In this section I utilize the concept of gap to consider the question of how well what we know about curriculum matches what we need to know. Two approaches taken to this problem include questions related to curriculum effectiveness and curriculum processes.
Does Curriculum Work?
A question that has formed the base for much of the research has been that of effectiveness, or does a curriculum accomplish its stated purposes? Studies of early childhood curriculum effectiveness initially flourished in the 1960s as researchers implemented preschool interventions intended to prepare children for school. The Consortium for Longitudinal Studies concluded, âIt appears that a variety of curricula are equally effective in preparing children for school and that any of the tested curricula is better than no preschool program at allâ (Consortium for Longitudinal Studies, 1983, p. 442). Powell (1987) later countered, after reviewing a wider set of the early literature, that the type of preschool curriculum âdoes matterâ (p. 205). He concluded that the research base provided both cautious suggestions and, ultimately, further questions.
In the early 2000s, preschool curriculum effectiveness was the purpose of the Preschool Curriculum Evaluation Research (PCER) initiative, an ambitious multi-site examination of 14 preschool curricula in comparison to teachers implementing practice as usual, in randomly assigned classrooms. Based upon examining the patterns of findings across child outcome data, the report offered the conclusion that only two of the curricula were more effective than the control for achievement outcomes over the preschool (pre-kindergarten) year (Preschool Curriculum Evaluation Research Consortium, 2008). An impact on math measures was found for a researcher-developed math curriculum, supplemented by DLM Early Childhood Express math software. Impacts across early literacy and language measures were found for a relatively prescriptive curriculum package, DLM Early Childhood Express with Open Court Reading. This particular curriculum continued to show effectiveness on similar academic measures at the end of kindergarten. Positive effects at the end of kindergarten on academic outcomes were noted also for a researcher-developed language curriculum (language effects) and for a relatively scripted curriculum developed by Success for All (reading effects). Overall, the lack of positive achievement effects across most of the curricula at both times of assessment was notable (Preschool Curriculum Evaluation Research Consortium, 2008). Future research efforts have tended to be much less on the grand-scale that was envisioned with PCER, perhaps in response to the largely null findings.
Concurrently with the implementation of PCER, the question of âdoes a curriculum workâ was taken up via the What Works Clearinghouse, an initiative of the U.S. Department of Education. Groups of researchers assessed the available evidence for a curriculum, determining if studies met conditions set for methodological rigor and then considering the findings. To this day, upon examining the early childhood curriculum reviews, it does not take long to reach two conclusions. First, there is a dearth of research that meets the conditions set for consideration of the data, experimental or quasi-experimental studies. Second, there are few conclusions established thus far regarding a curriculum found to âwork,â although practice guides for early math and reading instruction were released.
A variation of this basic question has been for researchers to examine the effectiveness of a curriculum relative to differences among children â what works for whom â with a tendency to focus on child characteristics that are common demographic measures. The investigators from independent projects who made up the Consortium for Longitudinal Studies pooled their analyses to examine family structure (presence or not of a father), maternal education, childâs sex, and childâs âethnic backgroundâ (which we would currently refer to as race, as samples were largely African American). In the pooled analyses, which admittedly may have masked differential effects of curricula in conditions where multiple curricula were being implemented, there were no significant effects of these variables in regression analyses predicting child outcomes such as IQ scores, achievement test scores, placement in special education, and grade retention (Lazar & Darlington, 1982).
On the other hand, some individual-site project investigators who pursued early curriculum comparison studies did report differential effects relative to these demographic variables. Miller and Bizzell (1983) reported that in middle school, boys who had been enrolled in non-didactic preschool models scored higher on reading and math achievement tests than boys who had enrolled in didactic models. These differences were larger and more consistent than the differences found in the subsample of girls, as well as being in a different direction (Miller & Bizzell, 1983).
A contrasting approach to analysis was taken by the Preschool Curriculum Evaluation Research initiative. In the report released about cross-site analyses, child demographic variables, including race/ethnicity, sex, and maternal education, were treated as co-variates and controlled. Potential interactions of these variables with curricula were not examined (Preschool Curriculum Evaluation Research Consortium, 2008).
As noted previously, more recent investigations about if curriculum works have been smaller-scale. Researchers have tended to investigate more focused and/or supplementary curricula and approaches toward specific goals, rather than comprehensive approaches designed to provide the sum of the classroom experiences. For example, Wang, Firmender, Power, and Byrnes (2016) offered a meta-analysis of studies of mathematics curricula in pre-kindergarten and kindergarten. They noted that these focused curricula showed moderate to large effect sizes for childrenâs learning in 29 studies, with effect sizes being larger when instruction was offered for 120â150 minutes per week compared to 23â90 minutes. Thus, it appears that clear and careful intent in curriculum design, in an area as highly researched as early mathematics learning, coupled with a critical mass of instructional time, did matter to childrenâs learning.
Interestingly, the dataset from the PCER study has been recently re-analyzed with new queries. Nguyen, Jenkins, and Whitaker (2018) separated what they termed the âwhole childâ curricula (Creative Curriculum and High Scope â designed to be comprehensive approaches) and the curricula focused on either mathematics or language/literacy for comparisons at the end of pre-kindergarten. They also examined if implementation of these curricula in a Head Start versus a public-school pre-kindergarten program mattered. They found that program auspice did not impact child outcomes, while the targeted curricula impacted outcomes in the targeted domains. In conclusion, they deemed the targeted approaches more successful.
In another analysis of the PCER dataset, Jenkins et al. (2018) reported that literacy-focused curricula resulted in higher literacy-related outcomes at the end of pre-kindergarten compared to Creative Curriculum and High Scope (considered together), though with a modest effect size; a more substantial effect size was found when literacy-focused curricula were compared to the locally-developed curricula that served as control conditions. There was no significant difference between the Creative Curriculum and High Scope compared to the locally developed curricula. Similarly, the math-focused curricula resulted in both higher math and academic outcomes at the end of pre-kindergarten compared to Creative Curriculum and High Scope. In spite of the lack of impact on child outcomes, Creative Curriculum and High Scope classrooms evidenced higher scores on classroom process measures (e.g., Early Childhood Environment Rating Scale) compared to the locally developed curricula. Based on the results of these re-analyses, members of the research team have concluded that âwhile it is conceivable that some kind of effective global, whole-child curriculum will be developed, there is currently no strong evidence to support these curricula as they currently existâ and âit may be best to focus more attention on assessing and implementing proven skill-focused curricula and move away from the comparatively ineffective whole-child approachâ (Jenkins & Duncan, 2017, p. 42).
Prior to moving forward, it bears noting that the PCER study was conducted in the early 2000s; these data are now almost 15 years old. The context that drives curriculum development has changed, with early learning standards ubiquitously entrenched and now in second or later renditions in most states. Both the Creative Curriculum and High Scope groups have continued to develop their approaches and resources for teachers. For example, the fourth edition of Creative Curriculum was published in 2002. In 2016, the sixth edition was made available, with widely expanded teacher materials and resources. What was once contained within a book is now an extensive set of materials. While the age of the dataset is mentioned previously as a limitation of the study by Nguyen et al. (2018), the recommendation by Jenkins and Duncan (2017) just is not accurate in referring to these curriculum models âas they currently existâ (p. 42) being the status of the comparisons made by this research group. The gap in the edition of the curricula under study is indeed a pertinent methodological question for any conclusions being drawn.
While the question of âwhat worksâ appears to be fairly straightforward, the evidence in the literature is not robust. This reflects a significant gap indeed. But is the question entirely straightforward? Why are the answers we have long sought so elusive? I take up these questions again later in the chapter.
How Does Curriculum Work?
The research interests explored in this section are focused on the workings of curriculum. What might make curriculum effective or not? What processes operating within a curriculum are important? What do children experience within curriculum? In contrast to the previous section, in addition to quantitative designs, qualitative methods have been employed here.
In an early investigation, Stallings (1975) examined first- and third-grade classrooms utilizing a range of curriculum models in Project Follow Through (intended to extend the benefits of Head Start approaches for low-income children). A large number of variables that captured teaching processes in the classrooms were included in the analyses, and they illustrate some of the complexity of curriculum and teaching. For example, Stallings noted that higher reading and math scores were associated with small group instruction in first grade but large group instruction in third grade. Systematic instructional patterns, with the introduction of information, followed by questioning and immediate feedback, were also associated with higher reading and math scores. On the other hand, Stallings concluded that children scored higher on a problem-solving measure when they were in classrooms with more flexibility, marked by things such as a variety of materials and activities and some child choice in grouping and seating during the day. The extensive report from Stallings reflects how dense an examination of the workings of curriculum can be.
More recently, Early and her colleagues (2010) detailed the experiences of children in preschool programs across several states. While they did not utilize the notion of curriculum in describing their work, they coded via observation the types of experience variations that occur within the curriculum. Overall, they found that on average children spent slightly more time in teacher-assigned activities (37 percent) than in meals and routines (34 percent) or than in free choice activities (29 percent). Perhaps not surprisingly, regarding the nature of the activity, various language and literacy activities (17 percent â categories were non-exclusive) were coded as comprising the content of childrenâs activities more frequently than science (11 percent) or math (8 percent), but only slightly more than social studies and art (15 percent each).
The Early et al. (2010) study included an examination of classroom differences relative to race and income. In classrooms where children were relatively better off financially, there was more free choice time and less time spent as âno coded learning activity.â When classrooms enrolled more African American children, there was more time spent in teacher-assigned settings, as well as in meals/routines. Classrooms with more Latino/a children were found to have higher proportions of time spent in teacher-assigned activities and more time spent in language/literacy activities.
These studies indicate that there may be systematic differences in the experiences children have in classrooms. In some cases differences may be the result of curricular emphases, but without more in-depth examination, it isnât possible to understand the relationship of these research findings to curriculum.
A closer examination of the experiences of individual children within the curriculum is provided by qualitative researchers. For example, Quintero (2015) reported on a multi-year project focused on building integrated curriculum that was strongly related to childrenâs local contexts (e.g., family, community) and that allowed childrenâs...