Search

Complex conditional instructions ("if X, then do Y") are prevalent in public documents, where they typically appear in prose form. Results of two previous studies have shown that conditional instructions become very difficult to process as the structure becomes more complex. A study was designed to investigate whether this difficulty can be alleviated by presenting conditional instructions in formats other than prose. It was hypothesized that the major sources of difficulty--disjunction, negation, hierarchical structure, and ambiguous punctuation--would be eliminated by the use of algorithms. Subjects were presented with a sample of complex conditional instructions in three formats--prose and two forms of algorithms, lists and flow charts. Results showed that conditional instructions were easier to follow when presented as algorithms than as prose when subjects were prepared to follow algorithms by practice with feedback. In general, the algorithm helped performance when the condition in the instruction was logically complex. The overall superiority of algorithms was qualified, however, by three complications: (1) the difference between flowcharts and lists changed the speed relationships between prose and algorithms, (2) the effect of partial processing changed the speech relationships, and (3) the difficulty with the first exposure to the instruction drastically reversed the algorithm advantage in both speed and accuracy. (HOD)

In an effort to assess the educational needs of limited-English-speaking personnel in the U.S. Army and to assess the effectiveness of existing programs in English as a second language (ESL), information was gathered from analysis of files on limited-English-speaking soldiers, 2,804 of whom were enrolled in the ESL program, and field studies of seven ESL programs using questionnaires, classroom observations, and oral proficiency testing. The data are intended for use in deciding whether existing programs should be curtailed, maintained, or expanded, and to decide about recruiting and training limited-English-speaking soldiers. It was found that the majority of limited-English-speaking soldiers are insular Puerto Rican and have at least high school diplomas. Those enrolled in ESL programs reported favorable feelings about programs and teachers, but also felt that greater emphasis should be placed on speaking and listening rather than on reading and writing the language. Gains in English comprehension test scores averaged two points per week during the six-week programs, and soldiers with final scores above 50 were more likely to succeed in the Advanced Individual Training than lower scoring soldiers. A high correlation was found between the reading and writing based English comprehension test and an oral proficiency test. Across installations, ESL programs differ in size, curricula, teaching techniques, and philosophy, but all produced improvements in English language skills. (Author/MSE)

Describes the role of natural language processing (NLP) techniques, such as parsing and semantic analysis, within current language tutoring systems. Examines trends, design issues and tradeoffs, and potential contributions of NLP techniques with respect to instructional theory and educational practice. Addresses limitations and problems in using NLP tools, and considers approaches to assessment. Contains 112 references. (JKP)

Forty adults served as subjects in a study of the comprehensibility of instructions conditioned by multicategory compounds, such as "If you are male, or both married and not employed, go to question 3." The stimulus materials were 96 sentence frames of the form, "If you are X, press (button) Y," in which "X" was replaced by coordinate compounds generated by different combinations of five factors: (1) number of categories in the antecedent statement, (2) organization of categories, (3) "and" versus "or," (4) presence or absence of negation, and (5) organizational subtypes. In individual sessions, subjects responded to all 96 instructions, presented one at a time on a computer display terminal. Analyses of subjects' response time and accuracy yielded a number of results, including that the pattern of results, for both speed and accuracy, was primarily of the interactive type and that each of the major sentence dimensions appeared to contribute to overall sentence difficulty. The results have both theoretical and practical implications and lend support to the premise that complex instructions should be experimentally evaluated prior to inclusion in a public document. (GT)

Viewed as discourse, official forms exhibit identifiable text characteristics and elicit strategies for comprehension that take into account these characteristics, as well as context and the user's prior knowledge about the world. However, forms also have unique characteristics that require processes and strategies which differ in systematic ways from those needed for conventional texts. Protocol analysis has revealed that expert form users operate on several levels as they try to understand and complete a form. The lowest level is represented by decoding statements, indicating that the user was devoting attention to the lexical and syntactic aspects of directions in order to figure out word meanings and to clarify sentences. The second level, form-using strategies, represents text-level processes for understanding in which the user goes beyond words and sentences in attempting to relate items across the form or to draw on personal knowledge to clarify the meanings of items. The third level, metacomments, reflects the global strategies that arise as the reader puts the document in a societal and institutional context. The preliminary results of protocol studies indicate that the comments of the expert form user are far more likely to reflect the higher two levels of strategies than are the comments of the novice form user. Results of these studies can be used to design forms to facilitate the use of higher level strategies and to exhibit more explicitly the discourse features--cohesive devices, scenarios, instantiation--to which successful form users attend. (HOD)

Features are discussed that are critical to the comprehension of texts and that readability formulas cannot handle. The critique and alternative analyses are confined to public and institutional documents and are based on research in cognitive psychology and psycholinguistics. Two types of comprehensibility complications are examined, those dealing with sentence length and syntax, and those dealing with context. It is demonstrated that readability formulas cannot grasp such variables as the effect of syntactical combinations, semantic and pragmatic relationships underlying a paragraph, and awareness of context. Examples of revisions in public documents that take account of factors beyond readability are given at the level of words, sentences, and the whole text. It is recommended that: (1) technical or special terms be tested for comprehension by the target audience; (2) content should be expressed by scenarios where possible; and (3) introductions to documents should be worded in such a way that the user will make the needed connections between items. Generally, use of design guidelines drawn from research into the comprehensibility of language and the usability of documents is recommended to writers, revisers, and evaluators of documents. (AMH)