Search

An attempt was made to find evidence for scanning residue in a modified Sternberg (1966) procedure. Lists of two, three, or four words drawn from a large pool were presented to four subjects in a completely within-subjects design. Trials were in pairs such that the list was the same for both trials within a pair. The test item on Trial 1 was either a word, which cued the subject to perform a recognition memory task, or a digit, which cued the subject to perform an irrelevant task. The test item on Trial 2 was always a word. No scanning residue was found: The slope of the reaction time/set size function on Trial 2 was not affected by the type of test on Trial 1.

A subroutine for use in any item analysis program is provided. This routine computes the efficiency of a test item, the exact probability of obtaining this efficiency, and the chance probability of obtaining a greater efficiency for an item, given its difficulty level, discrimination index, and sample size.

The Musical Aptitude Profile (MAP) and four levels of the Iowa Tests of Music Literacy (ITML) were administered to university non-music majors to investigate the use of the “in doubt” test item response. The “in doubt” response was not used indiscriminately by the university students on either test battery, and it appeared to function in a similar manner to use by public school students. The incidence of use of the “in doubt” response on ITML levels suggests that the students in this investigation felt quite confident of their answers on the achievement test levels and that this confidence remained stable throughout a semester of training. No reason to suspect the process validity of MAP or of ITML when used with university students was ascertained.

Self-scoring is a procedure for encouraging students to self-monitor their performance on classroom achievement tests. In our study, students instructed to predict their score on each test item earned higher scores than did students who were not so instructed.

It follows from the study-phase retrieval model of Tzeng and Cotton (1980) that age differences favoring young over elderly subjects should be revealed in a recency-judgment task. It may also be predicted from their “availability-principle” notion that the two sets of age functions relating accuracy of recency judgments to distance between test items (lag) will be parallel. Young and elderly females made recency judgments at lags of 0, 12, 24, and 36 intervening items. Performance improved as a function of lag, the young were more accurate than the elderly, and the two sets of lag functions were parallel.

Ss were presented given names as test items and asked either to indicate whether each test name was a member of a previously memorized list of names or whether each test name was the name of one of their siblings. The data indicated that the memorized hst of names was scanned in a serial fashion but that (disregarding the case where S had just one sibling) S did not have to scan the list of his own siblings' names to decide about a test name.

I HAVE ALWAYS BEEN IMPRESSED by the superiority of true-false items where, if the item is false, the person is required to tell why it is false, thereby lessening the opportunities for pure guessing and increasing the validity of the test items. But I have objected to the fact that this requires reading each test paper individually, substantially slowing correction time and, in addition, opening to debate the legitimacy of an alternative interpretation by the student. I believe that a "multiple-false" test item to a large extent overcomes these difficulties. First of all, it is answerable on a standard multiple-choice answer sheet with no supplementary information required. Second, the item is rapidly completed by the student, allowing a greater number of items to be included in any testing period. Third, this kind of test item tends to lessen the ambiguity of a question by spotlighting the crucial information and indicating information which the question does not contest. Also the nature of the construction of

THE CONVENTIONAL recognition test item can be interpreted as giving the examinee the task of matching a set of response options (r) to a set of stimulus options (s), where nr ^ns. This paradigm, which was introduced in an earlier paper by the writer, (1), applies to multiple-choice, double choice, true-false, rearrangement, and matching items alike, although the designations s and r are, to be sure, somewhat arbitrary. One important re striction imposed on the form of this conventional test item is that every ri, if it has a corresponding si, should have only one such si, and that for every si there should be one and only one ri. In addition, for every si, all nr response options should be equally distracting to the examinee who does not know the correct r match for si. The difficulty which item writers often have in meeting these re strictions is well known ( see e. g. 4)and there appears to be no easy solution to the problem. How ever, it is hoped that the test items constructed from some of the forms discussed in this paper will, by relaxing one or more of the restrictions on t h e conventional format, not only facilitate the writing of items but will also result in tests which are a bit more interesting to examinees and which will econ omize on the amount of paper needed for reproduc ing the tests.

It has become almost a tradition to consider as the only necessary equipment for one engaged in educational measurement some ingenuity in translating subject-matter content into short-answer forms of questions and familiarity with the statistical methods of computing central tendencies, percentiles, the standard deviation, and the coefficient of correlation. Probably this conception has developed because of the types of problems which historically have received the focus of attention in the construction of tests for elementary and secondary schools.

The learning process is complex and incompletely understood. Despite this, the general factors that appear relevant to the measured success of instruction are small in number. At least at the first level of approximation, they can be simply described, first there are those factors that determine that students will come in physical proximity of instructive events. In the case of written material, these factors include the incidence of instructive events in the text, and compliance by students with reading assignments and/or suggestions. The second class of factors are those that determine whether an instructive event, once it has been encountered, will be perceived and internally represented in instructionally appropriate ways. The third factor primarily determines test performance. This factor is related to (a) the semantic and structural disparity between instructive events and the test, as well as (b) forgetting. Whether the required skills are learned is determined by the opportunities for encounters with suitable instructive events, as well as by the likelihood of sufficient internal representation. The disparity and forgetting factor determines whether acquired competence is translated into appropriate test performance.

The development of intelligence during infancy is impressive and measurable. The anthropometrist measures growth in head size, which reflects the growth of the brain. The neuropathologist measures cerebral DNA to estimate cell number and possible damage from malnutrition or other causes (Winick, 1970). The psychologist measures behavioral change by means of careful observations of responses to specific tasks. In this chapter we shall review critically infant tests and their contribution to the understanding of mental growth in the first months of life.

At the turn of the century mental testing was in a state of crisis, at least in so far as the psychology of intelligence was concerned. This aura of crisis is well illustrated by Spearman’s (1904b) review of the early attempts to relate mental test scores to various criteria of academic competence.

This study examined one possible form of test development bias. We wanted to determine if supervisors exhibited 'item leniency,' and judged the items to be easier for enlisted personnel than the items were found to be. Thus we asked supervisors of a combat MOS (11H, TOE Gunner), an engineering MOS (12C, Birdge Construction Crewman), and an administrative MOS (71L, Clerk) to estimate what percentage of their troops would pass each item of the Skill Component (SC) portion f the appropriate SQT.

The preceding chapters described features of test items from a theoretical perspective. These discussions included a variety of topics, from a comprehensive description of a test item to various analyses involving requisite characteristics and conditions for good test items. The information in these chapters provides a foundation of knowledge necessary if one is to construct items that will contribute to sound measurement.

This concluding chapter identifies and explains some elements of item construction that do not conveniently fit into any of the previously discussed categories yet are important, even overarching, elements of constructing good test items. These features include ethical concerns, considerations related to using copyrighted materials and obtaining copyright protection for one’s own items, and concerns for preparing or modifying items for use with people who have handicapping conditions. The chapter concludes with a final remark on the task of constructing meritorious test items.

A recognition memory procedure was used to investigate the retrieval processes which Korsakoff and Alcoholic Control patients employ in locating a test item in a recently memorized short list of items. The reaction time data for both patient groups were consistent with a serial and exhaustive model of memory search. However, the Korsakoff patients' search rate differed from that of the Alcoholics, with the Korsakoff patients searching at a 40% slower rate than that of the Alcoholics. These results suggest that Korsakoff patients' short-term retrieval difficulties may stem from a decrease in search rate rather than an inefficient “strategy” of search.

When subjects were required to calculate answers for computable problems and answer questions, an interaction was found corresponding to that obtained by Kieras and Greeno (1975) from judgments of computability. With nonsense formulas, much longer times were required to identify noncomputable problems than to compute answers, with a much smaller difference when formulas consisted of meaningful concepts. The better performance on noncomputable problems and questions with meaningful formulas corroborates an interpretation that those items test the connection of algorithms with general conceptual knowledge. Finally, it was found that for relatively complex problems, solution times and time to judge computability were longer if nonsense formulas were learned in separate sets than if they were learned in a single set; however, no such effect was found with meaningful formulas. It was concluded that learning conditions influenced the integration of cognitive structure in the case of nonsense formulas, while subjects were able to adjust organization of the meaningful formulas.

Being familiar with the variety of formats in which test items can appear is one indication of a competent item writer. Familiarity with item formats means that writers understand the definition, structure, and advantages and criticisms for each of these item formats. It also means writers know the technical rules and the guidelines of editorial style for presentation of items in these formats. Whereas the preceding chapter presented information useful to understanding these features for items in the multiple-choice format, this chapter presents parallel discussions for items in several other popularly used formats, including true-false, matching, short-answer, sentences-completion, and cloze-procedure.

The identifiability of isolated vowels (/V/) was compared to that of vowels in consonantal context (/pVp/) when subjects performed a monitoring task. On successive blocks of trials in a test series, the subjects listened for instances of one or another of nine monophthongal vowels (/i,I,q,1,v,a,c,U,u/) and identified each test item as being an instance or not. On average, resulting false alarm errors occurred significantly less often in the /pVp/ condition, consistent with the previous finding that vowel perception may be aided by consonantal context. This beneficial effect of context was found to be restricted to the class of open vowels, however, with perception of the close vowels being somewhat hindered by context. The error data for misses also showed an interaction between context and vowel height. Various accounts of the interaction are considered.

Two experiments involving memory retrieval of auditorilv and visually presented materials were performed. In Experiment I, subjects were presented with memory sets of 1, 2, or 4 stimuli and then with a test item to be classified as belonging or not belonging to the memory set. In Condition 1, each memory stimulus was a single, auditorily presented letter. In Condition 2, each memory stimulus was a visually presented letter. In Conditions 3 and 4, each memory stimulus was a pair of letters, one presented visually and the other auditorily. Mean reaction time (RT) for the classification task increased as a function of number of memory stimuli at equal rates for all four conditions. This was interpreted as evidence for a parallel scanning process in Conditions 3 and 4 where the auditory item and visual item of each memory stimulus pair can be scanned simultaneously. Experiment II compared memory retrieval for a simultaneous condition in which auditory and visual memory items were presented as pairs with a sequential condition in which mixed auditory-visual memory sets were presented one item at a time. RTs were shorter for the simultaneous condition. This was interpreted as evidence that parallel scanning may depend upon memory input parameters.

The Rasch Model is offered as a valuable means to analyze what occurs when students take tests. Unlike Classical Test Theory, the Rasch Model is not sample dependent; it generates data about the test items and the students who took the test. Test item values, item fit statistics, person measures and person fit statistics are some of the results obtained when one uses this model. Through the application of the Rasch Model described here, one may assess its value particularly to nurse educators who address many variables when administering tests. This model facilitates analysis of these variables. Additionally, it allows for comparison of test item values and person measures on the same linear scale. The result of the application of the Rasch Model is a more refined measurement of test item difficulty and student performance.

We have so far acquired some insight into the nature of cognitive tests which might, on theoretical grounds, be considered likely candidates for the measurement of intelligence; we must now turn to a consideration of that part of the theory which asserts that intelligence is the general or all-round cognitive ability which mediates success in such tests whatever their nature. Spearman has called this “the indifference of the indicator”; in other words, if a test or test item fulfils the conditions for a “good” test or test item laid down in his laws of neogenesis, then it should not matter much which item or test was chosen for the measurement. This implication of the theory can of course be investigated empirically, and methods for doing this were worked out by members of the London school - Karl Pearson, the great statistician, Charles Spearman himself, and Sir Cyril Burt, who succeeded Spearman in the professorial chair at University College, London. These methods are essentially based on the use of correlation coefficients, and on factor analysis, i. e. the analysis of sets of such coefficients. There are some statistical complexities to analyses of this kind, but these are inevitably outside the scope of this chapter; there are many good books dealing with the technique of factor analysis (Thomson, 1939; Burt, 1940, and Thurstone, 1947, are three classics; among modern texts are Harman, 1967, Pawlik, 1971 and Lawley and Maxwell, 1971).

Item response models offer a number of advantages for test score interpretations and reporting of test results, but the advantages will be obtained in practice only when there is a close match between the model selected for use and the test data.

This chapter describes the characteristics of frontal lobe disorders. Damage to the frontal lobes produces disorders of personality, motor behavior, planning, and motivation. A variety of other behavioral manifestations may complicate frontal lobe damage, including the perseveration of motor movements, motor impersistence, confabulation, and reduplicative paramnesia. Abulia tends to be most marked after bilateral frontal lobe damage. The perseveration of visual, auditory, or tactile sensations may occur. Two distinct types of perseverative errors may be noted during language testing. A perseveration occurs when response from one test item is inappropriately repeated in the following test item. By contrast, an intrusion occurs when a test response is repeated but after intervening test stimuli. Intrusion errors appear to be characteristic of memory disorders, especially those due to Alzheimer's disease. Confabulation in the neurologically impaired patient is the unintentional, generally transient production of inappropriate and fabricated information, primarily in response to direct questioning. Environmental reduplication can occur occasionally and transiently in the normal individual.

The beginning of Herbert Simon’s Karl Taylor Compton Lecture on “The Psychology of Thinking” (see Simon, 1981) provides a highly generative metaphor. Simon described an ant making its way across a wind and wave molded beach. A drawing of the ant’s path is jagged, irregular, hard to describe (see Figure 1); each turn in the path and the length of each of the path’s segments is seemingly unpredictable. From a more distant view, however, the path has an overall direction. It appears to have a structure, to be governed, to seek a goal. Perhaps the ant is returning to its nest.

Eight Broca's aphasics at the Boston Veterans Administration Hospital were administered a story completion test, designed to elicit 14 different English syntactic constructions. The first scorable response to each test item was analyzed with respect to omission of obligatory elements. The patterns of error recorded a consistent order of difficulty of the constructions, as well as a variety of strategies employed by the subjects in attempting to compensate for their limitations in syntactic encoding. The results confirmed earlier observations that initial unstressed functors are particularly vulnerable. Correspondingly a typical compensatory strategy employed by Broca's aphasics is to initiate their utterances with stressed words, like vocatives, or nouns where normal subjects produce pronouns.

A product improvement test of T142 track blocks made of Aluminum Alloy was conducted by Materiel Testing Directorate at Aberdeen Proving Ground, Maryland. The purpose of the test was to determine the performance and endurance characteristics of the Aluminum T142 track in order to compare the results with previous tests of steel T142. Performance testing was conducted with both the aluminum and steel T142 track to obtain comparable data. Endurance testing consisted of 3545 miles (5604Km) conducted over paved, gravel, level and hilly cross-country courses. Acceleration and drawbar pull performance characteristics of the M60A1 tank were improved by the use of the aluminum alloy track. Fuel consumption for the M60A1 was not significantly changed by the use of the aluminum track. The test item successfully completed the 3545 miles (5404Km) of endurance testing. The endurance test results were compromised by conditions caused by the lateral movement of the blocks on the track pins. This aluminum alloy should be considered for use in making track blocks for use on the M60 series tanks.

If we as educators expect a youngster to have learned something by a certain point in his or her life, we have some obligation to make this possible. We often express this obligation by saying that we must provide all youngsters with an opportunity to learn.1 But just what is an opportunity to learn? Does it mean a teacher and a textbook? Does it extend so far as to include a supportive social climate, strong incentives, and individual tutoring?

Scientific revolutions are no everyday affair. So it is of some importance that Fortune (Alexander, 1972) and several other major magazines have recently proclaimed that we are in the midst of a major Kuhnian revolution in the social sciences, one that has significant implications for social policy. According to these magazines, ‘environmentalist’ theories, which assume that equalization of human environments and opportunities will increase equality of achievement between individuals, groups, and races, are in ‘crisis’. The crisis exists because the egalitarian reform programs of the 1960’s, which relied on such theories and thus constituted tests of them, failed to equalize achievement. In the face of this crisis, Fortune suggests, scientists are welcoming evidence from the study of ducks, baboons, and humans which points to “a basic intractability in human nature, a resistance to being guided and molded for improving society” (Alexander, 1972,p. 132). But if human nature is intractable, the argument continues, then social policy should be adjusted to recognize the inherited capacities and differences between individuals, groups and races, rather than continue to insist on unrealistic egalitarian reform programs.

When authors refer to style, they usually mean the expression of ideas in a smooth, orderly, pleasing manner. Each author develops an individual style of expression that allows for a personal presentation of his or her own thoughts and emotions. For editors, however, style connotes something different. Editorial style refers to the consistent use of a set of rules and guidelines. These rules and guidelines prescribe a consistent use of punctuation, abbreviations, and citations, a uniform and attractive format for tables, graphs, and charts, and a correct form for the many other elements that constitute written communication. Test-item writers are both authors and editors. As authors, they can express their ideas in novel, creative, and personal ways; as editors, they must be aware of and follow consistent editorial principles.

The evaluation of student progress in aviation is one of the most important duties of the aviation instructor. Many persons who have reponsibility in this area have minimal training in accepted test and measurement techniques. The author has been engaged in the last several years in developing test questions for use in private, instrument, commercial and instructor ground training courses at a major four year university. Included in this document are a review of testing theory and its application to aviation education, test item construction, and statistical results obtained from the author's investigation.

In Chapter Two we investigated the acoustic properties of the metrical foot by comparing iambic and trochaic lines in nearly identical lexical make-up. The hypothesis that the foot boundaries would be reflected in the temporal make-up of the lines was not confirmed.

In the previous chapter, a set of steps and techniques were introduced for conducting goodness-of-fit investigations. The purpose of this chapter is to highlight the applications of several of those techniques to the analysis of real data. Specifically, the results of fitting the one- and three-parameter logistic models to four mathematics tests in the 1977–78 National Assessment of Educational Progress will be described. This chapter is intended to serve as a case-study for how a researcher might approach the problem of assessing model-data fit.

As part of an effort to provide more economical methods for job-skill evaluation in the Enlisted Personnel Management System (EPMS), a study investigated the use of television as a means of providing test item presentation. The television scene provided the job setting for each item and then posed a question that required a real-time response from the examinee. All items were job-connected and therefore represented simulated skill items. The examinee responded directly to the face of the television screen by means of an electronic stylus or gun reticle. The test was compared with a paper-and pencil test which covered the same items and a hands-on performance test which covered many of the same items. The results indicated that television testing is very acceptable to soldiers and feasible. The validity of the prototype test could not be determined precisely because the criterion hands-on test was not usable.

Both competency testing and test validity have been receiving considerable attention in recent months. But the relationship between the two remains unclear. Indeed, as more attention has focused on the issue of competency tests’ validity, the more disagreement there appears to be among people billed as experts on testing and who might have been presumed to bring some coherent professional opinion to bear on the matter. But where the experts — given to usually abstract disputation — seem to have fostered complexity and provided more questions than answers, lawyers and judges — given more to dealing in the particulars of law and cases at hand — are producing some firm answers on the question of competency tests’ validity. It remains uncertain whether the clash between the disputation of testing experts and the decision-making of the courts is consistently making for better law, but it is at least leading to more direct attention to practical issues of test validity. This chapter’s purposes include the following: 1. To describe major points of dispute over competency test validity as evidenced in arguments in the Debra P. case. 2. To summarize how courts have made sense of such disputation in reaching their decisions. 3. To suggest some of the ways in which the relationship between a test and what is covered in school might be conceived. 4. To suggest some practical strategies for examining the validity of competency tests.

This paper deals with the use of computers in the evaluation of learning with a special concern for diagnostic evaluation carried out with a formative aim.

In Sternberg’s memory scan paradigm, subjects were required to scan for category membership or physical identity. In Experiment I, the scan rate when checking for either type of relationship at the same time was less than the sum of the scan rates when checking for either relationship separately. In Experiment II, presenting memory sets consisting of items from a single category affected the category scan function but not the physical scan function. The results were interpreted as evidence for the capacity to conduct parallel scans of independent memory encodings when retrieving remembered information.

Summary The development and utilization of multiple-choice test items and the utilization of computer technology have significantly altered evaluation practices in medical schools. This paper describes the development and implementation of a computer-based multiple-choice test item-bank for use in a junior medicine clerkship. The computer program, ITEMBANK, is an interactive system that stores and edits test items and constructs and prints an original copy of each test. Tests can be produced almost immediately by the clerical staff and the total developmental costs have been minimal. The program has reduced the amount of teaching time required to develop tests, and has enhanced the learning of essential information by medical students and residents. Teaching staff and student response has been enthusiastic and positive. Statistically significant Pearson correlation coefficients have been obtained between student performance on these tests and clinical evaluation and NBME Part II scores. The program has made a valuable contribution to the improvement of evaluation procedures in the Internal Medicine clerkship.

It has been well said that psychology has a long past, but a short history. People have puzzled over psychological problems for thousands of years, ever since the dawn of recorded history, yet the development of a science of psychology is scarcely a hundred years old. Plato and Arisotle already discussed the notion of “intelligence”, but it is only in this century that attempts have been made successfully to measure this important variable. Many misunderstandings have attended this venture, and it will be one of the functions of this book to clarify the points on which these misunderstandings have arisen. However, before doing so it may be useful to trace quite quickly, and without too much detail, the early development of the concept with which this book is essentially concerned. Like most scientific concept, this one arose out of everyday observation. The concept of temperature arose from the different feeling caused in human beings by fire and sunlight, on the one hand, and ice and snow, on the other; in this way were the notions of “cold” and “hot” born, and became the subject matter of science. Similarly, the concept of intelligence arose from observations of people trying to solve problems, to learn difficult and demanding things like mathematics, languages, and history; some seemed to find no difficulty in all this, and succeeded brilliantly, while others were very slow, and often failed altogether.

Recent experiments have provided evidence for an auditory locus of selective adaptation effects. The present experiment further tests this theory. A [pa]-[ka] series was constructed. The burst frication from the [ka] syllable was added to the vowels [u] and [i]. Subjects identified these syllables as [tu] and [pi]. These three syllables contained physically identical bursts but were identified by subjects as stops with three different places of articulation. The [pa], [ka], [tu], and [pi] syllables were used as adaptors on the [pa]-[ka] test series. The [ka], [tu], and [pi] syllables, which contained identical bursts, produced similar boundary shifts. The spectrally different [pa], although sharing its initial phoneme with [pi], produced an opposite shift. These results support an auditory locus for adaptation with little or no phonetic or linguistic influence. In a pairedcomparison procedure, [pa], [ka], [pi], and [tu] were used as exemplars. Both the [pa] and [pi] syllables produced fewer [p] responses to an ambiguous test item, whereas [ka] had the opposite effect of producing more [p] responses. The phonetic quality of the exemplar appears to have been the primary determinant of its effects in the paired-comparison procedure. Together, these results support a two-stage model of speech perception, in which neither of these stages are vowel contingent.

Educators readily agree that computers may constitute an important aid to teaching. There is less unanimity, however, as to how computers can best be employed. Neglecting such matters, three main functions can be distinguished in the practical application of computers: (a) as administrator, (b) as tester, and (c) as teacher.

To evaluate H2 production in the rat after ingestion of legume products as a predictive bioassay of their flatulence activity for man, responses of man and the rat to (1) a bland formula diet, (2) cooked California Small White bean (CSW) solids, (3) 70% ethanol extract of (2), (4) oligosaccharide fraction of (3), (5) non-protein nitrogen fraction of (3), and (6) combination of (4) and (5) were compared. Six human subjects were fed an 800 kcal breakfast of (1) or containing the amount of a test item equivalent to 100 g CSW. Each subject received all items once plus a single replicate of three items at the rate of one per week in random order. Total flatus volume, H2, CO2, and CH4 volumes and breath H2 concentration were measured for the interval 2.5 to 8.5 hours after breakfast. Young rats in metabolic chambers received known amounts of items 2 to 6 mixed with a bland basal diet. H2 trapped in the chambers was measured after 20 hours. Correlation of rat H2 and to human breath H2 measurements and flatus gas volumes were significant and positive except for negative correlation with flatus methane volume. Measurement of hydrogen production in the rat has potential for a predictive bioassay for flatulence in man.

An elementary rule for any writer is to know thoroughly the subject before beginning to write. This is the reason reporters ask questions, researchers investigate hypotheses, and novelists ponder their protagonists, all before they put pencil to paper. So too must writers of effective test items master their subject. Learning about test items means comprehending what test items are, understanding their purpose, and becoming familiar with their characteristics. Knowing the definition of a test item is the first step toward comprehending it. The definition forms the backbone upon which the purposes and characteristics of test items rest.

The development of mental abilities during infancy is impressive and measurable. The pediatrician measures growth in head size, which reflects the growth of the brain. The neuropathologist measures cerebral DNA to estimate cell number and possible damage from malnutrition or other causes (Winick, 1970). The psychologist measures behavioral change by means of careful observations of responses to specific tasks. In this chapter we shall review critically infant tests and their contribution to the understanding of mental growth in the first months of life.

This report describes the development and testing of the 155 mm Ballistic Simulator for laboratory testing of artillery fuzes. Present test methods using a 5 inch air gun are compared with the ballistic shock environment and are shown to have significant limitations. The design approach of the new simulator is described. The new system includes as its basic component a 155 mm field gun and uses compressed air as the high pressure fluid. Provision for shaping the acceleration-time history is included as well as simultaneous rotation of the test item. Initial tests on the 155 mm Ballistic Simulator have been performed up to 12,000 g's and sample acceleration-time histories are shown. It is concluded that within its operating range, the new simulator provides a significantly better simulation of ballistic shock than the present test method.