that the children in the follow-up sample did not differ from a random sample of the 269 children tested at Time 1, χ2 (100) = 117.93, p = .066. In addition, we measured several pre-intervention factors (the children’s cognitive level, their mothers’ educational level, and the children’s vocabulary and preschool phonological abilities) that are known to influence reading development. On none of these pre-intervention factors did the children in the follow-up sample differ at Time 1 from the children that were not followed up. The p values for the t tests were: .46 for the children’s cognitive level, .79 for their mother’s education, and .16 for the children’s vocabulary and .25 for the preschool phonological abilities composite. Also, comparing the reading level for all children in grade 6 for the schools included on the measures used in this follow-up, there were no statistically significant differences between the schools on any of the reading tests. For Word Reading, F (7,117) = 1.958, p = .074 and for the three text reading measures F (7,117) = 1.74, p = .107, F (7,117) = .895, p = .513 and F (7,117) = 1.674, p = .122, respectively.
The main approach to reading instruction in all of the schools in Grade 1 was a phonics approach with a clear focus on the alphabetic principle and on phonological strategies for decoding words (a grapheme-to-phoneme strategy). The various school classrooms received children from a large number of different preschools. However, most of the children stayed in the same classroom throughout their first 6 years of school. All school teachers had a 4 year teacher-education and had been in the same school for at least 4 years. Working in the same community meant that they attended the same continuing education courses and got the same teaching support from the community’s school service program.
Preschool and early Grade 1 measures
The following linguistic measures and information about mother’s education were collected before the intervention in the beginning of the last preschool year. Non-verbal IQ and vocabulary measures were collected early in Grade 1 and were used as additional control variables in the analyses reported below.
Initial phoneme matching
The children were presented with a row of three pictures and were asked to select the picture that started with the same sound that the tester pronounced (for example find the picture of a sun when the question was to find an item starting with/s/). Two practice items and 10 test items were given. Both consonants and vowels were used as target phonemes.
For each item, the children were presented with a row of three pictures. The phonemes in the target word were pronounced with an interval of approximately 1/2 s between them. The children were asked to mark the picture that matched the resulting word (for example finding the picture of a man when given the sounds/m/-/a/-/n/). The length of the words varied from two to four sounds. Two practice trials and nine test items were presented.
Each word was presented orally together with an easily recognisable picture. The children’s task was to count the phonemes in the word and to mark each phoneme by making a pencil stroke in an empty box next to the picture, The tester would say the word slowly and ask, for example, how many sound they heard in the word “cat” as well as telling the children to make one pencil stroke for each sound. The children were given one practice trial and six test items.
Deletion of initial phonemes
For each item, the children were presented with a row of three pictures. A word was presented, and the children were told that if the first sound of the word was deleted, then one of the pictures in the row would match the resulting word (e.g., What is left if you delete/take away the first sound/r/in rice?). Two practice items were presented, followed by 10 test items. The children were presented with words with both CV and CCV onsets.
Mothers’ educational level
The mothers reported their years of education and type of education. Their total number of years in school/education from Grade 1 is reported.
Data for Raven’s Progressive Matrices (Raven, Raven, and Court, 1988) were collected in Grade 1. This test is a nonverbal group test consisting of 60 multiple choice questions. For each test item, the subject is asked to identify the missing item that completes a pattern.
Vocabulary data were collected in the beginning of Grade 1 using the WISC-R [Norwegian version (Undheim, 1978)]. In this task, the children were asked to explain the meanings of the presented words (e.g., umbrella: can you explain the word umbrella). Children were given 1 point for a partly correct answer and 2 points for a well defined explanation of the word. It should be noted that this task was conducted after the intervention.
Grade 1 reading tests
Grade 1 reading data were collected at the end of Grade 1. We used Gjessing’s (1958) standardized test for word reading and text reading which has been widely used in Norway. Both the word reading and the text reading tasks are silent reading tasks.
For each item in the word reading test the children were presented with an easily recognized picture and a varying number (4–8) of words to match the picture. The children had to mark the word corresponding to the picture. There were 3 trial items and 36 test items. Children completed as many items as they could in four and a half minutes. The test–retest reliability is reported to be 0.87 for the total word and text reading tasks.
The children were presented with passages of increasing length. After each passage there is one question to be answered by putting a cross on a picture representing the correct answer. There were 19 test items. Children completed as many items as they could in 12 min. Each correct answer gave a score of two points.
Grade 6 reading tests
Grade 6 reading data were collected at the end of grade 6. These reading tests are part of a standardised Norwegian reading assessment battery (Nasjonalt læremiddelsenter, 1995). There is a tendency for ceiling effects for all subtests because the battery was developed to identify struggling readers. However, because this test is the only standardised Norwegian battery available for this grade and because reliable self-made tests would be impossible to develop within the limited amount of time we had to select tests and conduct the testing, we decided to use the existing test battery.
This test was a silent word reading task similar to the one given in Grade 1. For each item the children were presented with a word followed by 4 easily recognized pictures and asked to select the picture that represented the word. The test included 40 items and the children completed as many items as they could in 3 min. The score was number of words correct.
Continuous, narrative text reading
This relatively challenging, narrative text contained 1099 words. The reading task was followed by 16 multiple choice questions asking for facts and information given in the text, each of which had four possible responses for the children to mark. Children were allowed 8 min to read the text and answer the questions. The text remained visible while the children answered the questions.
Discontinuous expository text reading 1 (text and table)
This task consisted of a relatively short text and a table with information about proteins, fat and carbohydrates in various foods. This task is referred to as discontinuous because children had to go back and forth between the text and the table and the questions to find answers. Seven multiple choice questions with three possible answers each were presented. The answers to the questions were primarily found in the table, but the text also included information about the task. The children were given 3 min to answer as many questions as possible.
Discontinuous expository text reading 2 (text and map)
This task included a short text accompanied by a map of the world with written information on the map. Bars were placed at different places on the map to show the amount of production in each area. Some of the bars were grey (showing the amount of oil produced), and the remainder of the bars were white (showing the amount of gas produced). As for the first discontinuous test, the children here had to go back and forth between the text, the map and the questions to find answers to the questions. The answers to the questions were found primarily on the map with the bars but were also supported by the text. The children were given 2 min to answer as many questions as possible.