In a 2005 study, Mariana Achugar of Carnegie Mellon University and Mary J. Schleppegrell of the University of Michigan set out to examine how language shapes knowledge in history. Specifically, the authors looked at how the language used in history textbooks influenced the study of causality, that is, the link between particular actions and specific outcomes. They discovered that certain texts that set out to explain cause and effect contained wording that might prevent students from understanding this important concept.

The researchers focused on two historical accounts. One, from an eighth-grade textbook, concerned the expulsion of the Cherokee from Georgia. The other discussed causes of the Great Depression and came from a tenth-grade textbook. Since the writing in both was typical of middle and high school texts, Achugar and Schleppegrell looked for linguistic patterns that might help students draw a connection between cause and effect.

First, they identified two kinds of passages: accounts and explanations. An account was defined as a chronological narrative in which cause and effect emerge as a natural sequence of events, while an explanation frames events in an organized way, highlighting the key factors students should focus on.

To their surprise, the researchers found that the language in the explanatory passage (the one about the Great Depression) was no more help in explaining causality to students than the account about the Cherokee. While the "explanation" was organized differently, it suffered from the same problems that plagued the "account" of the Cherokees' removal. Both texts relied heavily on abstract nouns, failed to use explicit language linking cause and effect, and frequently employed the passive voice in describing events. Consequently, both passages created the impression that the course of history was somehow inevitable. Language, it seems, matters a great deal in shaping ideas about causality.

In the passage on Cherokee removal, the text constantly resorted to abstractions: It referenced Cherokee "resistance" without exploring what that resistance looked like or why it was unsuccessful. While the Great Depression text did focus on causes, it too used overly abstract language. Phrases like "economic overproduction" and "lessening demand" held little meaning for students trying to connect certain actions with specific actors.

As the researchers explained, rhetorical connectors, such as the word "because," can help students draw a direct link between cause and effect. Unfortunately, few of these connectors appeared in either text. They focused on "what happened," but stopped short of establishing the causal relationships that would help students understand why these events took place.

In each text, the use of the passive voice (a less direct, bold, or concise form of expression using wordy phrases like "it could be," "there were times that," "it was said," etc.) creates the appearance that events in question were inevitable. The eighth-grade text relates how the Cherokee were expelled from their land by the Georgia State Militia, but never identifies motives for ordering the tribe's removal. Similarly, in the high school passage on the Great Depression, the passive voice creates the impression that things were bound to turn out the way they did. The text cites an "uneven distribution of wealth," but never explores how or why that was the case.

The call to use primary documents to teach history comes from many quarters. While teachers are urged to use multiple documents in their classrooms, they are rarely told how to use them. To understand how students learn from multiple documents, researchers Steven Stahl, Bruce Britton, Mary McNish, and Dennis Bosquet observed how students approached different sources to make sense of a controversial historical event.

First, researchers asked 44 students in two Advanced Placement history classes to read multiple documents about the Gulf of Tonkin Incident. The documents represented a range of views and genres, from a Viet Cong cablegram to the actual Gulf of Tonkin Resolution. The students were told that once they finished reading, they would be asked to either describe the topic or develop an opinion about it. Students took notes as they read. After they read the documents, some students wrote a brief essay describing the event, while others put forth an opinion about the incident.

The students' notes revealed they had read all the documents in the same way and with one goal in mind: searching for basic facts and information. The researchers concluded that simply presenting students with multiple documents, regardless of any assigned purpose, did not enrich the students' understanding. Rather, students needed to be explicitly taught how to analyze and interpret documents.

A skilled historian doesn't simply accept a document as fact, but rather probes it, searching for bias and perspective: Who wrote the document? To what end? When? Where? To develop students' historical understanding, researchers recommended they be taught the same approach, incorporating specific elements:

• prior knowledge about the topic
• experience working with different types of documents
• explicit instruction in interpreting documents
• explicit instruction on using evidence when writing

This study supports the idea that documents play a central role in history education. But if students are expected to do the work of an historian, i.e., make sense of a variety of texts, they must learn how to navigate different types of documents in new and more meaningful ways.

Most educators use test results to measure student learning, but what are the connections between how students perform on tests and what goes on in their classrooms? Do certain teaching strategies help students score better on state tests? Julie Smith and Richard Niemi (of Oakland University and the University of Rochester, respectively) explored these questions in their 2003 study of National Assessment of Education Progress (NAEP) U.S. history results and student survey data. In particular, Smith and Niemi focused on whether taking history courses involving more active and extensive text analysis results in better test performance.

Beginning in 1986, the National Center for Educational Statistics (NCES) has overseen the development and implementation of the NAEP test in US history. Administered to fourth, eighth, and 12th graders across the nation, these tests contain multiple-choice and short-answer questions focused on themes and periods of American history. There are also questions that measure historical thinking. Smith and Niemi's study featured the 1994 test scores of 12th graders. They compared students' test scores with responses to a questionnaire about classroom practices that students filled out as part of the NAEP assessment.

Smith and Niemi found connections between the number of history courses students took, the content of the curriculum, teachers' strategies, and test scores. As the researchers expected, students taking more history classes scored higher on the NAEP. Of greater potential significance for history teachers, Smith and Niemi determined that classroom instruction is closely related to test performance. Students who had experienced "active instruction" such as complex writing tasks, in-depth reading, and extensive use of student discussion in their classrooms performed better on the NAEP. Specific examples of active instruction linked to high test scores included:

• reading primary documents
• small group discussion and projects
• student presentations
• creating research reports
• using a number of resources such as computers, film, maps, and globes

In 2005, Susan de la Paz published one of the few experimental studies that investigated teaching for historical thinking in real classrooms. The study was part of a Teaching American History grant and involved 70 eight graders, including 11 with learning disabilities. De La Paz's materials focused on 19th-century westward expansion and six topics in the middle school curriculum: the Indian Removal Act, the Whitman massacre of white missionaries, the Mountain Meadows massacre, Texas independence, women's suffrage, and the Mexican-American War. Students received 12 days of instruction in historical reasoning and 10 days of instruction in writing arguments. Their final essays were significantly better those of 62 control students who did not have the curriculum.

Results
At the end of the study, the students in the experimental group wrote essays that were longer, more persuasive, and more historically accurate than those of the control group.

De La Paz also interviewed 25 students in the experimental group before and after the intervention. She found that students deepened their understanding of why historians disagree and how they use evidence to support their claims.

Best Practices in History Instruction
De La Paz's curriculum included many elements of good instruction, and thoughtful, careful planning. We highlight the following five principles that characterize effective history instruction:

Use of Document Sets
Each topic was taught using document sets that consisted of the textbook account and at least two primary sources representing conflicting points of view. Having multiple accounts of historical events helped students understand that part of learning history involves reconciling conflicting accounts.

Historical Question to Focus Inquiry
Each historical topic was centered on a question (e.g., Who was responsible for the Whitman massacre of 1847? Who started the Mexican-American War?). These questions helped students formulate better arguments. The answers to such questions became thesis statements in students; argumentative essays.

Appropriate Scaffolds and Handouts to Support Student Learning
Students needed simple structures to develop their understanding of complex ideas. The teachers introduced a number of scaffolds to support student learning. For example, they opened the historical reasoning lessons with a mock trial that highlighted multiple perspectives and historical accounts. The teachers also used two mnemonic devices to help students organize their essays:

• S.T.O.P.: Suspend judgment, Take a side, Organize (select and number) ideas, and Plan more as you write
• D.A.R.E.: Develop a topic sentence, Add supporting ideas, Reject an argument for the other side, and End with a conclusion

Multiple Opportunities to Practice New Skills
The curriculum revisited the same reasoning and writing skills for all six historical topics on westward expansion. This approach offered students several opportunities to practice the new skills they were learning.

Gradual Release of Responsibility
The curriculum was structured so that it shifted the responsibility for thinking and analyzing from the teacher to the student. The intervention began with teachers modeling how to use the strategies. Next, teachers guided students as they began to apply the strategies in small groups. Finally, students learned to use the strategies independently.