Part 1: Take the Best Bets – Focus and Streamline the Curriculum

The demands on science curriculum, even at a middle-school level, consist of bloated, unnecessary, and unachievable expectations (Li, 2007; Li, 2006).  No one single entity or reform initiative (e.g., standards-based reform, accountability tests, textbook publishers, science teachers, researchers) carries the bulk of the blame, yet all contribute to the “obesity” of science education.  While such a system does not produce the achievement gap, it stands as a major obstacle in the narrowing of it (Li, Klahr & Siler, 2006).  In making such claims, we join a growing group of researchers and educators who have argued similarly since the onset of the science standards movement (Anderson, 2004; Anderson & Helms, 2001; Bauer, 1992; Donmoyer, 1995; Shamos, 1995; Wolk, 2004) – among them are stalwart science and standards advocates including a past president of National Science Teachers Association, a past president of the National Association for Research in Science Teaching, and the founding editor of Education Week and Teacher Magazine.

The goal of our project was to find pragmatic strategies for curriculum planning for middle-school that could survive or even thrive in such a context.  In order to do that, we must find a sensible way to cut down the sheer girth of the science content.  We propose to do that by both understanding exactly what is involved in such girth and how one may make the least costly choices to select which topics to pursue in depth and which ones to cover in breadth.  In both curriculum planning and later lesson planning, this is a consistent theme that frames our methodology.

At the curriculum level, we focus on two processes (click to expand and collapse):

 

 

 

 

 

 

 

 

 

 

 

 

References

Anderson, C. W. (2004). Science education research, environmental literacy, and our collective future.  NARST News, 47 (2). National Association for Research in Science Teaching.
Anderson, R. D., & Helms, J. V. (2001). The ideal of standards and the reality of schools: Needed research. Journal of Research in Science Teaching, 38 (1), 3-16.
Bauer, H. (1992). Scientific literacy and the myth of the scientific method. Urbana & Chicago: University of Illinois Press.
Donmoyer, R. (1995). The rhetoric and reality of systemic reform: a critique of the proposed National Science Education Standards. Theory into Practice, 34 (1), 30-34.
Shamos, M. H. (1995).  The myth of scientific literacy.  New Brunswick, NJ: Rutgers University Press.
Wolk, R. A. (2004).  Perspective: Way off course.  Teacher Magazine, 6 (2), 5

 

Part 2: Going Wide – Covering What Is Needed for Testing

Our curriculum blue print was led and driven by teachers, advised by researchers, and had buy-in and collaboration throughout its execution.  Yet it still faces an uphill race against limited time, poor materials, and disinterested students (how do we ever expect students to be truly interested in subjects ranging widely from rocks to cells to space shuttle?).

As the aforementioned blueprint got implemented, we realized that difficult choices still had to be made.  Particularly, when to teach deeply and when to merely “cover”?  Generally, a topic in which the teacher is highly knowledgeable and the students at least show signs of interest lends itself to deep pursuits.  A topic that does not readily excite the teacher or the student may have to be taught on the surface.  There is also the practical challenge of literally running out of time despite one’s best intentions.  Rather than being forced to make such choices arbitrarily or to skip topics entirely, it would be helpful to reflectively and deliberately make such choices.  The question for this section is, what would you teach within a topic once you have decided that you only have the time or resources to skim its surface?

Our implementation is simply to use tests as a guide.  “Tests” is plural, meaning that we undertake a serious effort to study and understand how any particular topic is to be tested across released items of standardized tests, whether they be TIMSS, NAEP, state, or commercial tests.  The tests, in turn, offer us a very good sense of which sub-components of the topic are most often tested.  This is clearly a version of the much-frowned-upon “teaching to the test” strategy.  We do not propose it because we believe this is the right way to teach science but because this is a practical means to enable a teacher, pressed by necessity, to skim topics without sacrificing students’ test performance.  Anyone who has done test-prep for SAT or GRE or other high-stakes tests can understand the short-term necessity and efficacy of this approach, without necessarily believing that it has long-term philosophical and educational merit.

What separates this teaching to the test strategy from those manic efforts that corrupt the entire curriculum?  We believe we are taking a highly selective approach to test preparation, in contrast to the blanket coverage approach often used.  One can teach to the test by making children remembering every single factoid that may possibly appear on a test, or, one can teach to the test by reminding children of the few factoids that would most probably appear on it.  Here, we take a strategy very much like the aforementioned merging of the standards – we sift the tests to find the “best bets” or “must cover knowledge” within each topic area.

Using the topic area of Life Science: Classifications as an example we can illustrate how this process is done.

Classification is a classic example of a science topic area that presents particular difficulty on the depth and breadth issue.  The science of classification is very interesting and can be taught “deeply”, yet the factoids involved in any typical science textbook under this topic are massive – just think of how many kingdoms, phyla, orders, and species that may fall under this subject.  Without some constraint and guidance, the amount of factoids a teacher would have to teach could stretch a mile.  This is where test items must come to the “rescue”.

First, we collect a large set of available test items and index thembased on the topic segmentation from the merged and clustered standards.

 

Second, we select the items pertaining to this particular topic (Classification of Organisms) and began to de-construct the items to identify the knowledge “chunks” a student must know in order to answer the questions.  With moderate training, we managed reasonable convergence on the extraction of the chunks from different coders and reliability on coding specific additional items by these chunks. 

 

 

Third, based on these specific chunks, we are able to construct very rudimentary test-preparation materials.  Teachers will cover these topics using lesson notes and some visual aids (e.g.: classification of organisms test prep worksheet, insects worksheet, insects/non-insects visual aid and vertebrates/invertebrates visual aid).  The goal is to ensure that the students “get” these basic chunks.  Students are then assessed using publicly released test items relevant to these topics, not for a grade, but merely to reinforce these factoids.  Here, our collected item-bank comes in quite handy for these exercises.

Why is this approach more efficient than the traditional slogging through the textbook and making students remember every single factoid at the whim of the publishers?  What, for example, prevents the number of chunks from expanding indefinitely and thus offering up no “good bets”?  We found out, a little to our own surprise, that chunks within a topic do not expand indefinitely.  In fact, for each topic area we tried, we found between 10 – 15 chunks even across many different tests and numerous test items (e.g.: classification of organisms, phases of matter, Earth processes).

There are very plausible reasons why such chunks “stabilize” across tests (just as the SAT or the GRE stabilize around certain academic skills and content), but it’s beyond the scope of this guide.  We are reasonably confident the same approach would yield limited-size knowledge chunks for most topics in science.

Part 3: Going Deep – Teaching for Mastery and Doing What is Meaningful to Students

With a focused, efficient curriculum/lesson planning method that makes real choices between where to invest in deep learning and where to suffice with surface coverage, teachers can then afford to invest time and energy when topics are important and/or have meaning to the students.

Such investment is less constrained by artificial timelines or curricular pace.  Rather, teachers can create deeply engaging experiences as much for the experiences’ own sake as for the sake of the content and skills to be acquired.  For example, topics deemed essential by our group of teachers and researchers include sound experimental design using control/contrast of variables, the three phases of water concept fundamental to physical, life, and earth science, and the embodiment of Newton’s laws in everyday phenomenon, which is a foundation for further study of physical science.

Once such topics are chosen (making sure that there are only a few per semester, or else it defeats the purpose of the aforementioned planning process), time is invested in planning these lessons for mastery, rather than simple coverage. 

• An example phases of water lesson plan created and implemented by a group of teachers who used weekly meetings for lesson planning and review. 
• Published descriptions of how the mastery-oriented teaching of “experimental design” is carried out to significantly narrow the SES/racial performance gap (Li, Klahr & Jabbour, 2006; Klahr & Li, 2005). 

We are one among the many other research and practice efforts that exploring the problem of deep learning and mastery vs. superficial coverage.  What distinguishes our overall method, however, is that we do not assume such approach can feasibly be implemented for all or even a majority of the content areas.  That’s why we explicitly design tools and methods to provide sufficient (for tests) coverage on the lesser topics in order to allow adequate emphases on important topics.  It is up to the teachers who join together for such ventures to identify for themselves which topics should be afforded priority over others based on their understanding of science education, standards, their own skills, and their students’ needs.

In addition to content depth, we also strongly advocate for creating a student-driven experience (at least for one extended stretch per semester) that allow the students to pursue science-fair like projects based on their own interests and motivation, so long as the methods are scientific.  To this end we provide a guide to such an approach developed for our urban classrooms and carried out with urban middle school students both with and without prior experience in making science fair projects.  These are challenging efforts, but doable and rewarding for both teachers and the students.  In the three urban schools where teachers banded together (not only within school, but across schools) to make this happen, two schools who haven’t had any science fair experience in over a decade were able to engage most of their middle-school students (and parents) to develop projects in just the first-year attempt; and the one school which has had a tradition of in-school science fairs have now created a 4-grade-level pipeline (students entering 5th grade doing science fair projects and continue for 4 years until they graduate) that send the students from the in-school fair, to an urban-school regional fair, to compete in an open-to-all regional competitive event.  We have just begun the explore the motivational implications of such efforts on urban school students (Siler & Li, 2006; Siler & Li, 2005).

Conclusion: A Sensible Solution for an Imperfect System

In this research effort, we set out to create lesson planning methods that are sound from the perspective of scientific learning, and are feasible and practical to serve the complex needs of today’s schools under the competing constraints of time, material, teacher knowledge, students’ interest, and the enormous pressure imposed by standards and tests.

We debunked our own naïve notion that one can “have the cake and eat it too” – that it would be possible, if only with research-based methods, to teach all science topics deeply and meaningfully within the present context of urban science education.  We have now understood and accepted that real choices have to be made to cross the mile-wide, mile-deep chasm which divides present day science teaching from meaningful scientific engagement in all learning communities.  With the help of our collaborating teachers and our students (whose engagement and participation were always the most honest and surest feedback for our methods), we have evolved, tested, and now advocate for a pragmatic approach to curriculum and lesson planning.  This method brings together seemingly disparate methods such as test preparation and student-driven projects, selective skimming of topics and mastery-oriented teaching.  It rises above the narrow ideological assumptions of advocates of standards-based reform and test-based accountability.  Its primary purpose is to create breathable space and flexible time under the existing constraints to offer teachers real opportunities to teach, and students real opportunities to engage in science.  It is by no means perfect, but we believe it is a viable solution to cope with the very imperfect system of science education likely to persist into the foreseeable future.

Resources and Works Cited in this Guide
Click to expand and collapse