Words: 3943
It could be the only way in which students who live far from a campus to take the classes or programs they need. Few studies have tried can characterize the online student. One found the largest group of online students to be women, over 25 years of age, who worked more than 30 hours a week. This characterization is essentially the same as the typical distance education student (Hoffman, AAA). A report by a faculty group at the University of Illinois, however, has found many on campus students take many if not all of their classes online (Regulator, 1999).
As distance and Web-based learning becomes more popular and more accessible, high school, college and graduate courses are being offered via he Web as part of complete diploma and degree programs by more and more institutions. Better, (1997) used an Invoices search with the term “online courses,” it returned 3. 5 million hits. Corporations have found online learning to be a more economical alternative than the typical corporate training session (Hamburger, 1998).
However, little research has 130 been done to understand some of the ramifications of this fast growing phenomena (Grossman, 1999). Different professors’ classes could have different class population characteristics, and could give those professors different impressions and pinions of issues confronting online instruction. In the future, as online education becomes more pervasive, the characteristics of the online student could change, and so too, the problems of online education.
Student Evaluation Practices Student evaluation strategies used by instructors not only serve to motivate, but to help students select strategies to organize their learning (Davis, 1999; Weakened, 1999). Decisions by instructors of which evaluation methods to use, serve as a hidden code to students directing them to the skills and behaviors that are important for them to succeed (Crooks, 1988; Science Education, 1997). Many students tend to invest their time as economically as possible, by studying only those aspects of a course that they expect will affect their grade (Science Education, 1997). The learning strategies students adopt are powerful predictors of educational outcomes, so that expertise in the selection and application of learning strategies is an important educational outcome” (Crooks, 1988, p 441) These evaluation cues are sometimes not obvious to the instructor (Tang, 1998). Crooks (1988) cited a study of the curriculum at the Massachusetts Institute of Technology by Snyder in 1971. Here the formal curriculum scribed an approach 131 characterized by problem solving and creativity, however, the evaluation procedures, which he called the “hidden curriculum” were oriented towards surface learning.
Fuhrman (1983) listed the importance of various data sources in assigning grades. This was gathered from a study of 700 faculty members at a major state university. Table 1 Importance of various data sources in assigning grades Source: (Fuhrman & Graphs, 1983, p 168) Two studies explore how assessment is accomplished in a distance learning environment. One study by the National Center for Education Statistics (Lewis t 1997) found that 98% of all institutions use testing for their credit- bearing distance learning classes.
One third of the classes used proctored group exams at remote sites, while another third gave proctored exams on campus. About 15% of the instructors sent the students their exams by mail or fax (so the students could take the test independently). About 8% take interactive tests at remote sites using either a computer, video, or telephone. Objective Exams 37 points Essay Exams 17 points Term Papers 11 points Recitation/Problem Sets 7 points Term Projects 6 points Class Discussion 5 points
Laboratory Performance 5 points Other 4 points Effort 3 points Attendance 2 points Laboratory Reports 2 points Ratings by T A. S 1 point 132 A second study by Dirks (1 998), where 20% of the distance classes were delivered by the Internet, found that the larger the class size, the more likely that exams would be used to determine more than half the grade. Many evaluation situations are focused towards a specific type of learning outcome by the nature of their logistics. Tests nag Tests are used to measure and encourage a student’s progress towards the accomplishment of course objectives.
Well-constructed tests can be used to valuate a variety of cognitive skills including knowledge, comprehension, application, analysis, synthesis and evaluation (Davis, 1999; Fuhrman & Graphs, 1983; Science Education, 1997). Many of these tests contain objective items, which are difficult to design, as a tool to assess higher order thinking skills (Travis, 1996). Garfield (1994) notes that test items usually evaluate skills in isolation and very rarely integrate them in a real world context (Garfield, 1994).
The Western Governors University and The University of Phoenix Online use a series of comprehensive examinations to know what classes their students re required to take. Students pass classes by passing only the exams, in some cases. Some of these exams include essay or multiple-choice items, while others are projects. These highlights, standardized tests are usually administered in proctored, online environments (Carnivals, 2001). 133 The College Independent Study program, a distance education program at the University of Nebraska-Lincoln, utilizes proctored, individual exams for all its classes (Sapwood, 2000).
Various item types have their own advantages and disadvantages. True-False questions are easy to construct and grade. However, guessing would be 50% erect, and it would be difficult to test gray areas (Davis, 1 999; Science Education, 1997). Good matching problems should have different numbers of items in both columns and there is the possibility of having more than one correct answer. These items assess associational knowledge by testing recognition rather than recall (Davis, 1 999; Science Education, 1997). Multiple-choice questions are probably the most commonly utilized objective test item.
Though the distracted are difficult to write, the items are easy to grade. These items can be constructed to measure simple recall and complex radical thinking skills. However, multiple-choice questions, which effectively test higher order thinking skills, are very difficult to write. These questions can be answered quickly, so instructors can evaluate many different objectives in a single session (Bolton-l_sews, 1 998; Davis, 1999; Gay, 1980; Science Education, 1997; Weakened, 1999). While short-answer questions are easy to write, they take more time to score.
They eliminate guessing, and stress recall of information rather than recognition. Critics feel that this type of question can place too much emphasis on rote learning. However, they can give limited insight into how students can express their thoughts (Davis, 1999; Gay, 1980; Science Education, 1997; weakened, 1999). 134 Essay tests can provide a way to recognize how students organize, integrate, interpret material, and express themselves. This gives instructors the opportunity to comment on these qualities as feedback. Studies have shown that students prepare for essay exams by focusing on broad concepts rather than details.
Students need more time to answer essay items, so fewer objectives can be covered in a testing session. These items are very difficult ND time consuming to grade. Instructors need to be conscientious in order to be consistent and fair in their scoring (Bolton- Lewis, 1 998; Davis, 1 999; Science Education, 1 997; Weakened, 1999). These tests can be given in different environments. There are several different testing formats: the group written test, individual written tests, and online/computer assisted tests. Each of these formats can be proctored or unprotected.
Combinations of these options give us a several testing strategies. Take home tests. Take home tests are convenient, both for the distance learning instructor and for the student. They have the capability of helping the student to develop deeper learning than the large group test These tests are usually taken over a few days to a week, and students have the opportunity to use books, the Internet, and other resources. Problem sets, short answers and essay type items are considered the most appropriate items for take home tests, when the students can have resources available and time is not a constraint.
Instructors of campus classes find this mode of testing useful because it saves class time. Sometimes this type of test is called a power test (Billie & Toothy, 1997). These tests can simulate authentic situations that professions face every day. Authentic 135 assessments are designed to reflect the behavior and skills required in the real world situations. Authentic learning focuses of students task performance and adds an aspect of relevance to the learning procedure (Travis, 1996). These tests are noted for reducing student anxiety.
Many students don’t study as intently for a take home test as they would for a regular test (Crooks, 1988). Researchers also note that the students who rely the most on notes and their texts during open book exams seem to be the rower achievers (Boniface, 1985; Davis, 1999; Science Education, 1997). Crooks (1988) notes that more research is needed on take home and open book tests. Most of the treatments have been very brief. Many of the studies have not given the subjects adds Tate time to develop the skills necessary in order to utilize resources effectively during a test.
Online Tests. Many instructors and other testing organizations are looking to computerized tests as an evaluation tool. They find the advantages of saving class time for traditional instruction, automatic grading, and immediate feedback of objective exams. These tests can be teacher made with Javascript or a CGI, plus there are many software products available (Gibson et al. , 1999; Greenberg, 1998; Natal, 1998). These tests are more secure because items and choices can be randomized (Solos, 1995).
Sometimes learning activities for the students can be guided by test performance. Test times can be flexible, and practice tests made available. There are notable disadvantages. Although objective exams are easier to create, unless the objective questions are well designed, they may only test factual information. Some worry that, in this situation, only surface learning ill occur, and students will not 136 deal with conceptual understanding. Computer anxiety or lack or computer skills might put some individuals at a disadvantage (Greenberg, 1 998; Lander, 1 997; Natal, 1998).
Collaborative Participation This technique, like online testing, was started as an attempt to simulate the traditional classroom on the Web. However, many instructors have found advantages to the Internet based collaboration. Web-based collaborations can take place synch ruinously or at a scheduled time, just as class participation is done in a traditional face-to-face classroom. Collaboration using the Internet can also be asynchronous. After the discussion is started with a kickoff comment, students and instructor can respond when they have time available. As people add to one another’s comments, the discussion grows.
These connected ideas are said to be “threaded. ” Students can participate, as they have time available, making it very convenient. Many articles report the better quality and quantity of interaction between students and faculty in online discussions during online courses. Comments from the literature attribute this to the fact that everyone’s point is heard, no one is interrupted, students have the time to think through their responses, and each student has more access to the professor than in a traditional classroom (Paint, 1998; Poole, 2000; Vicars, 1999).
Students are more likely to cheat when they feel isolated from the instructor (McCabe & Bowers, 1994). Studies have shown that electronic conferencing systems have contributed to the development Of learn inning communities. The grog. N/the of learning communities is associated with increased trust and respect (Whitewater, Gamer, 137 Robbins, & Shoemaker, 2000; MacDonald, 2001 ; Poole, 2000; Riel & Fulton, 001 With minimized alienation, the amount of cheating in an online course also may be minimized.
On-campus instructors have found that they can extend class discussion time by using this technique to encourage students to exchange ideas between scheduled classes. The grading of collaborative activities is difficult to do fairly and quickly. There is controversy on how or even if this educational activity should be evaluated (Mackinac, 2000). Papers, Portfolios, and Projects Projects and papers have been viewed as a means of encouraging discovery and research skills. After the information has been collected, students must analyze, organize, evaluate, and communicate the data.
These activities are seen as an alternative to testing, which also promotes the development of higher level cognitive and organizational skills (Marshall, 1999). Papers and projects have the student concentrate on a narrow topic rather than a wider scope of test objectives. The skills associated with doing an assignment or a project are similar to skills necessary for reports needed in the real world. These assignments take considerable time to grade fairly and consistently (Bolton-l_sews, 1998). A portfolio is a collection of a student’s work.
Students decide the examples they wish to include which demonstrate their growth and accomplishments through the course. Portfolios can be evaluated as the students completed efforts, portfolios also can be used to assess improvement (Davis, 1 999; Travis, 1996). The student is in charge of the evaluation of this method when they reflect on the direction of their work and 138 growth (Murphy, 1997). The grading and administration of portfolios can be very time consuming. It also is difficult to maintain grade consistency among different students (Travis, 1996; Wiggins, 1990).
Academic Integrity Academic Dishonesty in Testing Academic dishonesty is a long-standing problem. Some studies suggest that it is becoming more widespread (Bestseller, 1999; Collision, 1990; McCabe & Terrine, 1997; factor, Mencken, & Morris, 1990; Schwab, 1991). The Bowers Report in 1 964 pointed to academic integrity as the second most important problem on college campuses, after drunkenness and disorderly behavior. A 1989 Gallup survey of teachers revealed that academic integrity was the third worst problem faced (Schwab, 1991 ).
A survey taken of the New York United Teachers (K-1 2) in 1 995 indicated that it was a frequent or occasional problem n schools. This survey also showed that the vast majority of secondary teachers thought cheating was on the rise (Hildebrand, 1995). Other studies (Haines, Dickhead, Labels, & Clark, 1 986; McCabe & Bowers, 1994) found that the amount of cheating has not changed much over recent time. McCabe (1997) feels that dishonesty doesn’t carry the stigma that it used to. In 1994, McCabe and Bowers compared a survey of students performed by Bowers 30 years ago with a more recent survey by McCabe.
They noted that, although the rate of cheating appeared to remain the same, there were differences. There was a significant increase in the self-reported test and examination cheating among honor code students. The students who do cheat do it more often. Also noted was that the most serious test 1 39 taking behaviors (copying from another student, using crib notes, and helping other students to cheat) have all increased substantially. Students in the Nineties do not understand when collaboration should be used, and when it is considered cheating (McCabe & Bowers, 1994).
The use of “ringers” (experts who stand in to take tests for others) takes place in many large, traditional lecture classes on college campuses. Sometimes a “ringer’ is brought in to take a test and pass the information onto others. A flyer circulated on the University of Arizona campus offered students the services of having a substitute attend classes and take exams for them (Win, 1994). In an interview, Mike Wilding, the assistant dean of students at UCLA, cited situations where students swapped schedules and took other students’ exams (McKay, 1997).
Miramar and Mainline (1 993), listed in a report on college cheating for the US Department of Education, “Took an exam for someone else” (Miramar & Mainline, 1 993,p. 4) as one of the more prevalent heating activities. In a survey of college students and faculty, having a ringer take exams for someone else was found to be among the most serious forms of academic dishonesty. Faculty members only considered the acts of copying during an exam, or paying someone to write a paper to be more serious forms of academic dishonesty than utilizing substitute test takers (Nuns, 1984).
Testing is an area in education where identification of the test taker is essential to the validity of the process. It is obvious that picture IDs and passwords are not the answers. Photo identification can be forged. Sometimes photo IDs are just ignored. Shania p. Lemming, a middle school student, was invited to testify before Congress about her science fair project. The project involved dressing her mother in 140 different disguises and checking to see how successful she would be at cashing checks with photo identification.
She was successful in ten different disguises including that of a clown and a man (Barbaric, 1998). On May 1 9, 2000, a Congressional panel listened to testimony on the availability of false identification on the Internet (AP, 2000). Implications of Academic Dishonesty Gingerers (1992) points out that performance in a task is affected by the frequency of past rewards, and whether those rewards were given for high or low effort tasks. Infrequent reward Of high effort tasks results in a high work ethic. Constant rewards for low effort tasks result in a low work effort.
In another phase of this study, subjects with low work ethic were shown to cheat sooner than subjects with a high work ethic. The person who uses cheating to keep getting high rewards for little effort must do more cheating, and that further decreases work ethic. Series, Hendricks and Circle (1980) found significant correlations between deiced school cheating and “falsifying information about a patient from a laboratory examination, history, or physical examination and finding the patient as normal without obtaining the information” (Series et al. , 1980 p. 125).
In a study by Googol (1995), 83% of the accounting students surveyed believed that a correlation existed between academic and professional dishonesty, but noted that the only difference was that there was more potential for harm in the business world. These student beliefs were verified by Sims (1 993), who found a positive relationship between the level of academic dishonesty admitted to for college work and the level of 141 dishonesty engaged in during business practice. The study noted that people give similar reasons for dishonesty in school and on the job.
Furthermore, the surveyed professionals said they would “no longer engage in these behaviors if they knew that they were considered dishonest by their superiors (instructors and managers)” (Sims, 1 993, p. 209). Validity and Reliability Concerns with Integrity Research The Center for Academic Integrity places the percentage of students who admit to have cheated at about 75% (McCabe, 1998). However, the figures room various pieces of research van’ widely. One of the problems with studying cheating is that it is not just one behavior that a person performs or does not perform.
Nor do you have a group of students who cheat on everything and another that never cheats. Instead, there is a continuum of activities where people are not just placed at the ends but usually in between the two extremes (Newsstand, Frankly-stokes, & Armrests, 1996). Sometimes people are not really sure when asked if an action was cheating or not. Most of the studies of cheating are based on direct, self-reported information f college students, which assumes that these students will be honest about being dishonest.
Although these surveys are invaluable for the glimpse of the whys, wherefores, and situations surrounding cheating it is most likely that some of the information is skewed from not being able to assume completely reliable responses. This has to be kept in mind when looking at the demographic pictures of cheaters. Newell and Loafer (1997) point out that the demographics one is attempting to measure 142 may confound the degree of self-reported cheating and not give an accurate picture of the situation.
Ions and Swift (1998) also point out that what is available is a picture Of the cheaters who will admit to cheating rather than cheaters in general. When asked, 63% of students in the study admitted cheating. When the various cheating behaviors were inventoried individually and the responses compiled, however, the percentage of cheaters jumped to 84%. “Future research efforts concerning cheating behaviors need to recognize that identifying specific cheating behavior may unearth more acknowledgment of cheating than a general question asking whether the respondent had ever cheated” (Newell & Loafer, 1997, p. F 17). Efforts have been made to compare simulated behaviors with student selectors (Allen, Fuller, & Luckiest, 1998), experimental studies (Newell & Loafer, 1997), and random reports (Crinklier, 1994; Crinklier & Sigmund, 1999). Random response (OR) surveys utilize the last digit of a student’s social security number to direct students if they are to tell the truth or say that they cheated. This method has been successfully used to study drug and sex habits. Newell and Loafer (1997) used this technique to confirm the amount of cheating found in the study.
This technique was used to confirm the amount of heating which was found in experimental studies. Even the OR survey method significantly underestimated the amount of cheating. Surveyed people tend to under-report participation in practices which are considered to be unacceptable. People are generally more honest when reporting a friend’s habits. Many people share the same opinions, which their friends hold. Many surveys ask people if their friends participate in activities involving drugs, sex, or 143 cheating, rather than having the person state that they themselves participate in those activities (Sizes, 1999).
Two studies have used course outcome and exam grades in order to show he existence of cheating in specific instances. Riddled and Husband (1998) used a comparison of the grades 1 00 students received from Web-based class work and those received in traditional on campus classes to show that a significant difference did not exist in the rigor or integrity of the Web-based courses. Glottis, Smuggler, Fall and Answered (1994), conducted a study to compare computer based testing to in-class testing.
Among other things, the computer-testing group scored significantly lower grades than their in-class counterparts. The grades of the computer-based examination group were armorial distributed, while the grades from the in-class examination group were skewed upwards. Glottis et al. (1994) attributed this difference to the group that used computer testing having less opportunity to ask for clarification during the test. Solos (1995), attributed this difference to cheating. He noted that, with the computer randomly generating questions, copying would be impossible and meaningless.
The skewed results, he suggested, were the results of the poorer students cheating more than the better students. Glottis, Smuggler and Fall (1995) subsequently agreed with Slosh’s comments. Identification checks were done in the computer lab, while none were done for the in class-testing group (even though one lecture class had 260 class members). 144 Characteristics of Academically Dishonest Students Gender. Most of the studies revealed that males cheat more than females (Awaken, 1991; Allen teal. , 1998; Mccabe & Terrine, 1997; Newsstand et al. 1996). The McCabe and Bowers (1994) study indicated that women were catching up. McCabe and Terrine (1996) felt that this could be because women were entering traditionally male careers. Crinklier (1994) found that he opposite was true, and concluded that women were more likely to cheat than men. Maturity. Many studies concluded that there is more cheating in high school than in college (Davis, Grover, Becker, & McGregor, 1 992; Series, Shiner, & Krause, 1988). The frequency of cheating then decreases again from college to graduate school.
Many reports have noted the relation of maturity with increased honesty (Allen et al. , 1998; Haines et al. , 1986; McCabe & Terrine, 1997; Newsstand et al. , 1996). Others found that other components of maturity besides age, such as marital status and financial independence from heir parents, are related to greater academic integrity (Awaken, 1991: Dickhead et al. , 1996; Generous & McLeod, 1995). Academic Achievement. McCabe and Terrine(1997) had found that students with a lower GAP cheat more than students with a higher GAP.
With an experimental study (not by self-reports) Newell and Loafer (1 997), found that the student& grade in a particular class was negatively correlated with their likelihood of cheating. Houston and Jiff (1976) found that students may be more likely to cheat after a success, and suggest that a failure after a success would be unbearable, while failure after first failure is understandable. 145 peer approval or disapproval of cheating. Bowers (1964) studies have shown that peer approval or disapproval is likely to determine one’s cheating behavior in college.