The investigators of the study were Margaret Burchinal, Nathan Vandergrift, Robert Pianta, and Andrew Mashburn. Margaret Burchinal graduated from the University of North Carolina and currently work at FPG child development institute. Nathan Vandergrift graduated from Duke University. Both Robert Pianta and Andrew Mashburn graduated from the University of Virginia. The main goal of the study was to determine the relationships between child care quality and children’s academic, cognitive, and social development outcomes of children from low-income families in the public pre-kindergarten programs.
2. Who were the children in the study? (2)
–Give number, age, & sex of participants.
–Include where the sample was taken and any other defining characteristics.
The participants in the study involved one thousand one hundred twenty-nine children with a similar ratio of boys and girls from low-income families. Low-income families in the study were defined as a household that had less than one hundred fifty percent of the federal poverty measure for a certain number of people in the household. The participants were enrolled in six hundred seventy-one pre-kindergarten classrooms in the total of eleven states. All of the participants participated in two studies, which were the National Center for Early Development and Learning’s Multi-State Study of Pre-Kindergarten (Multi-State study) and the NCEDL-NIEER State Wide Early Education Programs Study (SWEEP study). Both of the studies used the same researchers and assessments. The Multi-State study was done in forty pre-kindergartens within six states from the school year of fall 2001 to spring 2002, whereas the SWEEP study was done in one hundred state-funded pre-kindergarten programs within five states from the school year of fall 2003 to spring 2004. The Multi-State study used a stratified sampling method to choose the pre-kindergarten institutes, whereas the SWEEP study used a stratified random sampling method to choose the pre-kindergarten programs. The stratified sampling method divided the pre-kindergarten programs into different groups, then some of the programs would be drawn from each group randomly or not randomly. In both of the studies, one classroom from each of the chosen pre-kindergarten was randomly selected to participate in the study. Each child would receive a take-home packet, which included a family contact form, a study consent form, and a demographic questionnaire. Then the data collectors would decide the children who were eligible to participate in the study based on the information provided on those forms. The eligibility criteria of the participants included parental consent, did not have an individual education plan, age and spoke English or Spanish fluently. After the eligible participants were determined, four children if possible two boys and two girls at each pre-kindergarten were chosen at random to participate in the study.
3. What was the design of the study? (5)
–Include experimental design (type of study and why).
–Describe procedure, in general.
–Include variables and how the variables were divided.
–Explain how the variables were measured.
–Give specific example of what the experiment was like.
The control variables in the study were the pre-test scores, children’s sex and race, and the mothers’ number of years of education. The dummy variables were used in the study for three reasons. Firstly, to determine whether the classrooms were part of the regular pre-kindergarten programs in public schools or the Head Start programs. Secondly, to account for the dissimilarities in the rate of children’s development across the classrooms. And finally, to indicate whether the classroom was in a high or low-quality range.
The researchers used the observation method and the Classroom assessment (CLASS) scoring system to measure the classroom quality. Trained observers would observe the classroom and the interactions between teachers and children for approximately every thirty minutes from the time the children arrive at the classroom until the children began to take a cap or left the classroom to go home. While observing, the observers would rate the scales on the classroom assessment scoring system. The total of seven scales for used in the study. The scales were divided into two main categories, which were emotional support and instructional quality. The first category was the emotional support involved the positive climate, negative climate, teacher sensitivity, over-control, and behavior management. Positive and negative climate described the classroom’s environment and the overall interactions between teachers and children. Teacher sensitivity described the warmth, respect, and encouragement that were provided by the teacher. Over control described the structure of classroom activities. Behavior management described the methods that the teachers used to prevent children from hurting one another. The second category was the instructional quality involved concept development and feedback quality. Concept development described the methods that the teacher used to encourage children to explain their reasoning when solving problems or expand their ideas. Feedback quality described how the teacher provided positive or negative evaluations of the children’s behaviors or works. Each of the quality within a category was rated from one to seven. The score of one and two indicated inadequate quality, three, four and five indicated adequate quality, and the score of six and seven indicate excellent quality. At the end of each observation day, the observed would compute the average rating in each category.
The children’s academic and language skills were measured by using the total of four assessments. All of the assessments took place at the beginning and at the end of the school year. The Peabody Picture Vocabulary test was used to measure the children’s vocabulary skills. The test involved reading a word that matched a picture on a card and indicated the picture that matched a certain word. The Oral Expression scale was used to measure the children’s spoken languages. The assessment involved listening to a stimulus while looking at some pictures and respond to the stimulus by answering questions, completing or creating new sentences. The applied problems and letter word identification were used to measure academic achievement. The applied problems assessed the children’s problem-solving skills, which involved completing simple math calculations. The letter word identification assessed the children’s reading skills, which involved identifying letters and pronouncing words accurately. The raw scores received from each assessment were converted to the standard scores corresponding to the children’s age group.
The social skills were measured using the teacher-child rating scale. The scale assessed the children’s social competence such as how well the children participate in group activities, and problem behaviors such as difficulty sitting still. The teachers would complete the scale by comparing how well statements on the scale described the children’s behaviors by rating from one to five. The rating of one implied that the statement did not describe the children’s behaviors at all and the rating of five implied that the statement described the children’s behaviors very well. There was a total of twenty statements under social competence and eighteen statements under problem behaviors.
4. Were there ethical concerns? (1)
–Describe the ethical concerns or specify that there weren’t any and why.
There was no ethical concern in this study. The investigators obtained the consent from the parents of all of the children participated in the study. Their parents signed the parental consent form. The measures had no obvious harm or risks to the participants.
5. What were the results? (3)
— Give overview of general results.
— Use actual numbers to represent how the variables are measured.
— Describe the statistical tests used and tell what was significant.
The results from the CLASS scales involved the mean and the standard deviation of the total score under the emotional support and the total score under the instructional quality category. The mean for the emotional support is 5.49 and 2.04 for the instructional quality. The standard deviation for emotional support is 0.68 and 0.85 for the instructional quality.
According to the children’s demographic questionnaires obtained in the study, 85% of the children of color ended up at the lower quality classroom based on the emotional support whereas 63% of white children were in the lower quality classroom. Comparing to white children, 18% of children of color were also more likely to be in the lower quality classroom based on the quality of instructions. There is no evidence suggesting that there is a different outcome between pre-kindergarten programs in public school and the Head Start program.
A spline technique was used to estimate the linear regression model, which predicted the children’s academics and behavior outcomes from the results of the CLASS scales. The spline model consisted of two separate slopes, one slope described the relationship between classroom quality and outcomes in high-quality classrooms and another slope described the relationship between classroom quality and outcomes in low-quality classrooms. The high and low-quality classrooms were selected from the sample distribution of CLASS. The range for the emotional support was between 5 to 7. There were three different ranges for instructional quality 2.5 to 7, 3 to 7, and 3.23 to 7. The data analysis involved fitting the piecewise spine models and estimating the effect sizes for the low and high-quality range. The first step was separating the slopes and intercepts in each quality range. This would create the two regression lines, one for the high-quality classrooms and one for the low-quality classrooms. The next step was to estimate the effect sizes by multiplying the coefficient by the standard deviation of quality indicator, then dividing by the standard deviation of the outcome. The effects size was used to imply the changes in the outcome with the increase in standard deviation.
The emotional support was a positive predictor of social competence, but it did not predict children’s behavior problems in high-quality classrooms. In low-quality classrooms, emotional support was a positive predictor of behavior problems, but it did not predict children’s social competence. The effects size of the relationship between emotional support and social competence the low-quality classroom is -0.01 whereas, in the high-quality classroom, the effect size is 0.09. The effects size indicated that in the high-quality classroom, emotional support was positively related to the ratings of social competence from the teachers whereas, in the low-quality classroom, emotional support was negatively related to the ratings of social competence from the teachers.
According to the effect sizes, children who were experiencing high-quality of instruction had a huge improvement in their math skills with the effect size of 8.30 and expressive language skills with the effect size of 3.97 whereas in the low instructional quality classroom, the effect size of the children’s math is 2.52 and 1.34 in expressive language. This concluded that high-quality instruction was a good predictor of expressive language and math skills.
6. What did the investigators conclude? (3)
–Explain why the investigators thought the results turned out the way they did.
–Give the “take home” message from this study (your headline).
–Give specific examples of how this may apply in real life.
The investigators concluded that the high-quality pre-kindergarten experiences may be crucial to improve children’s social skills, children’s behavior problems, and also encourage the increases in reading, math, and language skills only when the teachers maintain high-quality interactions and instructions. The high-quality pre-kindergarten classrooms involved the positive encouragement provided by the teachers, the ability of the teachers to engage in children’s conversations with respect, and the thoughtfully given instructions. In high-quality classrooms, higher quality interactions between teachers and children predicted an improvement in social skills and behavior problems. Levels of reading and math skills could also be predicted from the quality of instructions in high-quality classrooms. In low-quality classrooms, the quality of the interactions did not predict the level of social skills but predicted an increase in the behavioral problem. Levels of language skills could be predicted from the quality of instructions in the low-quality classrooms. For me, the take-home message from this study is that the quality of the interactions between teachers and children is the most important aspect that affects the children’s social and academics development. A positive teacher-children interaction promotes children’s creativity, language development, and social skills. A pre-kindergarten is a place where the children can use their imagination and abstract thinking. Children can gain insights from interacting and playing with teachers. Language is a key to communicate and make connections. The more communication that the children have, the more vocabulary sizes and stronger expressive language skills they will improve. Being in the pre-kindergarten classroom surrounded by peers and teachers is a great place to develop and improve social skills. The interaction between peers can sometimes be unpleasant, for example, fighting over toys or hitting each other. It is up to the teacher to take control of these unpleasant interactions by talking to the children nicely and patiently and explain to them the reason why their actions are not acceptable. If the teacher ended up shouting at the children loudly and punish them violently, the children might learn from those behaviors and lead to a negative outcome in the future.
7. What converging evidence would strengthen these conclusions? (2)
— Give ideas for how to improve this experiment.
— Give ideas for another experiment that would support the take home message.
There are limits to the findings. Firstly, the study only focused on children from low-income families. It is still unclear whether high-quality pre-kindergarten experiences would improve the social, cognitive, and academics skills of the high-income families’ children. Doing the experiment with families from different economic background would provide a more convincing conclusion that high-quality pre-kindergarten classrooms are essential for better child development outcomes. Secondly, only half of the children, who were approached for participating in the study, were successfully enrolled. Another half of the children failed to return their parental consent forms. The investigators believed that parents with low-income and less number of years of education were unlikely to sign the consent form for their children. As a result, the conclusions of the study might be biased. The conclusions would be strengthened if the study was replicated in families with a wide range of education background.