Search

Cut scores are used in a variety of circumstances to aid in decision making through the establishment of a clear cut line between adjacent categories. Community colleges regularly use cut scores on placement tests to decide the appropriate course for each beginning student: the first college-level course or a developmental course, depending on student test performance. This initial placement decision sets the stage for the student's college career and must be made in the most defensible manner possible to withstand any challenges that may arise. This paper provides a broad overview of important considerations in the generic standard setting process. Specific standard setting methods are only touched upon briefly, but many commonalities in actions and processes exist among the various standard setting methods. This paper covers both actions and processes at a level that is accessible to the non-psychometric layman with only a basic understanding of assessment. The paper also provides an overview of the key elements to consider in choosing a placement test, for which the cut scores are implemented and provides case studies. Placement decisions can only be as good as the instrument and cut scores upon which they are based. Areas for potential placement testing research are also provided. (Contains 2 figures and 6 footnotes.)

This study examined operational data from the SAT Reasoning Test™ to determine if students who tested under extended-time conditions were suffering from excessive fatigue relative to students who tested under standard-time conditions. Excessive fatigue was defined by significant (a) increases in differential item functioning (DIF) and (b) decreases in item completion rates, for items at the end of testing compared to the beginning of testing. Both of these factors were examined by comparing the performance of students who tested under standard time to students who tested with extended time on items administered early (section position 2 or 3) and different items administered late (section position 8, 9, or 10) during the 10-section test administration. The sample included students with learning disabilities and/or Attention-Deficit/Hyperactivity Disorder (ADHD) who tested with extended time (time and a half or double time) and students without disabilities who tested under standard-time conditions. Analyses were conducted on the critical reading and writing sections of the SAT® and examined item difficulty as well as item completion rates. Results indicated few changes in levels of DIF (early in the test compared to late in the test). In addition, item completion rates for students who received extra time were comparable to (or in some cases higher than) test-takers without disabilities who tested under standard time on both early and late sections. The following tables are appended: (1) DIF Results by Item for Form 1 Writing (Standard and Time and a Half); (2) DIF Results by Item for Form 1 Writing (Standard and Double Time); (3) DIF Results by Item for Form 2 Critical Reading (Standard and Time and a Half); and (4) DIF Results by Item for Form 3 Critical Reading (Standard and Double Time).

The introduction of the SAT Reasoning Test™ with a writing section in March 2005 and the concomitant elimination of the SAT® Subject Test in Writing after January 2005 have led many colleges and institutions to ask for guidance in using the new SAT Reasoning Test writing section scores for college placement and admissions. Standard-setting methodologies provide one possible set of processes that can be used to identify point(s) on a score scale to divide a group of examinees into categories. Many standard-setting methods have been developed, but only one, the Angoff Method (1971) with Mean Estimation for Essays (Loomis and Bourque 2001), is presented in this document. This document presents many of the decisions needed to conduct a standard-setting session and then provides step-by-step directions to enable colleges or institutions to conduct their own local Angoff-based standard setting for the SAT writing section. Many examples of the data collection documents and supplementary preparation and training materials useful during the standard-setting process, including test items for the March 2005 SAT writing section, are provided. Colleges and institutions are strongly encouraged to use the free Admitted Class Evaluation Service™ (ACES™) offered by the College Board for users of College Board tests to collect information regarding the validity of using specific cut scores on the SAT Reasoning Test writing section for placement or admissions for their institutions' applicants and students. The following are appended: (1) Information on the SAT Reasoning Test™ Writing Section; (2) A Sample Schedule for a Standard Setting Using Angoff with Mean Estimation to Set One Cut Score; (3) A Sample Schedule for a Standard Setting Using Angoff with Mean Estimation to Set Two Cut Scores; (4) SAT Essay Prompt and Multiple-Choice Items for Writing--March 2005; (5) Scoring Guide for the SAT® Essay; (6) Answer Key for the SAT Reasoning Test Writing Section; (7) Training Essay Prompt and Multiple-Choice Items for Writing Section; (8) Angoff with Mean Estimation of Essay Ratings Training Record; (9) Training Evaluation Form; (10) Angoff with Mean Estimation of Essay Ratings Record; (11) SAT Writing Composite Score Conversion Table; (12) Final Evaluation Form for the SAT Reasoning Test Writing Section; and (13) Dictionary of Terms.

The College-Level Examination Program® (CLEP®) consists of a battery of 34 tests covering courses in which entering college students often enroll. CLEP exams are tools that help accredited colleges and universities provide course requirement exemptions to students who have gained outside-the-classroom knowledge of certain subjects. Students may utilize course exemptions, which they earn through CLEP scores that are acceptable to the accredited college or university of their choice, in shortening the path to their graduation. An exemption could also allow students to take other courses in the place of a requirement that the CLEP exam is used to fulfill; this course may be a more advanced course in the same field of study or possibly a course in another area of particular interest to the student. The purpose of this paper is to evaluate the validity of CLEP score use.

Due to the high stakes that my be attached to placement decisions, it is imperative that the placement process be as solid and defensible as possible. An integral part of the placement process is the identification and use of cut scores, the point(s) on the score scale that classify students into adjacent categories for placement decisions. This paper is geared toward helping college administrators make valid decisions regarding setting cut scores, focusing particularly on selecting a method, but also discussing issues such as defining performance levels and validating the process.

Student attitudes toward the Internet were investigated in a study designed to develop an instrument that would provide a quantitative measure of the attitudes undergraduates have toward educational uses of the Internet while studying some behavioral correlates of student attitudes. The responses of 395 undergraduates to some form of the scale were used to construct the measure. Statements soliciting attitudes toward educational use of the Internet were written in two formats: the Thurstone equal appearing interval scale and the Likert summated rating scale. These pilot scales were administered with a social desirability response scale to ensure that students did not respond to scale items in a socially desirable manner. The final form, administered to 188 students, was an 18-item Likert-format "Attitude toward Educational Uses of the Internet" (ATEUI) scale that yielded a high internal consistency. Several behavioral correlates lent some credence to the scale's construct validity. Favorable attitudes were associated with: (1) keeping track of valuable educational Internet sites; (2) sharing information found on the Internet with friends; (3) choosing classes that use the Internet; (4) greater frequency of Internet use, both in general and for educational purposes; (5) greater number of reasons for using the Internet in education; and (6) greater number of Internet features used. Also, there were no differences between men and women or in class standing in ATEUI responses. Future research that considers using the ATEUI should continue to obtain new behavioral correlates of the domain. (Contains 2 tables and 14 references.) (Author/SLD)