The New Art and Science of Teaching. Robert J. Marzano

instructional strategies—a term I do not endorse. To illustrate, one need only enter the phrase “high-yield strategies” into an Internet search engine and thousands of results will be generated. In fact, in 2009, I wrote an article, “Setting the Record Straight on High-Yield Strategies” (Marzano, 2009b), to counteract the growing incorrect belief that research could ever produce a list of instructional strategies that would guarantee student learning. Specifically, I note that no single instructional strategy can guarantee student learning for a number of reasons. One is that many factors other than the use of instructional strategies affect student learning. Another is that instructional strategies work in concert or sets and should not be thought of as independent interventions. Still another is that educators have to use strategies in specific ways to produce positive results.

      Over time, I created an instructional model that ties the strategies, pieces, and points together in an interactive manner; it appears in The Art and Science of Teaching (Marzano, 2007). I chose the phrase art and science purposefully to communicate a message. Specifically, research and theory will never validate the notion that teaching is simply a set of preprogrammed moves manifested as strategies. Rather, instructional strategies are best likened to techniques an artist might develop and refine over years of practice. The artist then uses these techniques to create works that are not only unique and complex but elegantly focused. The more skill the artist exhibits with available techniques, the better his or her creations. Likewise, the more skill the classroom teacher has with the instructional strategies that research and theory have uncovered over the decades, the better the teacher will be able to create lessons that optimize student learning.

      The New Art and Science of Teaching, then, represents my perspective on the current state of knowledge about effective teaching. It is a perspective that certainly draws from the past but also shines light on the possible future.

The Research Supporting the Model

      The research supporting The New Art and Science of Teaching is extensive since it covers so many years and so many previous works. I discuss the complete research history in detail in “Research Base for The New Art and Science of Teaching” (Marzano, 2017). I briefly summarize it here.

       Narrative and Meta-Analytic Studies

      The first works constituting the ancestry of The New Art and Science of Teaching provide narrative reviews of the literature (Marzano, 1992; Marzano et al., 1988). These reviews are quite extensive. For example, the Association for Supervision and Curriculum Development (ASCD) sponsored Dimensions of Thinking (Marzano et al., 1988), but a consortium of twenty-eight organizations—the Association Collaborative for Teaching Thinking—supported it. That consortium included the American Educational Research Association, the International Reading Association (now International Literacy Association), the National Council of Teachers of English, the National Council for the Social Studies, the National Science Teachers Association, and the National Council of Teachers of Mathematics, among others. In effect, The Art and Science of Teaching is based on some of the most widely vetted narrative reviews of the literature to that point. As the name implies, narrative reviews are author centric in that they depend on the author’s ability to organize research and theory into logical categories. The author’s views on the extant literature greatly influence them.

      While narrative reviews were the norm at that time, the norm for research syntheses changed with the popularization of meta-analysis in the early 1990s. In How Science Takes Stock, Morton Hunt (1997) describes the nature and function of meta-analysis in nontechnical terms. Briefly, meta-analytic techniques translate findings into effect sizes that report how much increase or decrease in student learning can be associated with a particular intervention. Table I.1 reports the meta-analytic findings from a number of studies on the effect of goal setting as an instructional strategy.

Table I.1: Meta-Analytic Results for Goal Setting

      a Two effect sizes are listed because of the manner in which effect sizes were reported. Readers should consult that study for more details.

      b As reported in Hattie (2009).

      c Both Tubbs (1986) and Locke and Latham (1990) report results from organizational as well as educational settings.

      d As reported in Locke and Latham (2002).

      e As reported in Hattie and Timperley (2007).

      Source: Marzano, 2009a, p. 5.

      Table I.1 reports eighteen synthesis studies (each row represents a synthesis study) on the topic of goal setting. The third column lists the number of effect sizes in each study. In their 1993 study, Mark Lipsey and David Wilson report 204 effect sizes; in 2007, Steve Graham and Dolores Perin report 5 effect sizes. Each effect size represents a comparison between two groups—in this case, one group that used the strategy of goal setting and one group that did not. The fourth column reports the average effect size in the synthesis study, and the fifth column reports the expected percentile gain in achievement associated with the average effect size. For example, the average effect size of 0.55 from Lipsey and Wilson (1993) is associated with an increase of 21 percentile points for an average student. The average effect size of 0.70 by Graham and Perin is associated with an increase of 26 percentile points for the average student.

      Lists of meta-analytic studies like that in table I.1 appear in the following works: A Theory-Based Meta-Analysis of Research on Instruction (Marzano, 1998), Classroom Instruction That Works (Marzano et al., 2001), Classroom Management That Works (Marzano, 2003a), Classroom Assessment and Grading That Work (Marzano, 2006), The Art and Science of Teaching (Marzano, 2007), Designing and Teaching Learning Goals and Objectives (Marzano, 2009a), and Formative Assessment and Standards-Based Grading (Marzano, 2010b). These works as a whole include lists of effect sizes for virtually every element in The New Art and Science of Teaching. The studies in table I.1 include over one thousand effect sizes. If one were to list all effect sizes across these works that are foundational to The New Art and Science of Teaching, the final count numbers in the tens of thousands.

       Teacher-Designed Studies

      Since The Art and Science of Teaching was published in 2007, Marzano Research has conducted its own studies. Specifically, hundreds of teachers at various grade levels and in various subjects have undertaken studies of specific elements or strategies in their classrooms. In general, teachers selected a strategy they wished to study and identified content to teach to two different classes or sets of students. Instruction was the same for both groups with the exception that teachers used the selected strategy with one group of students but not the other. They used the same pretests and post-tests with both groups. Researchers at Marzano Research then analyzed the findings and reported back to teachers.

      To date, over five hundred teachers have been involved in such studies resulting in over one thousand findings reported as effect sizes. Visit marzanoresearch.com/research/database to access the results of each study. A series of studies summarizes many of the overall findings (see Haystead & Marzano, 2009). One of the more interesting aspects of these studies is that they were conducted with minimal and sometimes no teacher training. For the most part, teachers received either a very brief training (one half day or less) on a specific strategy, or they simply read a few pages about the strategy. This level of training probably represents the typical environment for a teacher, which involves minimal time for extensive training. The fact that the majority of teacher-designed studies demonstrated positive effect sizes in a short period of time (a few days to a