Testing: Uno, Dos, Tres!
Spanish teacher Sandy Gutierrez knows the path to a student's mind can be through the stomach. Turn in work that's like a yummy two-scoop ice cream cone, she'll say, and you've met expectations. A small, single-scoop cone would be OK and almost meet expectations. Submit a hot-fudge sundae with sprinkles, whipped cream and nuts, and you've exceeded expectations. But if an assignment resembles a dropped ice cream cone or one without any ice cream at all, it would not meet expectations.
"The food always works with kids," says Gutierrez, who has introduced this ice cream analogy to colleagues at several Fairfax County (Virg.) Public Schools' in-service sessions.
In the five or so years she has used the images in her classes, the 25-year teaching veteran has noticed that students really understand what they need to do to get by and--better yet--to exceed expectations. "If you just give them a grade, they're just going to stick [the assignment] in their folders and say, 'That's it.' The rubric gives them ideas for improving," says Gutierrez, who teaches at Falls Church High School.
The exercise is all in the name of becoming comfortable with the district's Performance Assessment for Language Students program. Since 1995, Fairfax teachers have been developing assessment tasks and accompanying rubrics for PALS so that they, along with students and parents, have a way to measure language development. The consistent message for students: Show me what you can do.
That goal replaces past practices of "catching kids at what they didn't know," explains Director of High School Instruction Marty Abbott. Language assessments used to be about verb tenses and other grammar points. Students "could do their homework and get the grammar 100 percent correct and not understand the sentences they wrote down," says Abbott, who is also immediate past president of the American Council on the Teaching of Foreign Languages. "What we have tried to develop is a generation of language students who can say, 'Yes, I can speak French (or Spanish or Arabic), and I can engage in a conversation with a native speaker.' "
"Kids are in foreign languages classes to learn to speak the language," points out Greg Duncan, president of InterPrep, a foreign language consulting firm for both K-12 and higher education. If they get into a program not based on performance of skills, they won't enjoy it. And they likely won't continue taking it.
Performance-based assessments, grounded in foreign language education standards, are catching on just in time. "What we've seen is a lot of pressure for foreign language teachers to show proof of results," says Lynn Thompson, a research associate in the foreign language education division of the Center for Applied Linguistics.
One source of that pressure is internal. "To keep languages alive in our district, I must produce quantifiable data," says Gregory Fulkerson, world language specialist for Jefferson County Public Schools in Kentucky, where his subject is the only academic discipline that's not assessed statewide. Thanks to a federal grant, the district is now implementing standards-based assessments to measure student proficiency based on ACTFL guidelines.
Without proof that foreign language education is working overall, the stakes for our country are greater than education accountability. "When the [government] says, 'Gee, we need people to fill these CIA jobs, they're not saying we need somebody who can conjugate these 12,000 verbs," Duncan says.
Yet, few ready-made options exist for assessing what foreign language students can do. There's the AP exam, which Abbott says has "much more of a grammar than an actual language use focus." The SAT II's for foreign languages also need changes to make them more performance-based, say some experts.
On the way is a National Assessment of Educational Progress foreign language assessment. The NAEP is getting kudos for its attention to standards, and it will provide the first picture of how the profession is doing nationally. But it still won't offer a look at individual student progress.
Raising a double-edged sword
While foreign languages definitely have some catching up to do--in data collection and then in being considered a valued part of the "core"--its educators consider themselves lucky. "On the one hand we're marginalized, but on the other hand we're escaping some of the ills of standardized assessments," says Judith Liskin-Gasparro, an associate professor in the department of Spanish and Portuguese at the University of Iowa. It's like a double-edged sword, say Liskin-Gasparro and other experts.
With high stakes comes accountability and funding," notes Paul Sandrock, an in-house consultant of world languages education at the Wisconsin Department of Public Instruction. But being national and state assessment-deprived means that educators can "really focus on classroom-type assessments that still capture our standards. ... Our 'high stakes' are being in sync with our standards," he says.
In that area, core subjects could envy foreign languages. "When others were trying to figure out standards of performance that were not tied to numerical scores on tests, the foreign language field already had a way to do that," notes Liskin-Gasparro.
And standards-based assessments are positively impacting instruction in foreign languages. There isn't the negative "backwash" that exists in subjects with high-stakes tests, where the tests don't deliver results that teachers are proud of--and teachers wind up teaching to the test, says Duncan.
"We're not going into it with the idea that we're meeting some bandwagon or some administratively imposed idea that we don't believe in," says Donna Clementi, world languages program leader for Appleton (Wisc.) Area School District.
Instead, Appleton teachers have been collaborating to develop standards-based assessments that are more open-ended questions than traditional fill-in-the-blanks. When asked to write a letter to introduce themselves, for example, students can "capitalize on what they want to talk about," whether it's school, family or favorite activities and foods, Clementi says.
"It doesn't narrow the scope and ask them to recall, for example, [only words about] school subjects."
This district's evolving assessments reflect what's happening nationally. "I usually [compare] the change to a jigsaw puzzle. I used to teach and test individual pieces in isolation: You will learn to conjugate these verbs; You will learn these vocabulary words. [Students weren't given] practice to put them together in a meaningful whole," says Sandrock, who taught secondary school Spanish for 16 years before moving to the state level. "The whole point of learning a little piece is, What can I do with it?"
This means moving beyond quizzes and end-of-chapter tests to recognize what good performance looks like for students at a particular level. Or, as Richard Donato of the University of Pittsburgh, says, it's an evolution from "paper and pencil tests to much more performance-based, authentic assessments and
systems." Ideally, the assessments are both everyday, formative ones and more formal, summative types, adds Donato, an associate professor of education, who works with school districts on foreign language assessments for early grades.
Bring on the tests
Just as students can find translation exercises frustrating, Jefferson County's Fulkerson has experienced that feeling while trying to convey extraordinary ideas to language teachers.
A regional representative of the National Association of District Supervisors of Foreign Languages,
he remembers one particular meeting that stimulated a lot of conversation among the organization's members. Carl Falsgraf, a director at the Northwest National Foreign Language Resource Center at the University of Oregon, had made a presentation about STAMP (Standards-Based Assessment and Measurement of Proficiency), an assessment instrument for which he had been principal investigator. The audience was impressed by Falsgraf's argument that assessments "could and should be the driving force in the curriculum, instruction, articulation, program evaluation, professional development and advocacy for language learning," Fulkerson says. In addition, the Web-based tool can be administered in one class period and it quickly breaks down results, so it doesn't "overpower the teacher's instructional time."
But teachers didn't immediately agree about the necessity of the tool, which is available through Language Learning Solutions. "District supervisors get all excited, [and then] something gets lost in the translation to teachers," Fulkerson explains.
Not all initial reactions have been thumbs-down, though. As Jefferson County teachers work with Falsgraf to pilot the online assessments in middle and high school Spanish and Japanese classes, some say they "can't wait" for the results, Fulkerson says. And principals, who are eager to get some sort of data for world languages, have largely been supportive.
Next year, the district will likely purchase LLS's ClassPak for its pilot program, Fulkerson adds. The tool helps language teachers develop proficiency-based lesson plans using authentic texts and will help everyday classroom instruction to mirror the STAMP assessments.
Also convincing teachers to embrace this type of testing is Menasha (Wis.) Joint School District, where foreign language classes are required and each curriculum unit ends in a performance assessment. "Change is difficult for everybody," notes World Languages Co-coordinator Lynn Sessler, who also teaches Japanese at Clovis Grove Elementary School.
-Marty Abbott, Director of High School Instruction, Fairfax County (Va.) Public Schools
Although, the highest hurdle for secondary school teachers is not the concept itself but the stress of how to make these assessments happen in the classroom, adds Sessler, previously a teacher at the high school level.
Educators across the country are becoming more sophisticated in recent years about teaching and testing for proficiency, Thompson has noticed. Today, when they go through the two-day training in using CAL's Student Oral Proficiency Assessment for elementary grades, they're not surprised that the SOPA interviews are so different from discrete standardized tests.
Overall, though, Donato says that "teachers need to get beyond thinking that testing means getting a score. It's more the systematic collection of student performance."
Traditional textbook formats have not helped in that aim, experts agree. 'We're trying to help teachers develop independence from textbooks. A lot are glossy and pretty and heavy, but many of them still follow the acquisition of a grammar syllabus," Donato says.
"Part of the difficulty with the traditional model of publishing is that one size must fit all," says Falsgraf, who directs the northwest resource center's Center for Applied Second Language Studies. The needs of a Spanish program in a small rural community with a large Hispanic population will be vastly different from the needs of another community without a lot of native Spanish speakers, for example. And, as Sessler has learned, when students have several years of a language in elementary school, middle school textbooks don't fit them anymore.
Student reactions to performance-based assessments are generally positive. Once they get over the shock of not being able (or having) to cram for them, that is.
"They're sometimes surprised by how much they can recall when given the opportunity to make those choices on their own [about how to respond to questions]. I don't have any students who sit and turn in a blank paper." Clementi says. Letting them highlight what they do know is an empowering shift.
And at the end of the day, week, marking period and year, students wind up with a better grasp of language use. "The performance that teachers see in students is so much greater when [overall proficiency] is the emphasis, rather than [when] only the jigsaw pieces [are emphasized]," says Sandrock.
Sessler has been especially pleased to see B and C students--who tend to take the most risks in communicating--excel in language classes where performance assessments are prominent. The tasks give "validation to the risk-taker, who's out there trying to do something," she says.
Results like these, of course, form the greatest promotional pitch of all. So what comes after educators are sold on the idea of performance-based assessments?
Teaching to the tests (proudly)
For Appleton teachers, the fruit of their labors extend beyond the home-grown assessments tracking student progress. The development process "has been a catalyst for some very rich discussions about teaching and learning," Clementi says.
In analyzing the strengths and weaknesses of a piloted assessment and rubric, teachers have stumbled across ways they might improve instruction. Recently, for example, beginning Spanish students in middle school were given a scenario where they had to prepare their families to host an exchange student. Each teacher could select any Spanish-speaking country of origin. The student responses revealed more than whether the assessment would work. "What we found was that, in our teachers' desire to open students up to the Spanish-speaking world, the kids confused Mexico, Spain, Nicaragua, (etc.). Everything was sort of just Spanish to them," Clementi says.
The district had wanted students to gain a broad picture of Spanish-speaking cultures around the world. "That global perspective translated into this stereotypical view of Spanish," Clementi explains. Now, teachers are adjusting their approach. "We've had discussions about how kids can become aware of the Spanish-speaking world and then focus on one country at a time."
"What assessment does is make us focus on what we want to happen at the end," says Duncan. He's referring to the first step in the process of backward curricular design, where educators identify the desired result. Next comes figuring out an assessment to measure progress made toward that goal. "Then you go back to the beginning and start to lay out an instructional path to develop that product," he says.
In other words, when testing is meaningful, teachers "do want to teach to the test," Donato says. "It's OK to teach to the assessment if it's a performance-based task. ... All of a sudden teaching to that task becomes part of instruction." And as assessments become more authentic, the line between assessment and instruction blurs.
"As we have [more] assessments that really look at how a child can perform, it makes us look at our teaching: What are we doing to help them perform in the language?" Thompson explains. Teachers might add more interactive tasks to the classroom routine, such as by asking students to interview each other, assigning group presentations or creating a scavenger hunt. These types of tasks are "in harmony with the way language teaching is going anyway," she adds.
Of course, adjusting instruction to aid in student performance on an assessment assumes that it's an appropriate one to begin with. "I have yet to find any single assessment instrument that's worth teaching to. No test is completely going to measure [proficiency]," Liskin-Gasparro says. "It would have to be a whole cluster of assessments."
While that idea may be impractical as a year-end measure, experts say that including performance tasks in the everyday classroom can give teachers a better indication of student progress--as well as of their own effectiveness as instructors.
Some textbook companies now assist teachers in that mission. In researching a Spanish program adoption for her district, Clementi has noticed an evolution in publishers' claims--from simply saying they match standards to explaining what standards-based instruction looks like. Thematic units end with a task that shows whether students can put what they've learned into context. Still, Clementi says German and French textbooks seem to be slower to adopt the new ways of thinking.
When commercial materials provide performance-based assessments with rubrics, teachers feel safe in trying them out, Sandrock says. "But [publishers] can't totally get rid of traditional elements because they wouldn't sell the products." So teachers need to take action, moving beyond those formative quizzes. Donato hopes that future materials will "get away from the little test in the back of the book" and provide educators with "inspiration and direction" for developing assessments with curriculum connections.
Meanwhile, 2004 promises to be a banner year for determining overall progress in teaching languages. When the first NAEP foreign language assessment is launched this fall, a nationally representative sample of 12th grade Spanish students will be asked to perform a variety of standards-based tasks. If all goes well, the same framework will be used in the future to test students of other languages, says Project Officer Janis Brown of the National Center for Education Statistics.
"Every effort was made ... to align the assessment with the most current thinking about performance-based assessments," says Dorry Kenyon, director of CAL's language testing division. The test's conversational tasks in particular model best practices, adds Kenyon, who directed development of both the assessment's framework and test items.
While the profession eagerly anticipates the NAEP data, there is some concern about whether a composite score will be reported. "We don't know if we'll be able to do that with the foreign language assessment," Brown says, since it covers four areas--reading, writing, listening and conversation--and tests students with various backgrounds and years of language study. Brown adds that the hope is to get a large enough sample of heritage language learners, or those whose families speak Spanish, to report on them separately.
"We're going to be sure that we have the backup information, so that when the media say 'Johnny can't speak Spanish,' at least we'll be there to say, 'Yes, but when Johnny started Spanish early and had a long sequence of Spanish, he'll speak it better than if he had Spanish for two years in ninth and tenth grade," Abbott says.
As for the whole movement toward standards-based assessments, Abbott and others hope that it will boost the perceived value of language learning. "Americans in general need to interact with the world, and they've only been able to do that in English. And that is such a narrow focus. It's one dimensional," she says. "When you can interact in another language, it really opens up another world to you."
Melissa Ezarik is features editor.