Introduction

In the 1990s, tests in reading became a high priority for teachers, administrators, and students.  As schools and districts attempt to improve student achievement, they place an emphasis on accountability.  Schools and teachers are expected to show that their students are achieving well on tests in reading, math, and content areas.  In an environment of school improvement through accountability, testing is a “high stakes” part of teaching and schooling.  This chapter discusses how teachers can prepare students to take high stakes tests in reading.  First, we address the issue of what high stakes tests are.  Second, we describe reactions of teachers, students, and administrators to high stakes testing.  Third, we present recommendations for preparing students for high stakes test taking.  Fourth, we discuss hazards in test preparation.  And fifth, we propose a framework for test preparation planning.  What are the long-term and short-term designs for improving test scores in your school?  While there are few answers that apply absolutely to everyone, a considerable amount of research has been conducted which sheds light on all of these issues.  References to this work are provided. 

Characteristics of High Stakes Tests

A test or assessment program is called high stakes when it is used to make extremely important decisions about individual students, teachers, or schools.  A recent report by the National Academy of Sciences entitled “High Stakes: Testing for Tracking Promotion and Graduation” (Heubert & Hauser, 1999) reviews many of these decisions.  One prominent decision is promotion and retention of students.  When a test is used to decide whether a student has “passed” fourth grade and should move into fifth grade, the decision is filled with consequences for the learner.  The test on which that decision is made is a high stakes test because the child’s future is affected, the parents are likely to be involved, and teachers will be influenced by it.  Across the nation, retention in a grade is a frequent practice.  For children 6 to 8 years of age in the 1990's, approximately 18% of the population was retained.  For the population age 9 to 12, about 24% was retained at some time in their school lives and for the population 15 to 17 of age, approximately 30% was retained at least once during their schooling.  Whether retention improves achievement or not cannot be discussed at length here.  However, one basis for retention is a high stakes test.  Related to retention is graduation from high school.  Increasingly states are making high school diplomas contingent on successful test performance.  A report from the Board on Testing and Assessment (Feuer et al., 1999) reviews this practice across all states.  In the review, 18% of the states had a high school exit exam, which is a high stakes event for students and teachers. 

The most prominent use of high stakes tests is improvement of instruction.  In 1998, 43% of the U.S. states reported using assessment to improve instruction.  The tests in these assessment programs may be standardized tests or performance assessments.  However, in every case, they are high stakes events.  In some cases, schools scoring low on a high stakes test are subject to “take-over” by the State Department of Education.  In other cases, Principals are providing teachers with welcome professional development opportunities. When school improvement is the aim of high stakes testing, the impact of testing on teachers can be substantial.

In addition to school improvement, tests may be used for tracking in high school.  Students are assigned to academic or general tracks based on a testing event.  This event has lifelong consequences as students rarely shift tracks during their high school experience.  In elementary schools, students may be assigned to special resources or special education based on tests scores.  Such a score is a high stakes piece of information as it determines the child’s educational experience dramatically.

Tests used for “high stakes” purposes vary widely.  In some cases, traditional standardized tests are used.  Frequently, tests such as the Iowa Test of Basic Skills or comprehensive achievement tests (e.g., Comprehensive Test of Basic Skills, CTBS) are employed.  Often these include a high proportion of multiple choice items.  Traditional items, such as short passages with brief questions and four alternatives are used in the primary grades to test word recognition, vocabulary, comprehension, and occasionally listening. In the intermediate grades, subtests often include vocabulary, word attack skills, and comprehension.  These tests are favored by administrators because they are machine scoreable and easily administered.  This reduces costs and usually increases frequency of test taking.  Multiple choice reading tests have the advantage of high reliability.  They are consistent and student performance from one form to the next does not change much.  However, many educators question the validity of multiple choice tests.  We rarely take multiple choice items in the real world and “naturalistic” items are increasingly preferred by reading experts.

Performance assessments are more popular with reading experts.  These assessments include longer passages with full stories, open-ended questions, and constructed responses.  Students write their reactions to text and the responses are scored according to a rubric from low to high.  Even more valid in the eyes of many reading experts are portfolio assessments.  In this approach, student portfolios are maintained by the teacher in the classroom.  They are coded for reading, writing, and ownership of literacy (Purves, 1993).  However, performance assessments and portfolios are generally less popular with administrators because they are time consuming and expensive.  Further, test reliability may be difficult to attain (Valencia & Place, 1994). The gain in validity is offset by a high administrative burden.  Nevertheless, all three types of assessment approaches in reading are used from high stakes purposes.

Reactions of Teachers, Students, and Administrators to High Stakes Testing

Teachers have many reactions to high stakes tests.  One junior high teacher in a study by  Haladyna, Nolen, and Hass (1990) stated that “. . . the test raises the anxiety level of everyone.  The Superintendent likes to use the scores to point out the teachers’ weaknesses and create competition among the teachers.  He thinks that good scores equal good teaching” (p. 62).  In an extensive study of how teachers react to high stakes testing programs, Smith (1991) reported five reactions.  First, publication of test scores produces feelings of shame, embarrassment, guilt, and anger in teachers.  This is most evident in low scoring schools or classrooms, but most teachers feel they could have done better.  Consequently, many are disappointed in their level.  This leads to the determination to do what is necessary to avoid such disappointments in the future.  Second, some teachers question the validity of the test and doubt whether a multiple choice measure of reading, in fact, represents “real reading.”  Third, many teachers believe that there is an adverse emotional impact of testing on young children.  They think that when young students are made to feel and incompetent, they are inappropriately stressed by the testing processes.  Fourth, teachers often resent and regret the instructional time required by testing.  In some cases, time spent preparing, administering, and recovering from tests may cost 100 hours of instructional time in one school year.  Since the tests are not teaching, students are not learning in this time.  Fifth, teachers often adapt to the tests that are high stakes in their school.  This adaptation often means narrowing the curriculum and restricting their teaching to specific test-like activities.  In reading, this is often regarded as a trivialization of the reading process (see Au in this volume).  In extreme forms curriculum narrowing leads to a loss of creativity, imagination, and critical thinking among students.

While teachers may have adverse reactions to high stakes tests, administrators frequently favor them.  Many administrators, from the school principal to the state superintendent of education, value high stakes tests.  Administrators expect tests to guide instruction, hold schools accountable, aid in curriculum planning, and evaluate school programs (Roeber, 1998).  Needless to say, administrative emphasis on results of testing is what makes them “high stakes” for schools and teachers.

Students also have reactions to tests.  In the elementary grades, students believe that a test measures how much they have learned, identifying their knowledge and intelligence.  In one study, students thought that the tests showed which teachers were doing the best job in the classroom.  They typically indicated that they tried fairly hard and were reasonably motivated to succeed on the tests. 

In contract, middle and high school students had different views.  They were less likely to say, “I gave my best effort on the test we took” than elementary students.  Some secondary students believed that the tests served the school’s interest but not their own personal interests.  Unmotivated students protected their self-esteem by not exerting effort on the tests. Because they expected to do poorly, they did not wish to appear that they had put forth maximum effort.  At the same time, secondary students felt less prepared to take tests than elementary students.  They reported they did not have good strategies for taking the tests were often anxious about their lack of skills.  For secondary students, high and low achievers often reacted differently.  High achievers were likely to report that they do well on reading tests, that the test was easy, and that they reread parts of the passage when needed.  They showed good strategies and positive self-perceptions.  However, low achievers were likely to feel resentment, anxiety, lack of appropriate test-taking strategies, and decreasing motivation.  There are many obstacles to success on high stakes tests.  As a consequence, it is important to address students’ needs for test preparation (Paris, Lawton, Turner, & Roth, 1991).

Taking Tests: Cognitive Processes

In this section, we first discuss what goes into performance on a reading test.  We next discuss how teachers can prepare students for all of the requirements for successful reading test taking.  The ingredients of reading test performance are displayed in Figure 1.  This pie chart shows that success on a reading test has many dimensions.  These ingredients for reading test performance are based on correlation research (Feuer et al., 1999).  By looking at how different types of tests correlate to each other and examining performance on different types of items, we found several important elements.  As the pie chart shows, the overwhelmingly prominent factor for explaining differences in test scores was general reading ability.  This was based on the fact that if a group of students were given two very different types of tests, such as a simple multiple choice test and a complex long-term portfolio assessment, their scores were substantially correlated.  Students who scored highly on the portfolio assessment also scored highly on the multiple choice test.  Of course, the agreement is not perfect and there are many exceptions.  However, it is extremely important to note that general reading competence is by far the most important, single ingredient for success on a high stakes test.  This has vitally important implications for teachers which we will discuss later. 

Second, the pie chart shows that student knowledge of the content of reading passages determines their success on tests.  Similar reading comprehension tests on the same topic were much more highly correlated than similar reading comprehension tests on different topics.  If students do not have the content background on volcanos, deserts, or early colonial history, their performance on a reading test with these contents will be relatively low.  Next in importance are two factors consisting of motivation and test-wiseness.  Motivation for succeeding in a test is essential.  Students who do not care about their success or who are highly anxious will score poorly.  Unmotivated students fail to complete items, do not attempt difficult passages, write incomplete answers, and they do not concentrate their cognitive energies. 

Fourth, test-wiseness is important.  Students need to understand the format of a test.  Strategies for taking an essay-type performance assessment are different from strategies for taking a multiple choice test. Studies have shown that when two tests were identical in format and different in content their scores were highly correlated.  However, their correlations were only slightly higher than two tests that differed in both format and content.  This small increase is due to knowledge of format and test-wiseness of students.  These format factors account for relatively few of the differences among students. The fifth factor influencing reading test performance is general intelligence.  Because a test is a mental activity, it will measure intelligence to some degree.  We know this because student scores on reading tests are correlated with their scores on math tests of computation.  These tests look different but they both require general mental activity and intelligence.  Last, it should be noted that no test is perfect.  As the pie chart depicts, some of the differences between students cannot be traced.  These differences are unknown sources of error in the process of measurement.

Preparing students for tests

Preparation for reading tests can be based on requirements for successful test performance.  The ingredients of successful test preparation are shown in Figure 2.  This is a second pie chart depicting the dimensions of effective test preparation.  The largest dimension of this test preparation plan is guided instruction in reading and writing. We recommend that good instruction in integrated reading-language arts is the most vital factor in test preparation.  The second element of test preparation is strategy instruction. The specific strategies needed for reading on the test should be taught explicitly.  If the test is designed for Grade 4 students and requires comprehension the plot, character development, and resolution in narratives, these element of story structure should be taught explicitly.  Third, a substantial emphasis should be placed on engaged, independent reading to learn.  All reading tests require speed, fluency, and comprehension.  This can only be learned in motivated, extended independent reading.  However, the objective of reading to learn during engaged reading is important, as the test situation requires this process.  

Fourth, practice on the format of the test is valuable.  This accounts for some of the differences between students in test success.  When taking multiple-choice tests, students should understand the strategies for responding to the items, selecting alternatives, and allocating their time appropriately.  Performance assessments often require extended writing to a prompt, such as “compare and contrast.” Students need to learn what this prompt calls for. However, research shows that such format practice should not be used excessively (Feuer et al., 1999).  Fifth, motivation for reading and test taking should be part of the preparation program.  Motivational support can include alleviating students’ anxieties. It can consist of providing meaningful reasons for test success, such as the future consequences of low test performance. Most important, motivation support consists of fostering extensive amounts of reading throughout the school year.     

The numbers in the pie chart are valuable guides.  Each number refers to the percentage of time that should be spent in that activity during a 10-week period preceding the test.  For example, in the 10-week period preceding an annual high-stakes test of reading, 20% of the time should be spent on strategy instruction.  This is related to general strategies in reading required by the test.  Approximately 10% of the time should be spent on format practice so the students are familiar with the types of items and questions.  The largest time block should be spent on guided instruction in integrated reading-language arts including a healthy array of texts.  Teachers should maintain their coherent instruction on fundamental objectives in reading-language arts.  This is because the test success is only slightly influenced by knowledge of format and slightly influenced by strategies for reading.  It is highly influenced by general reading competence gained through substantial time spent in a coherent curriculum that is aligned with the test.

Preparation in test formats

When children face a test, the format may be entirely new to them.  For example, primary students may face a multiple-choice format that is unfamiliar.  Help in understanding the structures of the items is necessary.  Providing direct instruction on what is expected in the testing formats enables young primary students to show their true reading competence.  For example, many students in a test of word meaning (e.g., The Woodcock Reading Mastery Test, word comprehension section) need to respond to analogies.  They are given one item, such as “dog is to elephant,” and then are given a word such as “stone” and asked to identify the analogous item from the alternatives of “mountain”, “water”, “wood”, and “book.”  Mountain is correct because it is large compared to the stone as the elephant is large compared to the dog.  Helping students understand the meaning of these analogies increases their reading skills as measured by the tests with this format (Franklin, 1983).   Some teachers have students analyze multiple-choice test items, or other test formats to determine what makes an item hard or easy. This helps students to understand the discourse structure of test items.

Actually providing students practice with items from a test has been shown to increase first graders’ test achievement scores (Eakins, Green, & Bushell, 1976).  First graders who received a unit of instruction lasting 1 to 2 hours in total time, had significantly higher scores than students with no preparation.  However, students who were given longer instruction did not surpass students given instruction limited to the 1 to 2 hour intervention.  Format practice is also valuable for high school students.  For example, practice on the multiple choice items in high school physics enabled low achieving students to increase their test scores (Kirkpatrick, 1934). Test practice is especially beneficial to students with low test taking skills, such as mildly handicapped or learning disabled populations. (Scruggs, Mastropieri, & Tolfa-Veit, 1986). 

The phrase “test wise” is applied to students who seem to have the ability to use the characteristics and formats of a test or a test-taking situation to receive a high score.  Millman and Bishop (1965) conducted an extensive study of test-wise students.  Their characteristics included:

            (1) Using time effectively by beginning to work rapidly.  Test-wise students omitted items which seemed unnecessarily difficult and used their time on items they can get correct.

            (2) Implementing error-avoidance strategies that included paying careful attention to directions, determining the purpose of the task, and checking all of their answers after completing the test.

            (3) Applying a guessing strategy to multiple-choice tests which encouraged guessing when alternatives within an item can be ruled out.

            (4) Using a deductive reasoning strategy which consisted of eliminating items known to be incorrect and choosing among other options.

            (5) Adopting a level of sophistication that is expected.  Often multiple choice items may be partly correct and finding “the closest” alternative is important.

            (6) Using cues in the items, such as selecting longer options if they appear to be plausible, avoiding extreme words, such as “never,” and considering the relevance of specific details in options.

Students at the middle and high school levels who get high test scores are more likely to use these strategies than low test scorers.  Some minimum time spent discussing and practicing these strategies may increase students’ competence and comfort in high stakes tests for middle and high school levels. 

In performance assessments, students are often expected to construct written responses to a wide array of different types of texts.  In these tasks, the criteria for an acceptable response are important to the test taker.  Teachers can help students prepare for these performance assessments by discussing the criteria for excellent constructive answers.  These criteria are often presented in scoring rubrics that can and should be discussed and taught directly (Mehrens, Popham, & Ryan, 1998).  To help students with constructed response items, teachers should foster students’ self-evaluation skills.  As good students read, take notes, write, and perhaps revise, they are continually self-monitoring.  This is a complex strategy that can be taught (Garner, 1982).  Students need instruction in generalizing these strategies should as identifying evaluative criteria and self-evaluating with respect to those criteria.  Teachers can provide practice on a range of reading and writing tasks in the normal reading instruction program to foster these test-relevant strategies (Mehrens et al., 1998).

Preparation in reading strategies  

The vast majority of reading tests depend very heavily on reading comprehension.  Beginning with third grade tests and extending through high school tests and college entrance exams (The SAT), text comprehension is a fundamental part of every reading test.  For many reading theorists, reading and comprehension are the same thing.  They believe that students who cannot comprehend, cannot read. In light of the importance of reading comprehension, good test preparation includes direct instruction in the strategies necessary for effective reading comprehension.  However, teaching reading comprehension strategies is a long term enterprise.  Reading comprehension grows constantly during a child’s period in grades K through 12. Needless to say, it cannot be accomplished in one week prior to the test.   Consequently, a brief burst of instruction prior to a test is not effective test preparation.  Effective test preparation depends on a sustained, long-term plan of providing instruction and opportunities to learn effective comprehension strategies.  For effective test preparation, a period of not less than three months might be devoted in the school curriculum for teaching reading comprehension strategies that will be required on a high-stakes test.

Strategies needed for success in high-stakes reading comprehension tests are well established.  An abundance of research supports the importance of the following: (1) Using background knowledge, (2) Searching to locate information, (3) Summarizing, (4) Self-monitoring, (5) Self-questioning, (6) Concept mapping, and (7) Self-explanation.  Good readers are known to use these strategies when necessary, but poor readers rarely use them.  Furthermore, these strategies can be taught and instruction in them will improve high-stakes reading comprehension test performance.  For example, Blanton and Wood (1984) suggested that direct instruction in these strategies would improve test performance.  The direct instruction recommended included: (1) teacher modeling of the strategies, (2) explanation of when and how to use them, (3) repeated opportunities for guided practice, and (4) extended independent reading.  Furthermore, primary children have been shown to benefit from direct instruction on reading comprehension strategies (Schwartz, 1988).

To help students succeed on tests, long-term strategy instruction on using background knowledge is valuable.  A number of researchers have documented the benefits of teaching students to use background knowledge and to make inferences between what they already know and new information in the text during reading.  For example, Hansen and Pearson (1983) showed that low achievers could improve their reading test scores by learning to use background knowledge and draw inferences as they encountered new text information.  In addition, researchers have shown that good readers monitor their comprehension, whereas poor readers are less likely to engage in self-monitoring.  For example, Garner (1982) showed that good and poor readers differed in how frequently they used a rereading strategy to improve their understanding.  In addition, students who were taught to reread and practiced this strategy frequently improved their tested performance.  Self-questioning has been identified as a powerful strategy for reading comprehension.  A large review of intervention studies by Rosenshine and Meister (1994) showed that self-questioning can be taught with a range of texts and it benefits reading comprehension if the instruction is extended over multiple lessons.  Self-questioning is frequently used in reciprocal teaching which is a technique known to increase reading comprehension on standardized tests (Lysynchuk, Pressley, & Vye, 1990). 

Many tests require students to locate specific information within a story or informational passage. The processes of searching and locating information can and should be taught as comprehension-relevant strategies.  Understanding text structure helps improve search. By understanding story structure, children can search for particular information about characters or events within stories.  By learning the organization of information text, such as “cause-effect” or “historical sequence,” children can learn to locate information needed to answer specific questions.  Involving students in inquiry projects is an excellent technique to foster their learning to search and retrieve information.  Within inquiry projects, students can be taught concept mapping.  With this strategy, they can synthesize information gained from a variety of texts. The strategy of self-explanation can also be included in inquiry projects.  As students encounter new text-based information, they can be encouraged to answer the question “Why?”  This fosters their integration of new knowledge into old background information.  Students who have learned to explain new information to themselves frequently are likely to score higher on reading comprehension tests.  These strategies are useful for the majority of students in the mainstream of the classroom and are also beneficial for students with learning disabilities (Hughes & Schumaker, 1991).  It is evident these strategies are complex. The optimal preparation for tests of reading comprehension that require these strategies begins several months prior to the administration of the test.  Incorporating the comprehension strategies needed on a test into the mainstream curriculum planning is the most likely technique for building student confidence and self-confidence for successful test performance.

Test preparation through engaged reading

One of the most well established findings in reading research is that reading comprehension is associated with extended, purposeful, motivated reading activities.  Students who read frequently and widely for their own interest are the high comprehenders and high achievers.  If we look at standardized test performance in Grade 3, 4, 5, or 6, the strongest predictor of achievement is amount of reading.  Students who read widely and frequently are high achievers; students who read rarely and narrowly are low achievers (Guthrie, Wigfield, Metsala, & Cox, 1999).  Amount of reading is such a strong predictor of reading comprehension that it out weighs intelligence, economic background, and gender.  That is, students who are active, engaged readers will be high achievers even though they come from backgrounds with low income or low education in the family.  Likewise, students who are infrequent, disengaged readers will be low achievers on standardized tests of reading even if their parents are well educated, their income is substantial, and literary resources are available at home (Reutzel & Hollingsworth, 1991). Note that this does not refer simply to sustained silent reading, but to conceptually-driven, learning and thinking with text that is strategic and socially interactive (see Baker, Dreher & Guthrie, 2000). However, engaged reading does not become part of the child’s lifestyle in a week or a month.  This most important contributor to standardized reading tests achievement requires a long-term commitment from teachers and schools.

Because reading comprehension test scores are highly correlated with students’ amount of engaged reading than any other single factor, preparation for test taking should include support of engaged reading.  A variety of innovative teachers and researchers have shown that long-term engaged reading can be increased with integrated instruction.  When students are provided opportunities to connect reading and writing activities in integrated language arts, their engaged reading increases and reading achievement improves (Morrow, Pressley, Smith, & Smith, 1997).  Further, when reading and writing are also linked to content learning in social studies and science, engaged reading is likely to be fostered. 

Integrated curricula have been reported by Au for the KEEP program (Au & Carroll, 1997), by Gaskins et al. at the Benchmark School (1994), by Lapp and Flood in science and language arts (1994), and by Guthrie and colleagues in reading, writing, and science (Guthrie, Anderson, Alao, & Rinehart, 1999).  In all of these instructional designs, students spend 90 to 120 minutes per day actively reading to learn science and social studies.  Students use all of the strategies described previously and participate actively in writing to express their understanding. In these contexts, students learn to gain new knowledge from information texts, to interpret narrative, and to integrate information from multiple texts (Guthrie et al., 1998).

It is no surprise that the most important ingredient in high-stakes testing is difficult to teach.  If reading comprehension were easy to provide for young children, the political controversies, accountability programs, and wide spread testing systems would not be as prominent as they are.  This implies that success on high-stakes tests is not provided in a quick fix.  Rather, success is nurtured over time in integrated curricula that are planned as part of the central instructional objectives of a school.

Preparing students for test taking through motivation

Scores on any test may be low for cognitive or motivational reasons, or both.  Students may lack the reading skills or they may lack the motivation and self-efficacy to put forth maximum effort on the test (Sugrue, 1995; Waid, Kanoy, Blick, & Walker, 1978). Tests are major sources of anxiety and worry for students.  As Paris, Lawton, Turner, and Roth (1991) found in a survey of 1,000 students in four states, many middle and high school students felt resentment, anxiety, cynicism, and mistrust of standardized achievement tests.  They did not feel satisfied with their scores and believed they were not well represented by tests. This increased their anxiety and decreased their motivation for test taking.  Elementary school students were less likely to be anxious, but the ones who did experience anxiety showed low performance.  Teachers can take steps to alleviate anxiety surrounding high-stakes test taking.  Simply allowing students to see a few items and practice the item formats on a test will reduce anxiety and enable students to focus their thinking on a test (Kalechstein, Hocevar, & Kalechstein, 1988).  It is likely that providing students specific “test wise” strategies for handling the formats will enable them to feel prepared for the testing formats and thereby reduce anxiety or increase self-efficacy for test taking (Paris et al., 1991). 

Some tests are “high stakes” where as others are “low stakes.”  An example of a test that is low stakes for the student is the National Assessment of Educational Progress.  Individual scores do not impact the student’s life, school prospects, graduation, or other outcomes.  While this test is valuable for the United States Congress and the U.S. Department of Education, it is a “low stakes” event for the individual student.  However, a standardized test or performance assessment that will be used to determine promotion to the next grade or high graduation is high stakes for the student.  Teachers should discuss the particular uses for the results of a given test.  Students are entitled to know how scores will be used.  If it is a high-stakes test, students deserve to know the consequences of successful or less successful performance.  Having concrete information is likely to be informative which in some cases is anxiety reducing.

While motivation surrounding the particular assessment event is important, more critical is motivation for reading in general.  Students who are relatively low achievers, with relatively few good reading comprehension strategies, will not improve their test scores very much through mere motivation for the test.  However, these students may improve their tests scores by becoming more actively engaged readers during the school year.  It has been shown that highly motivated students in Grade 5 read 30 minutes per day for their own interest. In contrast, less motivated students read 10 minutes or fewer daily for their own interest (Wigfield & Guthrie, 1997).  The difference between 30 minutes a day and 10 minutes a day is a large amount of time. It converts to a large number of pages per week, and a large number of books per month.  If a fifth grader is facing a high-stakes test in April, the best thing the student can do is begin reading widely and frequently in October.  If a teacher will encounter a high-stakes test for her classroom in April, her best preparation is to increase motivation of students for extended, learning focused, independent reading as early as possible in the academic year.  Moving students from 10 minutes of independent reading per day to 30 minutes of learning-oriented reading per day will be the strongest test preparation that can be provided.  By fostering the students in becoming engaged readers, the teacher empowers them with a sense of self-efficacy which will shield them against the shocks of tests’ conditions and foster their ability to demonstrate their level of excellence (Lane, Parke, & Stone, 1998).

Hazards in Preparing Students for High-Stakes Tests

Excessive format practice

In a previous section, we recommended providing explicit practice on the recognition format of the multiple-choice test and constructed response formats of the performance assessment.  When they encounter a new test, students should not be surprised by the specific tasks in which they will be expected to show their reading skills.  However, excessive practice in one or two formats can be hazardous for students.  The reason is that students may “lock in” to a particular format and this will be counterproductive.  From excessive practice, students will learn particular procedures too well and apply them inappropriately.  They will over-expect these formats in one or more tests and be confused if formats change.  While some exposure is effective, excessive time in format practice is likely to be misleading because the test is measuring general abilities not specific tasks. 

Curriculum narrowing

Where high-stakes tests are heavily emphasized by administration, curricula frequently narrow.  For example, if a multiple choice reading comprehension test in Grade 3 is required for promotion of students to the next grade, teachers focus the curriculum on multiple choice-like reading activities.  They use short passages, brief questions, and recognition responses.  While the narrowed curriculum may have some short-term benefits for some students who need it, the long-term effect is usually negative.  A narrowed curriculum taught for a period of several months leads to decreases rather than increases in scores (Yen & Ferrara, 1997).  As Smith  (1991) reported, where teachers focused instruction on a limited set of specific skills that appeared to be needed by a test, student achievement was likely to decrease in the long term.  There is a simple reason for the decrease.  A test is a sample of what the student can do.  While one test may sample Skill A with Content 1 in a given format, the next test faced by a student may sample Skill B with Content 2 in a slightly different format.  Therefore, excessive practice on any single format is unlikely to have the breadth needed by students to cope with the possible samples they may face in a testing situation.  Maintaining breadth of reading objectives, writing activities, and language arts integrations are the best way to avoid the hazard of a narrowed curriculum.

Incoherent curriculum

In states that are providing performance assessments, such as Maryland, Kentucky, and North Carolina, teachers are often encouraged to give explicit practice on the tasks in the performance assessments to familiarize students with them (Shepard, Flexer, Hiebert, & Marion, 1993).  The question is “How much task practice is effective?”  While students should understand the formats of the tasks, the formats themselves cannot become the curriculum.  Unfortunately, some misguided administrators expect teachers to provide practice on tasks in performance assessments for several weeks or even several months.  In these cases, the tasks become the curriculum.  However, the tasks were designed for testing and not for teaching. Consequently, they do not serve the curriculum well.  The hazard is that the curriculum becomes disconnected, fragmented, and incoherent.  As one teacher noted in an in-depth study, “Our social studies has become a hash.”  The technique for avoiding an incoherent curriculum is to place extreme limits on test practice. Teachers and administrators should retain the goals, contents, and procedures of the mainstream reading curriculum. It is advisable to design the curriculum carefully; place priority on full implementation; and commit to success with it.

Performance orientation

In the classroom, teachers may be either learning-oriented or performance-oriented.  The learning-oriented teachers emphasize understanding of major content themes. They aim to teach reading skills that can be used widely.  In contrast, performance-oriented teachers are concerned with test scores, student achievement, and external evaluations.  Researchers have studied the effects of these two orientations.  Flink et al. (1992) asked some teachers to show their students how to understand a problem and to enjoy trying to solve it.  These teachers were expected to be learning oriented.  The researchers asked a second group of teachers to make sure students scored high on a certain problem-solving test.  These were performance-oriented teachers.  The teachers were told they would be evaluated based on their students’ scores.  Teachers in the performance-oriented group were likely to become extremely procedural, showing students specific steps for problem solving.  The performance-oriented teachers did not emphasize understanding the problem, or being creatively flexible in solving it.  Performance-oriented teachers were anxious about their own instruction.  The outcome was that students in the performance-oriented group were lower in achievement on the problem-solving test than students in the learning-oriented group.  In other words, an emphasis on performance and high scores has a paradoxical effect. In fact, it decreases successful tests taking.  This principle applies to classrooms as a whole.  In middle school science classrooms, Anderman and Young (1994) found that teachers who emphasized performance on tests and rewarded students with extrinsic incentives had lower achievement test scores than teachers who were learning oriented, emphasizing understanding and encouraging students to be pleased with their comprehension of new ideas.  A performance orientation decontextualizes reading and therefore, it is likely to decrease the effectiveness of instruction.

Planning Instruction and Curricula in a High-Stakes Testing Environment

Linking teaching objectives to testing  

Optimal planning for test preparation begins with a simple step.  The objectives of the test should be clearly linked to the objectives of instruction.  If one test objective consists of comprehending character development in narratives, then one objective of instruction should consist of comprehending character development in narratives.  This objective can be taught with stories or novels.  It can be encouraged through reading, writing, and listening.  Pursuing this objective for instruction will not result in any of the hazards presented in the previous section.  This is a broad objective, not a narrow one.  It can lead to a coherent curriculum, not an incoherent one.  Other objectives, such as gaining the main idea from a paragraph or writing to integrate across multiple texts, can be formulated as teaching objectives.  When there is an alignment between the objectives of testing and instruction, test preparation is being optimized.

In a previous section, we described the value of practicing item formats.  However, we also suggested that this be extremely limited in time and scope.  Assessment tasks should not become the curriculum.  It is not the formats of the test that should be aligned with teaching. It is the objectives that should be aligned with teaching.

Time allocations.  

As indicated previously, we recommend three levels of test preparation for high-stakes testing in reading.  At one level, students are given instruction and practice with test formats.  The time allocated to this aspect should be a few days, but never weeks or months.  It will be counterproductive if this time allocation is excessive.  The second aspect of preparation is strategy instruction for reading strategies needed by the test, such as gaining the main idea, searching for information, or self-questioning.  Time allocated to this portion of the preparation process should be measured in weeks and months.  Strategy instruction is a long term endeavor and strategy learning is time consuming.  These strategies are often reflected in the objectives of a test and the objectives of a curriculum and can be incorporated into goals for instruction planned at the beginning of the school year.  The third phase of test preparation recommended previously consists of planning for an integrated curriculum that will encourage long-term engaged reading.  The time allocated to this phase of preparation should be measured in months and years.  The planning across the grade levels within a school and across the months for a given school year should include integrations of reading, writing, and content that will enable students to gain the high level skills most important on tests.  These skills include cognitive strategies, meta-cognitive awareness, and motivation attributes.

Support for reading motivation and self-efficacy

Planning for test preparation involves generating ideas for motivational support.  Teachers need to address the anxiety students face when they encounter tests. Teachers should help students feel efficacious about their test taking abilities.  However, beyond the immediate test context, teachers can develop students' intrinsic motivation for reading during the preceding months prior to the high-stakes test.  Intrinsic motivation will increase students’ amount of reading independently and amount of reading in school, both of which will contribute to test scores.  Plans for increasing motivation and principles for classroom application have been described by Guthrie and Alao (1997).  For example, supporting the choices of students and the use of interesting texts in reading instruction are described in detail.  These motivational supports can be combined and merged with plans to accomplish the content objectives in reading and the content areas.  Designing an instructional environment that nourishes motivation is best accomplished in concert with teaching reading skills and strategies. 

Concluding Questions

If the main elements of planning for test preparation are successfully implemented, the hazards that may arise in test preparation will be avoided.  If the objectives of instruction are linked with the objectives of testing, time is appropriately allocated to the three phases of test preparation, and motivation for reading in general and testing in particular are designed into the instructional framework, the potential pitfalls in test preparation can be circumvented. 

We close with several questions.  These questions may be used to design a test preparation program in classrooms or schools.

            (1) What are the objectives of the high-stakes reading test in your classroom and school?

            (2) How can these objectives be linked and integrated with the instructional objectives in your reading-language arts program?

            (3) How will you allocate time to the following test preparation phases: a) format practice, b) strategy instruction, and c) integrated curriculum for engaged reading?

            (4) What support systems for student motivation in test taking and in becoming an engaged reader will you provide in your classroom?

            (5) Looking at the two pie charts in this chapter, one for test performance and one for test preparation, explain the similarities and differences between them.

            (6) What are the potential hazards of high-stakes test preparation in your classroom or school?

            (7) What are the different reactions of teachers, students, administrators, and parents to the high-stakes testing in your school and classroom? 

            (8) How can these reactions be used to improve success on high-stakes tests?

References

        Anderman, E. M., & Young, A. J. (1994).  Motivation and strategy use in science: Individual differences and classroom effects.  Journal of Research in Science Teaching, 31,  811-831.      
        Au, K. H., & Carroll, J. H. (1997). Improving literacy achievement through a constructivist approach. The KEEP Demonstration Classroom Project. Elementary School Journal, 97(3), 203-219.
        Baker, L., Dreher, J. & Guthrie, J. T. (2000). Engaging young readers: Promoting achievement and motivation. New York: Guilford Press.
        Blanton, W. E., & Wood, K. D (1984). Direct instruction in reading comprehension test-taking skill. Reading World, 24, 10-19.
        Eakins, D. J., Green, D. S, & Bushell, D. (1976). The effects of an instructional test-taking unit on achievement test scores. Journal of Educational Research, 70(2), 67-71.
        Feurer, M. J., Holland, P. W., Green, B. F., Bertenthal, M. W., & Hemphill, F. C. (1999). Uncommon measures. Washington, DC: National Academy Press.
        Flink, C., Boggiano, A. K., Main, D. S., Barrett, M., & Katz, P. A. (1992). Children's achievement-related behaviors: The role of extrinsic and intrinsic motivational orientations. In A. K.           Boggiano & T. S. Pittman. Achievement and motivation: A social-developmental perspective (pp. 189-214). New York: Cambridge University Press.
        Franklin, M. R. (1983). The effect of practice and instruction on the word comprehension subtest of the Woodcock. Journal of Psychoeducational Assessment, 1(2), 197-200.
        Garner, R. (1982). Resolving comprehension failure through text lookbacks: Direct training and practice effects among good and poor comprehenders in grades six and seven. Reading Psychology, 3(3), 221-231.
        Gaskins, I., Guthrie, J. T. Satlow, E., Ostertag, F., Six, L., Byrne, J., & Conner, B., (1994).  Integrating instruction of science and reading, and writing: Goals, teacher development, and assessment.  Journal of Research in Science Teaching, 31, 1039-1056.
        Guthrie, J. T., & Alao, S. (1997).  Designing contexts to increase motivations for reading.  Educational Psychologist, 32, 95-107. 
        Guthrie, J. T.,  Anderson, E., Alao, S. & Rinehart, J. (1999). Influences of Concept-Oriented Reading Instruction on strategy use and conceptual learning from text. Elementary School Journal, 99(4), 343-366. 
        Guthrie, J. T., Van Meter, P., Hancock, G. R., McCann, A., Anderson, E., & Alao, S. (1998). Does Concept-Oriented Reading Instruction increase strategy-use and conceptual learning from text? Journal of Educational Psychology, 90(2), 261-278.
        Guthrie, J. T., Wigfield, A., Metsala, J. L., & Cox, K. E. (1999). Motivational and cognitive predictors of text comprehension and reading amount. Scientific Studies of Reading, 3(3), 231-257.
        Haladyna, T. M.,  Nolen, S. B., &Hass, N. S. (1990). Raising standardized achievement test scores and the origins of test score pollution. Educational Researcher, 20(5), 2-7.
        Hansen, J., & Pearson, P. D. (1983).  An instructional study: Improving the inferential comprehension of good and poor fourth-grade readers.  Journal of Educational Psychology, 31, 821-829.
        Heubert, J. P., & Hauser, R. M. (Eds.) (1999). High stakes testing for tracking, promotion, and graduation. Washington, DC: National Academy Press.
        Hughes, C. A., & Schumaker, J. B. (1991). Test-taking strategy instruction for adolescents with learning disabilities. Exceptionality, 2(4), 205-221. 
        Kalechstein, P. B., Hocevar, D., & Kalechstein, M. (1988). Effects of test-wiseness training on test anxiety, locus of control and reading achievement in elementary school children. Anxiety Research, 1(3), 247-261.
        Kirkpatrick, J. E. (1934). The motivating effect of a specific type of testing program. University of Iowa Studies: Studies in Education, 9(4), 41-68. 
        Lane, S., Parke, C. S., & Stone, C. A. (1998). A framework for evaluating the consequences of assessment programs. Educational Measurement: Issues & Practice, 17, 24-28. 
        Lapp, D., & Flood, J. (1994).  Integrating the curriculum: First steps.  Reading Teacher, 47, 416-419.
        Lysynchuk, L. M., Pressley, M., & Vye, N. J. (1990). Reciprocal teaching improves standardized reading-comprehension performance in poor comprehenders. Elementary School Journal, 90(5), 469-484. 
        Mehrens, W. A., Popham, W. J., & Ryan, J. M. (1998). How to prepare students for performance assessments. Educational Measurement: Issues & Practice, 17, 18-22. 
        Millman, J., & Bishop, C. H. (1965). An analysis of test-wiseness. Educational and Psychological Measurement, 25(3), 707-717.
        Morrow, L. M., Pressley, M., Smith, J. K., & Smith, M. (1997). The effect of a literature-based program integrated into literacy and science instruction with children from diverse backgrounds. Reading Research Quarterly, 32(1), 54-76.
        Paris, S. G., Lawton, T., Turner, J., & Roth, J.  (1991).  A developmental perspective on standardized achievement testing.  Educational Researcher, 20, 12-20.
        Purves, A. C. (1993). Setting standards in the language arts and literature classroom and the implications for portfolio assessment. Educational Assessment, 1(3), 175-199. 
        Roeber, E., Bond, L., & Connealy, S. (1998). Annual survey of state student assessment programs. Fall 1997. Washington, DC: Council of Chief State School Officers.
        Rosenshine, B., & Meister, C. (1994).  Reciprocal teaching: A review of the research. Review of Educational Research, 64, 479-530.
        Reutzel, D. R., & Hollingsworth, P. M. (1991).  Reading time in school: Effect on fourth graders’ performance on a criterion-referenced comprehension test. Journal of Educational Research, 84, 170-176.
        Schwartz, S. (1988). A comparison of componential and traditional approaches to training reading skills.  Applied Cognitive Psychology, 2(3), 189-201.
        Scruggs, T. E., Mastropieri, M. A., & Tolfa-Veit, D. (1986). The effects of coaching on the standardized test performance of learning disabled and behaviorally disordered students. Rase: Remedial & Special Education, 7(5), 37-41.
        Shepard, L. A., Flexer, R. J., Hiebert, E. H., & Marion, S. F. (1993). Effects of introducing classroom performance assessments on student learning. Educational Measurement Issues and Practices, 15(3), 7-18. 
        Smith, M. L. (1991). Meanings of test preparation. American Educational Research Journal, 28, 521-542.
        Sugrue, B. (1995). A theory-based framework for assessing domain-specific problem-solving ability. Educational Measurement: Issues and Practice, 14, 29-35. 
        Valencia, S. W. & Place, N. A. Literacy portfolios for teaching, learning, and accountability: The Bellevue literacy assessment. In S. W. Valencia, E. H. Hiebert, & P. P. Afflerbach (Eds). Authentic reading assessment: Practices and possibilities. Newark, DE: International Reading Association.
        Waid, L. R., Kanoy, R. C., Blick, K. A. & Walker, W. E. (1978). Relationship of state-trait anxiety and type of practice to reading comprehension. Journal of Psychology, 98, 27-36. 
        Wigfield, A., & Guthrie, J. T. (1997). Relations of children's motivation for reading to the amount and breadth of their reading. Journal of Educational Psychology,89, 420-432.
        Yen, W., & Ferrara, S. (1997). The Maryland School Performance Assessment Program: Performance assessment with psychometric quality suitable for high stakes usage. Journal of Educational and Psychological Measurement, 57(1), 60-84.  

Figure 1: Components of Test Performance	Figure 2: Components of Test Preparation