Avant Assessment

STAMP is delivered online over the Internet and is not platform specific. If a computer can display worldwide web content, it can be used for the STAMP assessment.

STAMP does require an Internet connection and a web browser. It is optimized to take advantage of the most recent versions of MS Internet Explorer (IE), and it minimally requires IE 5.1 or higher or Netscape 7.0 or higher or Safari 1.0 or higher. No additional applications or plug-ins are required for the reading and writing sections and only Flash Player 7.0 is required for the speaking phase of the STAMP test.

Structured proctoring is a requirement with the STAMP assessment and therefore it is preferable that the test be delivered to an entire class at one time or split over two testing sessions. It can be administered in a computer lab setting or from any Internet connected computer, as might be found in a secondary school classroom or college language department office. There is not a technical requirement that the test be given in a group setting, and individual make-up tests or other accommodations can be done on an individual computer.

STAMP test items are developed from authentic and semi-authentic texts. Students are required to interact with these texts in a "real world” context. For instance, a student may be presented with an actual bus schedule and asked to select the appropriate route and departure time required to arrive at a particular location in time for a concert. These test items require the student to demonstrate that they are proficient using the language. In addition, an understanding of second language culture is conveyed in connection with the task and text itself.

STAMP is a statistically validated test. Every question on the test is designed to measure one, and only one, aspect of a student's foreign language proficiency. This is known as 'unidimensionality'. Any extraneous factors, or confounding variables, that might get in the way of a test item's ability to accurately measure a student's ability with regard to that single aspect of language proficiency, can completely compromise the dimensionality, and thus the validity, of the assessment.

Common factors that can compromise unidimensionality include confusing or ambiguous directions, a poorly worded question, and more than one correct answer choice. If any of these variables causes a student to perform poorly on a given item, then the information obtained from that item is compromised. We have no way of knowing if the student's poor performance on this item is a reflection of low ability with regard to the skill that this item is designed to measure, or if one of these other variables has gotten in the way of that student's being able to demonstrate his/her actual ability.

In the case of a foreign language test, directions or questions that are delivered in the target language can be another type of confounding variable that compromises unidimensionality. If a student has to do TWO things to succeed on a given item - understand the directions in the target language AND perform the required task - then the unidimensionality of that item has been compromised. If a student performs poorly on such an item, then we really have no way of knowing if this is a reflection of low ability with regard to the skill that this item is designed to measure, or if he/she simply didn't understand the directions.

Thus, while requiring students to operate solely in the target language can be a very appropriate strategy for instruction, and a key tool for effective language acquisition, it can completely compromise the accuracy of the information obtained from, and thus the validity of, any assessment.

The current version of STAMP was originally developed by the Center for Applied Second Language Studies (CASLS) at the University of Oregon with grant funding. Avant Assessment is in the process of developing the technology and test items for Web-based listening assessments with piloting targeted for Spring of 2008.

The STAMP assessment measures student achievement based upon the ACTFL proficiency guidelines in the levels from Novice-Low through Intermediate-Mid, which are identified as benchmark levels 1-5 in our system. Multiple topics are covered in each level and students are measured not only by achievement level, but data on their strengths and weaknesses by topic is also provided.

STAMP report summaries can be customized to provide specific recommended courses for each student, based upon program descriptions. This is typically done for college placement purposes. STAMP is designed to collect critical diagnostic data that college second language departments can use for placement of incoming students or for measurement of program effectiveness as students matriculate. The test items are primarily designed to establish the student’s level of achievement based upon the ACTFL guidelines for Novice Low - Intermediate Mid. In addition, STAMP test items are topic-area specific.

A discrete grammar component is available and can be added to the stamp test if desired. These scores are reported independently, as a “raw score” showing the number of items answered correctly and incorrectly.

Foreign language instruction professionals at the University of Oregon Center for Applied Second Language Studies (CASLS) developed the original STAMP test items. Those original items were then piloted and statistically verified to consistently measure student achievement at specific levels. Over 30,000 middle, high school, and college students were involved in the piloting of STAMP. Avant Assessment continues to develop new STAMP test items using the same protocols and statistical testing regimen.

Through the use of CAT technology, the delivery of STAMP test items is unique for each student tested. Mid-level questions are initially delivered with randomly ordered answer sets. As the student correctly or incorrectly answers the questions, the computer generates higher or lower level test items, establishing the appropriate level suited to each student’s capacity. Additional test items are presented to the student covering topics at higher levels to probe for areas of strengths and weaknesses. The results generated not only establish the benchmark level that the student has achieved, but also indicate additional strengths that the student possesses and topics with which the student has difficulty. Since STAMP does not overly burden students with test items that exceed or do not meet their demonstrated achievement level, the testing experience is less frustrating to them.

CAT technology also produces grading results for the reading portion of STAMP instantly (written and spoken components are scored and generally reported within three weeks after test completion, although faster reporting is available for an additional fee). See below for more on Reporting.

Although computers lack the capacity to grade the limitless variations that can be produced by students on the written and spoken portions of the STAMP test, they do play a key role in assuring consistent accuracy on the part of our graders. Each STAMP grader undergoes a training regime originally designed by CASLS. After receiving formal training, potential graders, who are native speakers or language instructors, are asked to grade an extensive battery of sample items to establish that their grading standards are aligned with test requirements. Only after this process is successfully completed, are graders allowed to score actual student-generated writing samples. The computer system then continuously monitors our test graders by presenting test items that are internally generated to assess consistency on the part of the grader. Graders who fail to appropriately score the internally generated items are required to retrain before continuing to grade student writing and speaking samples. All writing and speaking samples graded as failing are automatically re-graded by a second, and if necessary a third rater to confirm the scoring.

The best way to prepare for the STAMP test is by using the ClassPak online instructional support tool to create and assign online formative tests that look and feel like the STAMP test to students. Additionally students can take the STAMP demo to familiarize themselves with the Web-based testing experience. While the presentation of authentic items may be new to some students, the proficiency skills measured by STAMP will accurately reflect the skills the student has acquired regardless of the manner in which the curriculum has been presented. Students with the most basic of computer skills will easily understand the STAMP interface. No specific test preparation is needed.

ClassPak is an online, data-driven instructional-support-tool that helps teachers build, deliver and manage reading, writing and speaking quizzes, and lessons, assignments and activities online. ClassPak is pre-loaded with STAMP-like, realia-based content and assists teachers in creating a more proficiency-oriented classroom. It also can familiarize students with the STAMP test experience, since ClassPak quiz items resemble STAMP test items. ClassPak saves teachers the work of searching for authentic texts to use in lessons and quizzes, while affording them the flexibility to use pre-set test items, or to customize them utilizing the large database of authentic texts. Using ClassPak regularly often results in higher proficiency levels among students tested with STAMP.

There is no time limit on the test, so exam takers are not pressured, and can read instructions at their own pace. You may impose a time limit if you wish, and we can make recommendations for what that limit should be.

Immediately after the exam is completed, the results of the STAMP reading test are available online at http://report.stamptest.net for authorized personnel. Reports include individualized test results that certify specific benchmark level achievement and also indicate specific topic area strengths and weaknesses exhibited by the student’s reading test responses. The writing and speaking sections of the reports include Web links to each individual’s writing response, and audio access to each speaking response.

In addition, reports are furnished at the class, school, and district or college/university level, which enables analysis of how instructional efforts are proceeding at various levels. STAMP reports also help identify where resources may best be directed. Authorized access to varying levels of reports is determined by our partners during the STAMP service setup process.

Avant Assessment will provide STAMP test results that identify an individual only to the teacher of the student and administrators of the school, district or college. Avant will provide results from district-wide tests that identify an individual only to administrators of the school, district or college. We will use the aggregate individual results to calculate a national score for each topic tested without identifying the names or scores of individuals. In no event will test results that identify an individual be released to anyone other than those identified above.

Avant Assessment will only release test results to the individual school, district, or other school department or college that has retained Avant. Aggregate test results from all participating schools, districts and colleges using STAMP will be used by Avant for statistical analysis and to help Avant identify and share achievement trends information with the professional education community. All data will be anonymous. No data used for research purposes will be tied to student identities.

Personal information will only be used by Avant Assessment for test delivery purposes and will not be provided to any third parties.

Avant Assessment does not utilize cookies in any of its testing services. We may use cookies to identify users that visit our information and marketing web site. Please note that we do not control, nor have responsibility or liability for the content, privacy policies or security policies of any of our business partners or sites to which we provide hyperlinks in our information and marketing web site.

Please our headquarters toll free at (888) 718-7887 or send an email to info@avantassessment.com if you have any questions.

Avant Assessment has developed security measures to mitigate the risk of exposure of our services, including STAMP, to theft, manipulation, loss and vandalism. We have installed appropriate firewalls, password systems, and other software and hardware systems designed to thwart such possibilities. Furthermore, we have instituted internal management systems to protect the integrity and privacy of the data that we collect. System penetration tests are regularly conducted by third party security consultants to simulate attack on our systems and assure sufficient safeguards are in place. However, no system that relies on the Internet for the administering of tests and the transmission of data can be guaranteed to be 100% secure. We continuously monitor our efforts and rely on the cooperation of our partners/clients in adhering to proctoring guidelines to ensure that STAMP is a reliable and secure proficiency test.

The STAMP test is a computer adaptive test, assuring that each student receives a test that customizes itself to each student’s ability level. This makes it highly unlikely that two students will encounter the same items in the same order on a test. However, in addition, our computer servers draw questions from a deep pool of test items, and display questions randomly so that students who may be tracking along the same leveling path will still be asked questions in a different order and will almost certainly receive a different set of questions. Moreover, on the reading test even if two adjacent students were to receive the same question at the same time the order of the four answer choices displayed within any given test item is also always randomized.

So called grabber programs are difficult to completely defeat, however, there are a number of techniques that we have employed to substantially diminish the ability of students to use these programs successfully to copy questions. We can describe these in more detail to administrators who are concerned.

We have protected our servers with hardware firewalls to protect against hacking.

We have created a set of proctoring guidelines that instruct test administrators how to create and maintain a secure environment locally for the tests. In addition, passwords for accessing the test are provided to students at the beginning of the test and expire as soon as the students have completed their test and logged off to prevent them from accessing the test at a later time. Nevertheless it is impossible to entirely prevent the leakage of items from the test. Therefore, we are continually creating and statistically validating new STAMP test items to expand the number of test items in our item database. We also periodically remove items that have been in the database for a prolonged period of time.

There are a number of issues that can cause disruption in the delivery of an online test to large numbers of students. The STAMP servers have been "load tested” to ensure that they can deliver the tests as scheduled. It is required that the school run through a diagnostic setup process well in advance of testing to ensure that computer settings are correct and that headsets are operating properly for recording and playback during the speaking section of the test. This diagnostic can be accessed here: http://www.avantassessment.com/diagnostics.php.

Generally, students are able to complete a three skills STAMP test in approximately 60-75 minutes. Avant Assessment places no time limits on the test-taker, although we report the length of time the student took to complete. It is up the test provider at the local level (the school, district or college) to set a time limit, if so desired. During the Speaking portion of the test, STAMP will time-out if there is no submittal for 10 minutes. (In that case, the student should simply log-in again to proceed from where they left off). Please note that to accommodate scheduling needs STAMP is designed to enable students to stop in mid-test, and resume the test at later date, within 14 days of the individual student’s test initiation. Our computer server will store all partially completed tests and allow the student to resume from the point at which he or she logged out until the 14 day window has closed.

For security reasons, to prevent cheating and protect the integrity of the test, STAMP does require a proctor’s presence at all times while students take the test. A proctor may be any responsible adult, providing he or she will have read the proctoring guide, and agreed to observe the protocols outlined therein.

The reading section is automatically scored instantly. Scoring for the writing and speaking sections is reported within three weeks. Faster reporting is available where required, for an additional fee. A price sheet is available in the STAMP section of our Web-site.

All results can be viewed by authorized personnel immediately on the password protected STAMP reporting website at http://report.stamptest.net. Authorized teachers/administrators wishing to view the results will use the login codes that were provided to them when we transmitted the login codes that were used by students to take the STAMP test.

There is no training required. However, proctors must agree to review and adhere to the proctoring guidelines, including assuring the STAMP computer diagnostic is run in advance on each computer and headset to be used for testing. Avant Assessment will not be responsible for incomplete tests resulting from proctor failure to follow STAMP set-up and proctoring guidelines, including pre-test computer and headset preparation to ensure that computer settings are correct, and that headsets are functioning properly, i.e. that the record and playback functions work. Proctors are also responsible for providing the required information to students before testing, and to act as a facilitator during testing, in case students do not understand what is required of them, or if a computer encounters a problem during testing. Where possible, if a problem occurs, proctors should attempt to solve it locally, and then failing that, during testing they should contact our help desk toll free at (888) 718-7887 or email support@avantassessment.com.

Please visit the Educational Partners section of our Website under About Avant. Of course we would be happy to provide personal references to state-level administrators as well as school, district, or college instructors who use STAMP. Please e-mail us at info@avantassessment.com or call us at (888) 718-7887 with your request for referrals.

Prices are per student/test, including all skills indicated and all scoring and reporting. The STAMP speaking component requires a headset; Avant offers durable, competitively-priced headsets as a service to its educational partners. Please see the pricing section of the STAMP page on our Website www.avantassessment.com.