Os IGNORANTES, que acham saber tudo, privam -se de um dos maiores prazeres da vida: APRENDER.
Início Avaliação da Aprendizagem EVALUATING STUDENTS

EVALUATING STUDENTS

 

GILBERTO  TEIXEIRA (D.B.A.)
 
 
Throughout your years as a student, you have experienced a variety of educational assessment procedures that teachers used to gather information and make decisions about the degree to which you had achieved their learning objectives. They also used this assessment data to make decisions about their teaching and to assist them in finding approaches that would help students succeed. As you now assume teaching responsibilities, you will need to think about the methods you will use to provide feedback to students and to yourself. The large number of assessment techniques available to you range from informal observations to more formally gathered data from tests, homework assignments, lab work, papers, projects, and oral presentations. In planning your course, it is important to consider the kinds of assessment you will include. You must also decide how much weight you will assign to each of the assessments that will contribute to the final grade. This section will introduce several topics related to the evaluation of student performance, an important aspect of teaching. Even if you do not have primary responsibility for teaching an entire course right now, you will probably be asked to assist with tasks related to the assessment and grading of students.
When you have primary responsibility for teaching, it is essential to communicate to your students at the beginning of the course your course objectives and the manner in which you will assign grades. It is best to do this both on the written syllabus and in your oral introduction at the first class meeting. Students are concerned about how their grades will be determined, and you can do much to relieve their anxieties about the course by making your standards clear.
Assessment results can be used both formatively (to diagnose learning needs and help plan instruction) and summatively (to evaluate student learning after one or more units of a course). Although you will undoubtedly use many different sources of information, teacher-made quizzes and tests play a large role in the evaluation of student performance. For this reason, and because test construction is probably not an area that you probably have encountered in your program of studies, some suggestions for preparing good tests will be introduced here. There are many excellent sources for you to further your study of this aspect of teaching, and several will be mentioned below.

TEST CONSTRUCTION AND SCORING

 
As Nitko (1996) and others emphasize, [1] the following three principles are important to follow when constructing classroom assessments:
  • Focus on your teaching and learning objectives.
  • Develop your assessments so that they elicit from your students only the knowledge and performance that are relevant to your teaching and learning objectives.
  • Write items that do not inhibit or prevent the student’s ability to demonstrate attainment of your learning objectives.
When you construct an assessment, you should ask yourself whether each item measures something that you really think is important for students to know. Even if a test is difficult, students will consider it fair if it covers the learning objectives and content you have stressed. If students have mastered the learning objectives, they should be able to perform well on your tests. Likewise, if students are having difficulty with the learning objectives, these deficiencies should be reflected in their test scores. If items are ambiguously or poorly worded, if directions are not clear, if examples use material that is not familiar to all students, then some students who have attained your learning objectives may still answer items incorrectly. If you are aware of and make an effort to follow the three principles, you will develop tests that have the desired technical properties referred to as reliability and validity.

PLANNING YOUR TEST

The first step in test construction is to develop a test plan. This is sometimes referred to as a "table of specifications" or "blueprint" for your test. It serves to help ensure that the test will be a valid representation of your learning objectives and that the various content areas covered by the test reflect the emphasis you have placed on them in your lectures, readings and class discussions. By listing the content areas/topics and the percentage of emphasis each will have, you develop a structure for your test. You may also want to consider the level of understanding you want the students to have about each topic. For example, do you want students to be able simply to recall information, or do you want them to comprehend or apply the information to a novel situation. You may also have written some teaching and learning objectives for your course that require students to analyze, synthesize or evaluate. (For more information on instructional objectives and learning skills, see Chapter 10.) You can create a matrix by listing your content areas/topics as rows, and levels of the thinking process (such as recall, comprehend, apply, analyze, synthesize or evaluate) as columns. When you start to write or select items from other sources, you should follow your plan and the resulting sample of items will be representative of your course content and emphasis.
This planning before the actual writing of test items is very important, but it can be simplified so that it does not become a burden for you. For example, you could use "recall," "comprehension," and "higher levels" as the categories of learning. Each time you teach the course, you can make adjustments and refinements to your test plan.
Developing a test plan also helps you determine the types of items you want to write for your test. Your choice between essay items and objective items depends on the types of objectives you have developed, the level of thinking you wish to test and, practically speaking, on your class size. You may want to combine several types of items in your test. Essay items are so named because of the need for a written response. The term "objective," however, refers to the scoring procedures used rather than to the type of response given. If a student receives the same score regardless of who scores the test, the test is said to be objective.
Objective tests allow you to sample adequately from a large amount of course material, thereby increasing content validity. They are easy to score, but it does take time and skill to write effective items. Item formats include true/false, multiple-choice and matching. One common criticism is that these items often test only basic recognition, but test developers have demonstrated that effectively constructed items can also be used to test higher levels of thinking.
Essay tests include items requiring short answers and/or more extended responses. It takes less time to develop essay test items, and they give students an opportunity to demonstrate skills of synthesis and evaluation as well as organization and expression. However, the scoring of essay items is time-consuming and must follow certain guidelines in order for the test to be reliable (i.e. others reading the essay would assign the same score as you). Remember that it is important to consider your test plan in deciding on the type of test items to use.
Depending on your discipline, you may use problems as test items. They may be categorized as either objective or essay items, depending on the scoring system that is used. If the problem is simple and you are scoring only for a correct or incorrect response, it would be considered an objective test item. However, many problems involve complex solutions requiring several steps and perhaps allowing alternative approaches to their solutions. These problems are more difficult to score and they present the same reliability challenges as the essay item.

WRITING TEST ITEMS

It has been said by measurement specialists that writing good test items is a combination of science and art. There are a number of guidelines that should be followed to produce the qualities desired in your items. Even when these suggestions are followed, however, some item writers are better than others at creating clear, concise items that measure their specified learning objectives. Item writing will become easier as your teaching experience increases and you learn from the feedback provided by statistical data generated from your tests and from student feedback. There are entire chapters in measurement textbooks covering the guidelines for writing various types of objective and essay test items).
 A brief summary of these guidelines is given below.
If you are planning an objective test, you may have several sources of previously written items available to you. Some textbooks supply items that have been developed for each chapter. Other instructors who have taught the course may be willing to share their items with you. In either case you will have to review carefully the items and adapt them to your specific test plan. Published items are not always well written and even those that are may not cover concepts that you consider important. Remember the three principles mentioned above when you adapt or write your items.
The most widely used type of objective item format is multiple-choice, consisting of a stem that poses the question/problem and a series of options, each representing possible answers to the stem. Suggestions for writing multiple-choice items include, but are not limited to the following:
  • Select content related to important aspects of the course objectives.
  • Have the stem clearly present the problem and include all the words that would otherwise need to be repeated in each option.
  • Avoid irrelevant information in the stem or options, unless deciding what is necessary information is part of the learning objective.
  • Have only one correct response. (This is essential if your test is going to be scored electronically by OMET. See the section on scoring.)
  • Do not provide verbal clues that might enable students to eliminate options or select the correct one.
  • Make all options grammatically consistent with the stem and also parallel in form.
  • When there is a logical sequence in which alternatives can occur (such as time or numbers), use it to structure the multiple-choice options; answer choice "A" might be 1800, "B" might be 1812, and "C" might be 1819.
  • Formulate the correct option so that it is not consistently different in appearance from the incorrect options. (Often there is a tendency to include more information in the correct option, which makes it longer).
  • Try to have at least four options per item unless doing so requires using implausible options.
  • Be very careful in using "all of the above" and "none of the above." If students perceive one answer to be truer than another answer, they may be reluctant to select an all-or-nothing option. You want your test to reflect accurately your student’s understanding of the material.
  • If you use a negative ("Which of the following is NOT . . .") in the stem, emphasize it by capitalizing and bolding it so students will notice it.
  • Vary the position of the correct answer in a random manner and have each position (A, B, C, D) represented as equally as possible.
Make sure that you give yourself enough time to write your test items. It takes time to write good test items, and it is helpful for you to review the entire test after several days have elapsed. It is a good idea to proofread and develop the answer key for your objective test before printing copies. This will allow you to catch errors that might otherwise require writing corrections on the board the day of the test or even having to drop items from your test (if, for example, you discover you have no clearly correct answer or more than one correct answer). Check the numbering of the items as well to make sure it is accurate.
When writing essay items, try to:
  • Keep in mind whether the learning objective you are measuring requires the higher-order thinking process of analysis, synthesis or evaluation. This will help you with the wording of the question.
  • Write the question in such a way that the task is clearly defined for the students.
  • Start the essay item with words such as " compare," " contrast," "give reasons for," "give examples of," "criticize," "differentiate," "explain how."
  • Be sure the essay question asks for the specific behavior that you want the student to display.
  • Adapt the length and complexity of the question to the level of the students.
  • Provide an indication of how questions will be weighted by giving the number of points for each question. (This allows students to plan their time.)
With essay items, it is also a good idea for you to specify the ideas that you will be looking for when scoring if you are going to use an analytic scoring method (see explanation under Scoring). This will assist you in determining the total number of possible points for each question and will enable you to make adjustments if the weighting is not what you had intended. Even if you are going to use a holistic approach to scoring, you can describe the quality of the papers that you plan to put in each scoring category.
If you are developing problem-based test items consider the following:
  • Develop problems that test students on the strategies and concepts they have been taught in ways similar but not identical to the problems that they have covered in class or homework assignments.
  • Don’t let the setting for the problem become so complex that your scenario will confuse students who know how to solve the problem.
  • Review your problem to make sure that all of the students will have the contextual knowledge assumed.
  • Work out the solution yourself to help you decide how conceptually complex it is. If there are a number of sub-problems embedded in the problem, you may want to provide some of the assumed information such as formulas.
  • Think about how much time students will need to think through and solve the problem. (This will be longer than the time it takes you to solve the problem since you are much more familiar with the concepts being tested.)
In writing test items, you must consider the question of test length. Consider both the length of the class and the reading level of your students. You do not want students to feel rushed and frustrated because they were not able to demonstrate their knowledge of the material in the allotted time. Some general guidelines regarding time requirements for high school student test takers (as reported by Nitko in1996) are as follows:
 
Task
Approximate Time Per Item
True/False items
20-30 seconds
Multiple-choice (factual)
40-60 seconds
Multiple-choice (complex)
70-90 seconds
Matching (5 stems/6 choices)
2-4 minutes
Short-answer
2-4 minutes
Multiple-choice (with calculations)
2-5 minutes
Word problems (simple math)
5-10 minutes
Short Essays
15-20 minutes
Data analysis/graphing
15-25 minutes
Extended essays
35-50 minutes
If you are combining multiple-choice and essay items, these estimates may help you decide how many to include. One mistake often made by first-time teachers is having too many questions for the time allowed.
Once your questions are developed, make sure that you include clear directions to the students. For the objective items, specify that they should select one answer for each item and indicate the point value of each question, especially if you are weighting sections of the test differently. Students must use pencils if they are responding using the NCS optical-scanning sheets for machine scoring. For essay items, indicate the point value and suggested time to be spent on the item. For problems, indicate whether partial credit will be given or only the correct solution will be awarded credit.
If you are teaching a large class with close seating arrangements and are giving an objective test, you may want to consider administering several versions of your test to decrease the opportunities for cheating. Multiple versions have become quite easy to develop with the availability of word-processing software. You simply create versions of your test with different arrangements of the items. It is not even necessary to change the position of the correct response for each item. You can also then print these versions on different colors of paper so that it is obvious to students that they have a different test than the person sitting beside them. (Some instructors simply use different colors of paper without changing the order of test items, but it is best to do both.)

TEST ADMINISTRATION

On the day of a test, it is a good idea to take extra pencils with you if you are using the NCS answer sheets for your objective items. Many departments have pencils available to you for this purpose. If your class is large, try to seat students in every other seat and have every row the same so students are seated directly behind one another. You may also want to talk to faculty in your department who have taught large classes to see how they have handled the potential problem of students taking tests for others. Many faculty members announce that students must present their IDs at the completion of the test when they submit the tests and answer sheets. Announcing this in class ahead of time and even including it on your syllabus will deter students from asking others to take the test for them. Even if your class is small and/or you are giving an essay test, taking precautions against cheating and making clear to students that this will not be tolerated will create a fairer situation for all students. Cheating on an exam is unethical and a violation of the University’s policy on Academic Integrity.
Some instructors may wish to post the scores for tests outside their offices or outside of the classroom. You need to be aware that posting scores by name or by social security number is not allowed under federal law unless the student has given you permission,. It requires students to give permission for scores to be posted by social security number by filling in a specific circle on the answer sheet.. Alternatively, some faculty use a unique number that they assign to each student.

TEST SCORING

If you have given an objective test, you may have your test electronically scored by the Office of Measurement and Evaluation of Teaching (OMET). A description of the service is available at OMET. Answer sheets are provided at no cost to you and can be obtained from OMET any time before your test. (Many departments maintain a supply of OMET answer sheets as well.) After your test, you submit to OMET the student answer sheets and a key that you have recorded on an answer sheet. Multiple versions of the test are submitted separately with a key for each version, but you can complete a form indicating the item order if you would like OMET to combine all versions in the printout described below. You will be asked to complete a cover sheet with identifying information about your course and the length of your test. Tests submitted by 5 p.m. are available for you to pick up after 10 a.m. the next working day.
In addition to the scores printed on each answer sheet, you will receive an item analysis printout that summarizes the test scores for the class, lists individual scores for your record book, and also provides you with valuable information about each of your test items. By examining the statistics related to item difficulty and item discrimination, you can identify problems students are having with items, items that need to be revised before you use them again, and items that worked in terms of discriminating between those students who knew the material well and those who didn’t. It also can include an optional list of the items that each student incorrectly answered, which will help you answer questions that students may have about their performance. OMET stores a computer record of the actual responses that students made to each item on your test and keeps that record for one year. If you are handing the answer sheets back to students, you may want to mention this to discourage students from trying to cheat by changing answers or filling in items they had left blank while taking the test. Attempts to cheat in this way have never been successful because of the computer record.
As you can see, scoring of objective tests is quite simple. Scoring of essay tests and problems takes considerably more time and you have to worry about being consistent across student responses. This is why essay items should be limited to those learning objectives that require the essay format. Once you have used an essay item, it is your professional responsibility to score the responses properly and to provide useful feedback to students.
There are two methods of scoring essay items, the analytic approach and the holistic approach. To use the analytic method, you develop a list of the major elements you believe students should include in the ideal answer. It was suggested above that this be done before the test is printed so that adjustments can be made if necessary. If you are assisting a faculty member in a course and have been given responsibility for scoring essay items either you or the faculty member has developed, you will need to meet with him/her to discuss the points to be assigned to the responses.
You may find that the analytic approach is easier to employ when the essay item requires a restricted or short answer. You can decide to give partial credit for each element in your list or simply to give full or no points for each item. This may be difficult to do if you are teaching this material for the first time since you will not be aware of the types of answers students will give. You may want to read a sample of the papers to check on your scoring scheme and make necessary adjustments. If you keep track of how students perform on each element in your list, you can identify the points that are weak and need to be reinforced in your teaching. If you provide the students with the elements of the ideal answer, they can gain information about their strengths and weaknesses.
The holistic approach to scoring essay items involves your reading an entire response and assigning it to a category identified by a score or grade. You can develop a description of the type of response that would illustrate each category before you start, and then try out this draft version using several actual papers. After reading and categorizing all of the papers, it is a good idea to re-examine the papers within a category to see if they are similar enough in quality to receive the same points or grade. It may be faster to read essays holistically and provide only an overall score or grade, but students do not receive much feedback about their strengths and weaknesses. Some instructors who use holistic scoring also write brief comments on each paper to point out one or two strengths and/or weaknesses so students will have a better idea of why their response received the score it did. Again, if you are assisting a faculty member who is responsible for the course, you might check with him/her after both of you have read several papers to ensure that your scoring is consistent.
Regardless of whether you decide to use an analytic or holistic method of scoring, there are several guidelines to consider with respect to the scoring of essay tests:
  • Develop scoring criteria or model answers.
  • If there are several essay questions, score all of the students’ responses to one question at a time. This improves consistency and reduces any "carryover" effects.
  • Decide whether you are going to score factors other than content, such as spelling, language usage, and organization. Make sure that your students are aware if you are, and give a separate score for these factors.
  • Score essay responses anonymously to avoid any bias resulting from your familiarity with students and their previous work. One idea is to have students write their names on the back of the answer sheet or booklet.
  • Periodically check to see whether you have applied the criteria in the same way to later-scored answers as to earlier-scored ones.
  • If you have a large number of papers to score, stop when you get tired so that your frame of mind will not cause your scoring to be inconsistent. When you start again, read over the last few papers you scored to be sure that your scoring was objective.
  • Provide students with feedback so that the test provides an opportunity for students to learn their strengths and weaknesses. You can provide short written comments or verbal feedback to students in a brief conference.
It is obvious that you will need to spend a great deal of time and effort to do a thorough job in scoring essay items. You may want to start out by using short-answer essay items and as you feel more comfortable with the process, refining your items to require a more extended answer. Use essay items judiciously; consider whether you can test the learning objective you have set with complex multiple choice items instead. If you decide that essay items are the best way to test student understanding of your learning objectives, give yourself enough time for the scoring of the test so that your information about students will be as reliable and valid as possible.

MATH AND SCIENCE ASSIGNMENTS

Math and science classes pose special challenges for teaching assistants. Although the general principles outlined in Dr. Carol Baker’s essay on grading apply to almost any assignment, grading problem-solving exercises (such as math or logic problems, computer language assignments, etc.) and lab reports often require that the instructor assign partial credit for process. In both cases, the student may understand how to solve the problem but may make simple mistake that lead to an incorrect answer. There are no strict rules for assigning partial credit except that graders should have clearly established criteria that has been shared with students and is applied equitably. Teaching assistants in math and science courses should provide students with a written statement of their grading criteria before the first assignment or lab is due.
In the following sections, we provide some general advice for grading lab reports, problem-solving exercises or numerical assignments, and we offer some advice on preventing cheating. Teaching assistants are encouraged to speak with their supervising instructor and their peers about the specific standards required in their discipline.

GRADING LAB REPORTS

At first glance, lab reports seem easy to grade. Since lab reports follow specific guidelines, it is not difficult to judge whether the basic requirements have been completed. If the student has finished the experiment and has submitted an accurate lab report, the student has met the fundamental requirements of a lab. However, not every experiment works out and not every lab report equally distills the larger themes from the experiment. When grading lab reports, teaching assistants should ask the following:
The real key to grading a lab report is to establish standards in advance (preferably in a written lab syllabus) and to let students know how their work will be graded. Students need to know how much weight has been assigned to performing the actual experiment, to getting a correct result, to exploring alternative approaches and explanations, and to recognizing the theoretical implications of the experiment (regardless of the success of the lab)? Although lab instructors apply their own standards, keep in mind that the overall objective is for students to experience the scientific method first-hand and to apply the theoretical models that they have learned in the lecture class.
While traditionally lab grades are based on two criteria, completion of the experiment and a written lab report, it is possible to grade students on other observed aspects of the lab. Teaching assistants may consider student’s preparation, their ability to perform the lab techniques, their understanding of the procedures, and their observance of safety standards. [2] When using observation to access grades, it is especially important that you establish clear guidelines and keep accurate notes on students’ performance.

GRADING NUMERICAL PROBLEM-SOLVING

Numerical problem solving, everything from an engineering exam to a math quiz, poses special considerations for teaching assistants. Partial credit is often given to students who follow the correct procedures even if they ultimately get an incorrect answer. Once again, establishing clear guidelines and making your grading criteria available to your students avoids justified frustration on the part of your students and can make grading easier. The following advice may be helpful:
Take the Exam. Taking the exam has two advantages. First, if you can not solve a problem, it may indicate that the problem is poorly worded or is not appropriate for the class. Keep in mind that questions posed on an exam should reflect the teaching objectives of the course
Second, taking the exam will help you to establish grading criteria. Are the formulas employed simple? If so, how much credit should you assign to a student who identifies the right formula or procedure, but does not know how to use it? Is there only one answer? Does the math (or other task) require multiple steps creating natural divisions to assign grades to?
One Standard. Even before the exam is distributed, the instructor and the teaching assistant (or teaching assistants) should meet and discuss the criteria that will be used for grading. All graders should use the same guidelines so that students are graded fairly.
Patterns. As a teaching assistant you may not have written the exam and the grading criteria may not be your own. However, watch out for patterns that may indicate that an exam problem was poorly written. If all of the students in the class are making the same mistakes, bring the situation to the attention of your supervising faculty member. If a disproportionate number of students have come up with the same wrong answer, it is often a sign that the question was poorly constructed or misleading.
Alternative Solutions. Your grading criteria will probably be based on how you yourself solve the problem. Often the procedures that you, other teaching assistants and the professor work out are the most efficient and even most appropriate means of answering the question. However, there may other acceptable answers. In advance, determine how alternative correct answers will be handled and be alert to detect creative solutions so that credit can be assigned fairly.
Take Breaks. Grading 200 problems in a night, or even a week, can be difficult. If you are grading numerical answers, grading fatigue can lead to serious mistakes. Take breaks. Only grade papers when you are alert.

WRITTEN ASSIGNMENTS

DESIGNING WRITTEN ASSIGNMENTS

There are no hard and fast rules for designing a written assignment. A writing assignment can be as tightly structured as a lab report, where specific rules guide the students in formatting and presenting information, or as loosely structured as a creative writing assignment where a poem or graphic essay might be considered acceptable. Many written assignments, however, share the same objectives. Papers demonstrate students’ knowledge of the subject matter, their capacity to evaluate information, and their ability to craft an argument creatively and logically. In other words, papers often test the depth of a student’s understanding of the subject matter.
Be creative when writing essay or paper questions, but be realistic. Consider your teaching objectives and what your students can reasonably be expected to know or to figure out. Keep the language clear, precise and simple, but include enough information that the student is given a basic outline of your expectations. Avoid using phrases that could be misinterpreted and avoid topics that cannot be answered within the page length and time you have set. For more information on developing paper topics, see Dr. Carol Baker’s discussion of essay questions in the previous section.

GRADING PAPERS

No matter how well a question is worded and how many guidelines are offered, no two papers are ever alike. Two very good writers may marshal the same facts and come to the same conclusions, but how they present their argument—the weight given to the evidence and the stress placed upon the conclusions—are bound to be different. Grading written assignments, then, is an inherently difficult chore. Two very different papers may both be A papers. Jim, a University of Pittsburgh teaching assistant, described his frustration:
I read all the papers before I actually begin to grade them and sort them into piles. All the As here, the Bs there, the Cs . . . Usually after reading thirty or forty out of sixty papers, I have a pretty good idea what is a good paper and what is a bad paper. But there is always one quirky paper that spoils my system. It is too short and needs to be edited, but it makes a point, answers the question and does so in a way that no one else in the class has thought of. Now which pile do I put that paper in?
Most of us feel some discomfort with our own writing so perhaps it is not surprising that we are reluctant to critique and criticize others. [4] Yet when taken seriously, grading writing assignments is an important part of teaching. Since few of us are born great writers, the feedback we receive on our written work teaches us how to organize, present and communicate our ideas. And good writing is not "just an academic concern;" it is a practical skill that will help your students achieve success in almost any field that they explore.
Each discipline has its own style of writing and, as a result, its own standards by which papers are graded. If this is your first time grading, you should sit down with your teaching mentor or a fellow teaching assistant and discuss writing in your discipline and what is typically expected of undergraduates in your department. Bring a few papers along so that you can both read them and then discuss specific issues that the papers raise.
In the following section, we provide a guideline for grading papers. This checklist walks you through the process of grading papers. In addition, we have provided two online handouts. The first, "Grading Standards," offers criteria for assigning letter grades to papers. The second handout, "Paper Evaluation," is an evaluation sheet that you might use to insure that you are giving your students thorough feedback. Although your grading standards and assessment tools should be fitted to the course, your teaching objectives and the department you are teaching in, these may serve as starting places for thinking about what is an appropriate grade.

GUIDELINES FOR GRADING WRITTEN ASSIGNMENTS

Establish Clear Criteria for Your Students. If you do not provide your students with a clear guide to your grading policies, few of them are going to meet your expectations. Your syllabus should contain a detailed written description of the assignments or you should provide your students with a handout stating your policies. Even if you are not the one who designed the syllabus and the assignments, if you are the one who is grading the papers you should let your students know what your expectations and guidelines are. For more information on grading policies, see Dr. Carol Baker’s comments on grading essay exams in the previous section.
Graders Unite. If more than one person is grading papers, you need to establish a consistent standard for your grading. One simple way to do this is for each of you to grade five papers and then to trade papers. Would you have given the same grades? If not, why? As you discuss the differences in your grades, outline a rough standard for the rest of your grading. When you are done, you should each have a set of matching criteria for assigning grades. If as you are grading, you come across a paper that does not seem to fit the criteria you have established, assign a grade and then pass it to your fellow grader for a quick review.
Some instructors believe that when two teaching assistants are grading papers in a class, they should grade each other’s recitation. This would seem to establish a greater level of objectivity since you probably do not know the students in your co-teaching assistant’s class. But the approach also has some significant disadvantages. It is seldom that two teaching assistants present exactly the same material. If the material you presented in your recitation is different from the material your coworker presented, you may find it difficult to grade the other teaching assistant’s recitations. In addition, your students have adapted their writing (as good writers should) to their audience. The verbal feedback you give in class helps to build students’ expectations of how their papers will be graded. Switching graders can put your students at an unfair disadvantage.
Read All the Papers First. Before you start to grade, read all (or at least most) of the papers through once. Reading the papers will help you to establish the range of responses and to distinguish the best papers from the worst. Many graders, like Jim, even sort the papers into rough representative piles as they read through them. Then they grab the B paper pile and assign grades with respect to the other B papers.
Be Objective. Needless to say, you should not allow your personal feelings about a student to influence your grading. What is often more difficult, but less obvious, is that you should not let your personal feelings about the position taken in an essay or paper influence your grading. If you hate capital punishment but your student just wrote a clear, well-argued essay advocating the electric chair, you should give it a good grade. When grading, you should focus on the quality of the argument not the specific values the student advocates. Thus, you are looking at how the student addresses the reader, explains his or her position, establishes the facts and draws conclusions.
Be Consistent. If one student lost three points for not mentioning new research into the optic nerve, every student who missed that argument should lose three points. Sometimes, after forty or so papers, keeping track of these little deductions can be difficult. Many graders keep a "diary" in which they make notes about grading decisions so that they can apply fair and consistent standards across a large number of essays especially when those essays are graded in three or four different sittings.
Provide Clear and Concise Written Comments. There is nothing more frustrating for a student than to get a paper back with two or three scribbled and indecipherable comments in the margins. The grade then seems arbitrary and the conscientious student is left with very little advice on how to improve his or her writing for the next assignment. Clear, concise comments that offer advice on how to remedy mistakes are the most useful to students.
Place short notes on grammar or style in the margins near the text you are referring to, but save longer notes on the presentation, organization and argumentation for the end of the paper. Until you finish reading the paper, you are not going to be able to make comprehensive remarks. Many educators believe that "holistic" comments on errors that suggest illogical or incomplete thinking, rather than picky corrections of minor mistakes, are more useful for developing students’ writing skills. If you want to comment on a specific section of the text, place a symbol in the margin and then reference it in your closing comments.
Provide Corrective Comments. It is not enough to tell your students that they made a mistake; you also need to explain to them how to make their writing more effective in the future. Perhaps one of your students has demonstrated a clear knowledge of the subject you are teaching, but in his papers he consistently makes arguments without drawing conclusions. You might do the following:
 
Explain the problem to the student. "Joe, you seem to know the material, but you have not drawn any conclusions from the evidence."
Select an example of the problem and explain why it is a problem. "This was a problem throughout the paper, but let us look at one example. On page three you provide a number of quotes that demonstrate that Freud had a low opinion of women. Good, but what is the effect of Freud’s bias? Does it mean that we should take Freud’s theories less seriously or can we trust that modern psychology has routed out these biases?"
Offer a solution. "Joe, the paper is weak unless you draw some conclusions from your analysis. You might, for example, explain how Freud’s bias taints the foundations of his theory. I might conclude the section you wrote by saying: ‘Freud’s bias can not be simply dismissed as a product of early 20th century chauvinism since it fundamentally influences the psychology that is practiced today. Freud’s entire theory of infant development, a cornerstone of this larger work, is . . .’" Avoid rewriting the entire passage; instead, offer students some direction on how they might rewrite it themselves.
 
Obviously, time will probably not permit your making such extensive comments about every problem that a student faces. Select the most pervasive problem, the one that you feel the student needs to work on the most. Learning to write is long-term project best accomplished by tackling one problem at a time.
Consider the Four Factors of a Good Essay. First, the essay needs to make clear its purpose. Addressing its audience, it should frame the argument that is being made. Second, the essay should marshal evidence and present that evidence with arguments in a reasoned and logical manner. Third, the arguments should be organized in a persuasive way that builds toward the conclusions. Finally, it must be readable. Let your students know that the use of good grammar, correct spelling and a precise vocabulary make it easier for the reader to understand the argument that they are making.
How these four factors (purpose, reasoning and content, organization and presentation) determine the grade is up to you to decide. Not every instructor places the same weight on spelling as on reasoning. However, in most cases the grade you assign will be determined by looking at these four categories (see the "Grading Standards" handout). Communicate this clearly to your students in your comments. If they have organized the paper well but have failed to draw logical conclusions from the evidence, compliment their organization and then show them how they might have used their evidence more effectively.
Provide Positive Feedback. Your goal in grading papers is to encourage students to perform better on the next assignment. Offering only negative comments merely discourages students. Be sure to comment when the student has made an impressive argument or has demonstrated a clear understanding of the problem being addressed.
Allow Yourself Plenty of Time. Grading papers takes time. After your preliminary reading, allow at least twenty minutes for each three to four page paper.
Use a Grading Sheet or Handout. Often when grading a stack of papers, you will find yourself writing the same comments over and over. Telling each and every student to indent quotations is painstaking work with little reward; in fact, such comments often distract you from the longer and more significant comments that you may want to write. One solution is to make a list of common problems that you found when grading a set of papers and to prepare a handout to distribute when you return the papers. Another possibility is to create a grading checklist. A grading checklist is a form attached to each paper that you grade with typical comments listed and mistakes checked off. (A grading checklist is not, however, a replacement for in-depth comments. You should use it only to mark consistent grammatical and organizational errors, and you should review it in class so that your students are clear on what the pre-fashioned comments mean.) We have provided a typical grading checklist you might use or revise.

ACADEMIC INTEGRITY

Each discipline also has unique concerns when it comes to students’ cheating. In a literature class, plagiarism may be the most common form of academic integrity violation. In math and science, where answer sheets are easily altered, cheating can occur after the exam or assignment has been graded. Teaching assistants in math, engineering and science classes have developed their own unique safeguards against academic dishonesty. (For a fuller discussion of preventing cheating during an exam, see Dr. Carol Baker’s essay in the previous section.)
You can reduce the occurrence of academic integrity violations and catch offenders by doing the following:
  • Include the University of Pittsburgh’s academic integrity policy (or a distillation of it) in the course, lab or recitation syllabus. Remind students before exams that you take the policy seriously and will enforce it.
  • Consider making copies of exams before returning the answer sheets to your students. Photocopying answer sheets makes it impossible for students to change an answer and then return it to you to be re-graded. Although it may be difficult to make photocopies of all the answer sheets in a large class, even copying a handful and announcing your photocopy policy to your students, will deter academic dishonesty. Alternatively, some instructors copy only the papers of students who have, in the past, requested a grade change. Other instructors make copies of exams where the student erased excessively (a common practice of grade alterers used to disguise later changes they plan to make in their answers).
  • Use a marking pen and record both the mistake and the correct answer. Using a marking pen, especially a red pen, makes it more difficult for students to change your original markings and demand unearned credit. As additional insurance, consider crossing out the incorrect answer or repeating the incorrect answer in your grading comments (for example, you might write, "Not 42, should be 45.") Although these actions make it more difficult for students to adjust their answers, you should keep in mind that in a large class the additional time it takes to grade the papers may be prohibitive.
  • Collect unmarked answer sheets or blue books. Require that students leave a random number of pages blank at the start of the blue book (so that smuggled in completed blue books can not be substituted).
  • If question after question, two students use the exact same procedure to answer a problem (where multiple procedures are possible), it may be a sign of cheating and you may want to discuss the problem with the students. Do not act hastily, however; students have rights and unless the pattern is so consistent that coincidence is impossible (even if the two students studied together), you should have some collaborative evidence (for example, you or another teaching assistant saw the alleged cheaters passing notes during an exam) that proves that the academic integrity violation has taken place.
  • Avoid using the same exam repeatedly. Change the questions and the order of the questions.
  • Consider open-ended quizzes. Although not always possible, in some disciplines it is feasible to replace the close-ended quiz (exams with one objective solution) with an open-ended quiz. Open-ended quizzes pose questions with more than one possible solution while still testing for problem-solving skills. For example, in an open-ended quiz you might ask students to solve a problem where all of the information has not been provided. Students then explain what information or materials they would need to answer the question and then describe how they would solve the problem. Since open-ended exams do not pose standard questions with objective answers, it is substantially more difficult to cheat on an open-ended quiz. [3]
. Cheating is an ethical violation and teaching assistants have an obligation to report instances of cheating to their supervising instructor or, if they are the primary instructor, to confront the student. If you believe a student has cheated, carefully document the violation. (For example, if the cheating takes during an exam while another teaching assistant or the instructor is present, ask your colleague to verify the academic violation.) Once you have documented the violation, speak with the student or students involved. If an instructor and a student can reach a mutually agreed solution, it should be put in writing and submitted to the appropriate dean’s office. If, however, an agreement on what has happened and how it should be remedied can not be reached, then the violation should be reported to the school’s academic integrity hearing officer. For more information on the adjudication of violations of academic integrity at your University  or College

PLAGIARISM

POLICY

The University  or College  well structured  has a Code  of Academic Integrity  in
 Which  is included plagiarism as violation of its code of academic integrity.
That  Code  usually states that "students have an obligation to exhibit honesty, and to respect the ethical standard of his [sic] chosen profession in carrying out his [sic] academic assignments." In the case of plagiarism, the policy goes on to say that a student has "violated this obligation" if the student:
  • Presents as one's own, for academic evaluation, the ideas, representations, or works of another person or persons without customary and proper acknowledgment of sources.
  • Submits the work of another person in a manner which represents the work to be one's own.
  • Knowingly permits one's work to be submitted by another person without the instructor's authorization.
In addition, the policy states that a student has violated his or her obligation to the University when he or she has given or received assistance on an assignment where the instructor has expressly prohibited that assistance.
The policy establishes strict guidelines for how an alleged violation should be adjudicated. It states that: "No student should be subject to an adverse finding that he [sic] committed an offense related to academic integrity, and no sanction should be imposed relating thereto, except in accordance with procedures appropriate for disposition of the particular matter involved." These procedures are for both the student and faculty member’s protection. They are described below.
Advise the student that there is reason to believe that the he or she has committed an offense related to academic integrity. Once the student has been informed of the charge, the student must be afforded at least an informal opportunity to respond. If the teaching assistant and the student can come to a specific and mutual resolution, they should put it in writing and submit it to the dean’s office. The matter is then considered closed.
If, however, a resolution can not be agreed upon, the instructor  must file a written statement of the charges with the designated academic integrity hearing officer (and, if it is the end of the term, issue a G grade for the student). Once the statement is filed, a hearing will be convened to evaluate the case.
It should be stressed that this procedure must be followed to protect both the student and the teacher. University policy holds that failure to follow these procedures is grounds for dismissal. In addition, some universities have concluded that in a public institution of higher learning, an individual faculty member does not have the legal right to punish a student. [6] Although you may give a student a failing grade for not meeting the academic standards of the class
Despite these strict guidelines, plagiarism is fairly common in first-year students’ papers. In many cases, it is not that the student wants to cheat, but rather that the student does not know that what he or she is doing is cheating. Sometimes students just forget that they need to cite material drawn from a published work; other times, given an assignment that is supposed to be based on one or two texts, students paraphrase the original source because they cannot think of a better way of presenting the material. These are serious errors that need to be corrected, but instructors should use some discretion in determining the right action to take with first-time offenders whose errors seem to stem from ignorance rather than intent to defraud. In accordance with University policy, you should meet with the student, outline the problem and work out a solution with the student. If the student agrees, having the student rewrite the paper can help the student to learn the rules that govern academic writing at a university.
Other cases are not so ambiguous. Papers that have been lifted wholesale from a text, have been copied from a fellow student, or have been produced by a paper mill are clear violations of the University’s policy. Handle such cases in person and as quickly as possible as outlined in the policy.

PREVENTING PLAGIARISM

Ignorance and panic are, more often than not, the two major causes of plagiarism. As an instructor, you can play a role in preventing both these causes.
In your syllabus, include a clear statement explaining what plagiarism is and that it is a violation of the University policy. Review this statement when you give your overview of the syllabus and the assignments during the first class and stress the seriousness of this offense. In classes where many of the students are writing university papers for the first time, you may want to offer more extensive guidelines. These guidelines should specifically explain what plagiarism is and should tell students how to properly cite others’ works.
Plagiarism often occurs when a student has waited until the last minute to start the assignment and, facing a deadline, looks for a quick way to finish the paper. You can reduce the likelihood of this happening by offering your students as much assistance with the writing process as possible. Set deadlines for each stage in the paper writing process, schedule special "writing workshop" office hours, and offer to look at outlines and read drafts.
Finally, reconsider your assignments. Assigning the same paper year after year or asking students to write "research" papers using only one or two sources does nothing to discourage plagiarism. New, creative assignments not only make it more difficult for students to plagiarize, but also inspire interest in the work so that students are less tempted to plagiarize.
 

THE LOGIC OF GRADES

 
As you plan your course and develop your syllabus, you should also consider your grading criteria. You will be using various types of assessment during the term to help determine the students’ achievement of your learning objectives. At the end of the term, however, you will be required to use a system of letter grades  of  your  University. There are several important decisions that you need to make regarding grading so that you can inform students of your grading policies on the syllabus and in class.
Grading is a difficult aspect of teaching for many instructors. Making judgments about others is often uncomfortable, especially since, as teachers, we want to be helpful and motivate students to do their best. There are different philosophies about grading, and these result in different methods of