Nira Herrmann

A methodological study of a stochastic model of an AIDS epidemic

Abstract A model of an AIDS epidemic in a population of male homosexuals was formulated as a stochastic population process. The paper is a methodological study in the sense that computer-intensive methods were used to investigate some properties of the model statistically rather than relying solely on classical methods of deductive mathematics. Three factors of importance in the evolution of an AIDS epidemic were studied in a numerical factorial experiment. These factors were the distribution of the latent period of HIV, the probability of infection with HIV per sexual contact with an infected individual, and the distribution of the number of contacts per sexual partner per month. The numerical experiment suggested that the distribution of the latent period of HIV will have a decisive impact on the evolution of an AIDS epidemic but this impact will depend crucially on the levels of the other two factors. A Monte Carlo experiment suggested that if forecasts of an epidemic were made solely on the basis of deterministic nonlinear difference equations embedded in the stochastic population process, then predictions of the number of individuals infected with HIV and AIDS cases may be overly pessimistic.

Redesigning introductory computer programming using multi-level online modules for a mixed audience

We report here on an extensive redesign and unification of the Introductory Computer Programming sequences offered to computer science, computer engineering, information science and digital media majors. The redesign is intended to improve student learning while reducing costs. The approach makes use of substantial Web-based course material and course management tools, including multi-level online modules that individualize instruction and enable students to self-schedule learning each week. Each module covers a particular aspect of computer programming at different levels of knowledge. Students are assigned work and reading from the module at a level appropriate to the objectives of the long-term goals of their major. This allows students in different majors to acquire the appropriate skill level for each technique and concept. Peer mentors and teaching assistants provide assistance online or in person. In the future, we plan to expand the self-scheduling aspect of the course to allow students to enter the course at different modules, depending on their previous knowledge.

Assessment of a course redesign: introductory computer programming using online modules

We assess the effectiveness of an extensive redesign of the first Computer Programming course offered to computer science and computer engineering majors. Our goals were to improve student learning while reducing costs by making use of substantial Web-based course material and course management tools, including multi-level online modules that individualize instruction and enable students to self-schedule learning each week. DFW rates and costs were significantly reduced by the redesign.

Using HTML and JavaScript in introductory programming courses

Students with little or no computer programming experience prior to entering college often have difficulty keeping up with the fast pace of college-level programming courses, even at the introductory level. For the past several years we have developed a curriculum for teaching fundamental language concepts to this population of individuals using the programmable features of a variety of software packages --- thus giving students nontrivial results with relatively little syntactic overhead. These pre-programming courses prepare students to succeed in subsequent language sequences, or they can serve to provide computer literacy credits for non-technical majors.Here we report on a course designed to exploit students burgeoning interest in the World Wide Web (WWW), where we used HTML and JavaScript to teach programming concepts. These languages allow students at different skill levels to work side by side, learning common abstract ideas while implementing them at different levels of complexity, motivated by the rewarding and exciting interactive environment of the WWW.

Mail merge as a first programming language

The ‘mail merge’ or ‘form letter’ feature prevalent in most word processors provides many features which make it desirable for a first introduction to the art of computer programming. We demonstrate how key concepts are introduced in an environment easily understood by the student, including variables and identifiers, the importance of order in specifying input, data files, output files, IF/THEN/ELSE statements, nested and compound IF statements, issues of syntax vs. semantics, garbage in/garbage out, debugging, and the creation of output through specifications written in a language understood by the computer.

Creating an authentic learning experience in introductory programming courses

We have developed an integrated, software-based course in scientific and statistical programming consisting of an introduction to computer programming and data analysis concepts. This course is being taught in an innovative way to non-majors: “stretched” over two quarters rather than taught in a single term. Classes meet in a computer classroom so students have a seamless lecture/laboratory experience to reinforce the concept that the computer should be made use of whenever needed, rather than only at specified lab times.nIn addition to presenting key programming and data analysis concepts, we are giving the course an applied research focus to illustrate to students the importance and utility of programming and statistical concepts to their own fields. This focus provides motivation for students to learn material they often perceive to be difficult and not relevant to them. It also provides a mechanism for addressing the increasing perception of faculty in a variety of technological fields that many students have problems with abstraction, the use of symbolic notation to understand or express ideas (e.g., through mathematical models), the interpretation of graphical information, and written communication, since all of these skills are needed in applied research.nThe software-based approach to teaching programming concepts dovetails nicely with the applied research orientation of the course in that the software we use is widely applicable to a variety of activities, from word-processing to data handling and analysis.

An integrated, software-based approach to teaching introductory computer programming

We have developed a course in scientific and statistical programming consisting of an introduction to computer programming and data analysis concepts using a variety of software packages. This approach addresses the problems inherent in introducing programming to non-computer science majors, particularly those in engineering, the sciences, and the social sciences where computing and statistical data analysis techniques are essential professional tools, as well as to computer science majors with minimal or nonexistent programming backgrounds.nKey programming concepts are introduced, including variables and identifiers, absolute versus relative addresses, assignment statements, IF/THEN/ELSE statements, nested and compound IF statements, truth tables, precedence of operations, use of built-in and user-defined functions, dummy variables, passing by value and reference, the importance of order in specifying input to functions, modular program design, subprograms, debugging and testing techniques, properties of good programs, and iterative loops. Elementary statistical concepts and data analyses are covered within a computing environment context that emphasizes data analysis and interpretation of results.nAssignments and examples are developed in collaboration with the students major departments to insure relevance and interest to the students.

Why everyone should know how to program a computer

The notion of programming a computer usually connotes the idea of writing computer programs in general purpose, block structured languages such as Pascal, C, FORTRAN, etc. The need for people to become proficient in such languages, even for scientists and engineers, is perceived to have diminished as powerful software tools have become available. In this paper we argue that the need for students to be familiar with the basic fundamentals of programming a computer are stronger than ever, regardless of whether they intend to become computer programmers. We discuss a software based approach developed for students in all curricula which addresses this need by introducing the concepts of programming a computer in more intuitive and friendly environments than those afforded by traditional programming languages.

A prototype statistical advisory system for biomedical researchers I: overview

Abstract A prototype for a statistical advisory system has been developed to explore the possibilities of providing assistance to biomedical researchers who lack the statistical expertise to apply appropriate methods for group mean comparisons. The components of the system are a statistical inference machine, an object-oriented statistical taxonomy and a dynamic multi-window interface, all programmed in LISP. The advisory system is coupled to PROPHET, a National Institutes of Health molecular biology computer resource familiar to many biomedical researchers, which provides the statistical algorithms and graphics capabilities. The objective of the system is to guide the biomedical researcher who is a not an expert in statistics to appropriate methods, to flag potential pitfalls in applying statistical methods, and to suggest alternative statistical directions, including consulting with a statistician.

A prototype statistical advisory system for biomedical researchers II: development of a statistical strategy

Abstract The goal of a computerized statistical advisory system is to incorporate into software specific and sound advice on appropriately applying statistical techniques and interpreting their results for a given set of data. At the heart of such a system is the statistical strategy, the conceptual model used to represent the reasoning of an expert statistician. this paper presents our experience in developing a statistical strategy for a prototype knowledge-based advisory system. The statistical strategy was conceived in two stages. First, a hierarchical structure of decision trees at different levels of abstraction, together with rules for traversing the structure, was developed as a vehicle for eliciting the strategy from expert statisticians. In the second stage, the underlying objects and essential modes of decision-making implicitly incorporated in the decision-tree structure were extracted and a specific taxonomy for the problem was developed. A more general, object-oriented paradigm based on the results of this second stage was used to implement the prototype model. The statistical strategy was developed with a specific user audience in mind and focused on a well-defined subset of statistical techniques often used by that audience.