Hdykf003.sum

HOW DO YOU KNOW, FALL 2000

Tu-Th 4:30-6:00, Logan 402

How do you know is a new course that attempts to teach students how scholars advance knowledge in various disciplines. The focus in the course is on the nature of evidence, and how a case is made for a particular claim. Five different faculty members, from five different disciplines, in two weeks each, using principally original materials, illustrate the process of discovery, confirmation, and change. Students write a short paper relevant to each of these five modules. In addition, the course integrator compares the scholarly process in the different disciplines, teaches some basic statistics, and leads students through an exercise that takes them from scientific findings to media representations of these findings. The course is taught, for Fall 2000, by Professor Paul Rozin (Psychology) as integrator, with modules offered by Professor Dennis DeTurck (mathematics), Professor Rebecca Bushnell (English), Professor Ian Lustick (Political Science), President and Professor Judith Rodin (Psychology), and Professor Ingrid Waldron (Biology). The course meets for lectures with some discussion for 3 sessions out of 4, the fourth session being a graduate student led discussion section.

Discussion sections will be led by three graduate students: Derek Isaacowitz, Vani Kulkarni, and Christel Lutz.

THE FACULTY AND THE MODULES

Paul Rozin, Course integrator

Paul Rozin is the Edmund J. and Louise W. Kahn Professor of Psychology. He is among the developers of the How do you know course, and taught it once. He has also been Director of the Benjamin Franklin Scholars and General Honors Programs for eleven years, at different points in time. His research focuses on the interaction of culture, psychology, and biology, in the domain of food attitudes, and includes studies of the emotion of disgust, how people think about diet risks, how food preferences are formed, and how food functions in different cultures (particularly France, India, and the United States). He has been awarded the Ira Abrams prize for excellence in teaching at the University of Pennsylvania, and is a member of the American Academy of Arts and Sciences.

In addition to integrating and comparing the different approaches in the five modules of the course, Rozin will teach some elementary statistics, basics of scientific method, and an exercise on the relation between scientific findings and their representation in the media.

Summaries of modules

Dennis deTurck Mathematics

The Mathematics component of HDYK has two goals. The first is to develop the students' appreciation of mathematics as a living discipline. All of our students have studied mathematics for many years, and are (sometimes painfully) aware of the ubiquity of mathematics as a means of communicating quantifiable ideas and information, and of its usefulness as a tool in activities arising from the needs of engineering, technology, science (and more recently economics, social science) etc... But much of the mathematics used today was developed not by users for practical applications, but rather for its own sake by research mathematicians.

Examples include the application of chaos theory to studies of turbulence, number theory to cryptography, and abstract algebra to error-correcting codes. Thus, the first goal is to emphasize the notion that mathematics is of interest in its own right, and that the development of new mathematics depends in an essential way on a supply of problems, which can be generated from within the subject itself as well as from other disciplines.

The second goal of the Mathematics component is to emphasize the essential place within mathematics of precision and proof. The idea of mathematical proof is a peculiar one, which arose uniquely in Ancient Greek culture. The nature and standards of proof have evolved both in the direction of proving more advanced theorems and in understanding explicitly the assumptions that lie at the very foundations of arithmetic, algebra and geometry. Thus the second goal is to emphasize the nature of mathematical precision and proof, to apply it in examples, and to examine its evolution through the last few centuries.

The goals are addressed through a classroom dialogue that revolves around

a set of problems (click here) whose solutions introduce notions of precision, language and technique that are central to mathematics but foreign to most students. In particular, they lead to a discussion of various aspects of infinity, including mathematical induction, the definition of infinite sets, countable versus uncountable sets, and ideas of convergence and definitions of real numbers. The specific topics discussed are accessible to most students but are enough off the standard math/science majors' track so that it is rare that any student will be able to use previous content-specific knowledge for an advantage. In the course of the discussions, students discover the necessity for precision in the language of mathematics, experience the nature of mathematical logic, engage in real mathematical discourse, and realize that "mathematical proof" is not as well-defined a thing as they might have thought. A class project has students writing proofs of a "controversial" mathematical statement or of its negation, and then exchanging papers and writing criticisms of proofs that oppose their own.

An example of such a statement is "0.99999...=1". In the course of grappling with the truth or falsehood of this statement, students realize that they need a much more precise definition of number than they have ever had (indeed, the first question that needs to be addressed is whether 0.9999... is a number at all). The class discussion and problem-solving is accompanied by a set of readings from various sources (usually excerpts from the books of Morris Kline, Martin Gardener, Davis and Hersch, Devlin, Lakatos, Hofsteader and others).

Dennis DeTurck is the chairman of the Mathematics Department and the Davidson Kennedy Professor in the College of Arts and Sciences. He is a Lindback-Award winning teacher, who has participated in the How Do You Know course since its inception and is active in other innovative programs in undergraduate education, such as the PennAdvance (web-based distance learning) program, and the integration of computers into the mathematics curriculum. He was founding director of the Middle Atlantic Consortium for Mathematics and its Applications Throughout the Curriculum, and now directs Penn's "Access Science" outreach program to West Philadelphia High Schools. His research specialties are differential geometry and partial differential equations. Differential geometry is the study of curved surfaces and their higher-dimensional generalizations. It is the mathematics behind general relativity as well as the basis for the study of dynamical systems. Partial differential equations concern how quantities that depend on more than one variable change with time. At the moment, his research centers on the problem of understanding the shape of DNA, particularly how something that big (human DNA molecules can be up to

3 meters in length) can fit into a cell. As is often the case with mathematics, it turns out that the mathematics used to understand this problem can also be used to understand other things. In particular, the shapes of laboratory and stellar plasmas (used in experimental attempts to

control nuclear fusion) are determined by the same geometry and the same partial differential equations as the shape of DNA molecules.

Rebecca Bushnell English

Literary evidence/Literary Knowledge: Dr. Faustus

This section of "How do you know" is centered on the play Doctor Faustus, which was written by Christopher Marlowe and first performed in London at the end of the sixteenth century. It is a play about a man who sells his soul to the devil in exchange for knowledge, and as such, is surely relevant to the themes of this course. For many reasons, this play works well to exemplify many vexed questions of literary knowledge: because it exists today in two completely different texts, it makes us ask, "what text should we read? How do we know we are reading the 'right' text?"; because the writer, Christopher Marlowe, was a particularly mysterious and scandalous figure, it makes us ask, "does it make a difference who wrote the play?"; and because it is either very religious or very profane, it makes us ask, "how are we supposed to interpret this text? How do we know that we have the 'right' reading?" In our sessions together we will be talking about the assumptions that underlie the methods of literary interpretation, biographical criticism, and textual criticism, at the same time as we will engage in those methods ourselves.

The texts for this section of the course will include the play itself, a sample of criticism from the 1930s, short sections from two conflicting interpretations of the play (one emphasizing its orthodoxy and the other its subversiveness), and the two versions of the final scene from the "A" and "B" texts of the play.

Rebecca Bushnell is a Professor of English. Her books include Prophesying Tragedy: Sign and Voice in Sophocles Theban Plays, Tragedies of Tyrants: Political Thought & Theater in The English Renaissance, an annotated selected bibliography of scholarship on King Lear and Macbeth, and, most recently, A Culture of Teaching: Early Modern Humanism in Theory and Practice, which considers the links between modern and early modern humanist education. She is now working on a book on the social uses and aesthetics of early modern English gardening books, and so she is currently fascinated by the history of science, Renaissance prose style, and changing ideas of nature (to mention a few things). She has received an ACLS research fellowship and the Lindback Award for Distinguished Teaching. Professor Bushnell is currently serving as the Associate Dean for Arts and Letters in the school of Arts and Sciences.

Ian Lustick Political Science

How do you know whether beliefs about the political world have been transformed into commonsensical understandings that have been so successfully and pervasively naturalized that they are no longer experienced as beliefs? How, in other words, can we determine whether beliefs are "hegemonic," and how can we know when a hegemonic belief ceases to be so?

In my classes I aim to teach students the meaning and importance of hegemonic beliefs in politics (and in science). I will highlight measurement challenges and difficulties faced by analysts trying to operationalize the concept. I will try to show how I have developed one way to approach the definitional, analytic, and research problems involved. Students will each read 3-4 chapters from my book, UNSETTLED STATES, DISPUTED LANDS. These groups will read those sets of chapters, respectively, dealing with hegemonic analyses of the Irish problem in British political history, the Algerian problem in French political history, and the Palestinian problem in Israeli politics.

Ian S. Lustick is Professor of Political Science at the University of Pennsylvania where he holds the Merriam Term Chair in Political Science. Professor Lustick's books include Arabs in the Jewish State: Israel's Control of a National Minority (1980); State-Building Failure in British Ireland and French Algeria (1985); For the Land and the Lord: Jewish Fundamentalism in Israel (1988); and Unsettled States, Disputed Lands: Britain and Ireland, France and Algeria, Israel and the West Bank and Gaza (1993). His articles on ethnic conflict, Middle East politics, American foreign policy, social science methodology, and organization theory have appeared in many journals, including World Politics, International Organization, The American Political Science Review, Foreign Affairs, Foreign Policy, The Middle East Journal, Politics and Society, Israel Studies, The Journal of Palestine Studies, and Cornell International Law Journal. He is a founder and past president of the Association for Israel Studies. He is a former President of the Politics and History Section of the American Political Science Association and currently serves as Associate Director of the Solomon Asch Center for the Study of Ethnopolitical Conflict. He is the recipient of grants from the National Endowment of the Humanities, the United States Institute of Peace, and the Carnegie Corporation. His current research focuses the question of "who is a Jew" in Israel, the future of Jerusalem, and on development and applications of agent-based modeling techniques to the solution of theoretical and policy related problems pertaining to identitarian conflict, globalization, and the requisites of stable democracy.

Ingrid Waldron Biology

Environment, Behavior and Health

This section will discuss the different types of evidence used to evaluate the effects of environmental and behavioral factors on health and disease. We will discuss the strengths and weaknesses of various types of epidemiological studies, clinical trials, and animal and laboratory

experiments. We will discuss the reasons why different types of studies often yield apparently conflicting results, as well as the reasons why different studies of the same type may also yield differing results. This methodological discussion will be illustrated by considering evidence

concerning several specific topics, which will probably be (1) the effects of diet on cancer risk (specifically, effects of beta-carotene), (2) the effects of postmenopausal hormone replacement therapy on ischemic heart disease mortality, and (3) the causes of the substantial socioeconomic

differential in mortality. We will read an introduction to methodological issues, original research articles, and analytic reviews which evaluate current research findings and attempt to identify reasons for apparently conflicting research results.

Ingrid Waldron is Professor of Biology and Undergraduate Chair of the Biology Department. She also holds the Donna and Larry Shelley Term Chair in Women's Studies. Her research focuses on causes of gender differences in health-related behavior and mortality, as well as behavioral and social influences on health and disease. She has been awarded the Lindback prize for excellence in teaching.

Judith Rodin Psychology

How Do You Know: Body Obsession

Genetics, biology, psychology and culture determine body weight and size, but most people are not satisfied with their weight and appearance, and in some, body preoccupation and eating disorders occur. Many believe that these are problems of modern society, but throughout recorded history people have tried to alter their bodies to conform to cultural norms. Even so, body concern has escalated dramatically in recent years. As we discuss the importance of culture to weight and appearance, we will consider what may be different about this particular period in history. We will also explore how body image develops, the relationship between body image and self-esteem, and the biology and psychology of weight regulation and weight control, especially in relation to dieting and exercise.

Judith Rodin is the seventh president of the University of Pennsylvania, beginning her duties on July 1, 1994. Dr. Rodin holds faculty appointments as a professor of psychology in the School of Arts and Sciences and as a professor of medicine and psychiatry in the School of Medicine. She

returned to Penn after 22 years on the faculty of Yale University where she was provost from 1992 through 1994.

Renowned for her work on the relationship between psychological and biological processes in human health and behavior, Dr. Rodin has published more than 200 articles and chapters in academic publications and authored or co-authored ten books, most recently, Body Traps, which examines the role of physical appearance in the psychological health of women. Dr. Rodin has been elected to the American Academy of Arts and Sciences, the American Philosophical Society, and the Institute of Medicine of the National Academy of the Sciences.

Fall 2000

Derek Isaacowitz, Psychology

Vani Kulkarni, Sociology

Paul Grant, Psychology

Most of the course readings will be on the course web site. However, students are asked to purchase two books:

Texts

Doctor Faustus, SIgnet Edition, NAL: ISBN 0451524772 (Paperback)

Ian S. Lustick, UNSETTLED STATES, DISPUTED LANDS: BRITAIN AND IRELAND,

FRANCE AND ALGERIA, ISRAEL AND THE WEST BANK/GAZA (Ithaca: Cornell

University Press, 1993). (paperback)

TEXTS CAN BE PURCHASED AT THE PENNSYLVANIA BOOK CENTER, ON 34^TH ST. AND SANSOM ST.

Details of each of the sections, with assigned readings and writing assignments

Module 1 (Paul Rozin): Introduction, science, and statistics (first 3 classes)

Thursday, September 7

Rozin: Introduction: Science and evidence

Complete in class survey to provide a basis for demonstration of data analysis

Rozin, P. (1995). The nature of science (21 pages). (Chapter 2 from book in progress by Paul Rozin and David Kritchevsky: Thinking sensibly about food, eating, diet and health: The French solution)

Friedman, D., Pisani, R., & Purves, R. (1998) Statistics (Third Edition). New York: W. W. Norton.

Chapter 1. Controlled experiments (3-11)

Chapter 2. Observational studies (12-20)

Tuesday, September 12

Rozin: Introduction to Statistics: 1 Describing the data

Gleitman, H., Fridlund, A., and Reisberg, D. (1999). Psychology (Fifth Edition). New York: W. W. Norton

Appendix A: Methods of Scientific Research A 1-21

Appendix B: Statistics B: 1, 3-9, 11-14

Thursday, September 14

Discussion section

Module 2 (Dennis DeTurck) The nature of mathematics: Precision and proof

see website

www.math.upenn.edu/~deturck/hdyk/main.html

Module 3 (Rebecca Bushnell).

General introduction: Literary Knowledge

This section of "How Do You Know" is centered on the play Doctor Faustus, which was written by Christopher Marlowe and first performed in London at the end of the sixteenth century. It is a play about a man who sells his soul to the devil in exchange for knowledge, and as such, is surely relevant to the themes of this course. But for many reasons, this play works well to exemplify many vexed questions of literary knowledge: because it exists today in two completely different texts, it makes us ask, "what text should we read? How do we know we are reading the "right" text? Because the writer, Christopher Marlowe, was a particularly mysterious and scandalous figure, it makes us ask, "does it make a difference who wrote this play?" And because it is either very religious or very profane, it makes us ask, "how are we supposed to interpret this text? How do we know that we have the "right" reading?" In our sessions together we will be talking about the assumptions that underlie the methods of interpretation, biographical criticism, and textual criticism, at the same time as we will engage in those methods ourselves.

October 3.

Part I: Reading a Play

Assignment: Read Christopher Marlowe, Doctor Faustus in the Signet edition.

For our first class together (Oct. 3), I am expecting you to have read the entire play of Doctor Faustus in the edition that was ordered for the class. Do not worry about understanding every detail of the play: your most important task is to understand the plot and the play's basic themes.

Make sure you read the editor's introduction to the play, which will provide you with general background about Christopher Marlowe and his play. As your introduction mentions, Marlowe's source for the plot of this play was the German Faust Book, a book published in Frankfurt in 1587 (and soon translated into English), which related in lurid detail the adventures of one Johannes Faustus, a wandering conjuror. If you are interested in finding out more about the Faust Book, you can find it on-line at the Perseus Project's website on Marlowe's plays, at http://www.perseus.tufts.edu/Texts/faustus.html.

It will help you to remember that Marlowe was roughly Shakespeare's contemporary -- except that he died long before Shakespeare. Marlowe was writing for the most part in the London of the late 1580's and early 1590's (he died in 1593; Shakespeare died in 1616). Shakespeare's earliest work went on the stage in 1590; in the early nineties he was writing Comedy of Errors, Taming of the Shrew, and Titus Andronicus, among other plays. You might also remember that you should not trust or believe everything you read in an editor's introduction: an introduction is just one scholar's interpretation. This part of the course is designed to teach you to ask questions about all the parts of this edition: the biography, the editor's interpretation, and the nature of the text itself.

Marlowe himself produced a remarkable group of plays in his short lifetime. In all these plays Marlowe showed himself to be obsessed with characters who in some way challenge social norms, and who are "overreachers," in the term used by Harry Levin. Besides Doctor Faustus, These plays include:

Tamburlaine I, Parts I and II ( circa 1587), the story of an ambitious Scythian shepherd-robber who cuts a swath across the Near East, conquering Persia, Turkey, Syria, and Babylon, until his sudden death.

The Jew of Malta (circa 1592), which depicts Barabas, a rich Jew in Malta, and his murderous career of revenge against the equally corrupt Turks and Christians who rule Malta.

Edward II (circa 1593), a history play about the fall of an openly homosexual English king, who defies all his counselors to pursue his love for his favorite, Gaveston.

Massacre at Paris (unknown date), depicting the corruption of French religious politics

Marlowe also wrote a famous erotic poem, "Hero and Leander," about two doomed classical lovers.

As you read the play, please focus on the following questions about the play:

What does Doctor Faustus want to know, and why does he want to know it?

From where/whom does he get his knowledge, and what does he in fact do with it?

What would seem to be the purpose of the Robin-Dick comic interludes? (If you have read Shakespeare's plays, you will recognize this habit of Renaissance English plays for the popular stage, in which "low" or common character often mimic or burlesque the serious themes of the play while providing "comic relief").

Does the play present him to be praised or condemned in these desires and these actions? HOW DO YOU KNOW?

Oct. 5

Part II. What is an Author?

In this session of HDYK, we will focus on the questions surrounding biographical criticism of literary texts: what can we know of the life of an author, and why would it matter that we do know anything about the author? Should knowledge of an author's life influence our reading of a literary text? Should we care about his/her beliefs? And how do we know what those beliefs were, anyway (are we getting them from literary texts, after all)?

You will have two different samples of biographical criticism to read:

The first is a section from C. Tucker Brooke's Life of Marlowe, which was included in a "standard" edition of Marlowe's works published in 1930. In that sense, it is presenting itself as the "standard" biography to accompany the "standard" work. Brooke's biography appeared in the wake of Leslie Hotson's celebrated discovery of documents relating to Marlowe's life and death, including the inquest on Marlowe's spectacular death in a tavern brawl, and the famous "Baines document," in which are recorded the accusations that one Richard Baines made concerning "the opinion of one Christopher Marly concerning his damnable Judgment of Religion and scorn of God's word." In the sections of Brooke's biography that you will read, you will see Brooke not only trying to reconstruct a life from some paltry evidence, but also trying to interpret that evidence and to reconcile the documentary evidence with the image of Marlowe that he has formed from his own reading of the works.

To understand the significance of the Baines document, you need to know that this period in English history was defined by intense religious conflict between Catholics and Protestants. Henry VIII declared England's independence from the Pope and the Church of Rome in 1534, and thereafter, the reigns of his descendants saw wild swings between the stricter Protestantism of his son Edward VI and the return to Catholicism imposed by his sister Mary I. When Elizabeth I came to the throne in 1558, she reestablished the Protestant religion, but with a sense of moderation. In Marlowe's time, however, Catholicism was still associated with treason (and indeed, Catholic plots against Elizabeth were often being hatched abroad). The average English Protestant would have associated Catholicism with elaborate ritual, rebellion, and corruption.

The second reading is from Andrew Butcher's introduction to William Urry's unusual biography of Marlowe, called Marlowe and Canterbury (published in 1988). Urry's biography of Marlowe is very different from Brooke's: instead of being a literary historian trying to give an overview of Marlowe's life, Urry was a archivist of the city of Canterbury, and he wanted to give as full as possible a picture of what Canterbury might have been like during Marlowe's childhood, and how this might have shaped him as a writer. Butcher, who provides an posthumous introduction to Urry's book, is a literary (and linguistic) historian who has been influenced by the new movements in literary studies that have made the practice of literary biography problematic. As you read the section from Butcher's introduction, look for the ways in which he struggles with what it means to account for Marlowe as an author in a critical climate that has abandoned the notion of the author as originating and determining the meaning of what he/she writes. Note, for example, the way that he dwells at length on details of small events, such as the betrothal of Dorothy and Richard, instead of the "big events" of political history, as a way of defining Marlowe's world.

The title of this section and Butcher's preface have been influenced by the work of Michel Foucault. Michel Foucault (1926-1984) was a French intellectual whose work on social and intellectual history and theory(including Discipline and Punish, The History of Sexuality, and Madness and Civilization) has had a powerful effect on literary studies since the 1980s. His essay "What is an Author" (first published in French in 1969 and in English in 1979), offers a provocative attack on the traditional notion of "what is an author," and the kind of questions we ask about the relationship between authors/writers and their texts. Here are some questions that Foucault wishes that you would not longer ask about authors -- and the ones he would like you to ask (do try to make your way through his jargon):

"I seem to call for a form of a culture in which fiction would not be limited by the figure of the author. All discourses, whatever their status, form, value, and whatever the treatment to which they will be subjected, would then develop in the anonymity of a murmur. We would no longer hear the questions that have been rehashed for so long: Who really spoke? Is it really he and not someone else? With what authenticity or originality? And what part of his deepest self did he express in his discourse? Instead, there would be other questions, like these: What are the modes of existence of this discourse? Where has it been used, how can it circulate, and who can appropriate it for himself? What are the places in it where there is room for possible subjects? [By "subjects" here Foucault mean something like "self" or "subject" in a grammatical sense, the originator of an action] Who can assume these various subject functions? And behind all these questions, we would hear hardly anything but the stirring of an indifference: What difference does it make who is speaking? [here Foucault is quoting a line from Samuel Beckett]

Questions to ask yourself as you read:

Why is Brooke so anxious about the facts of Marlowe's biography? How does he use evidence -- both the evidence of the Baine's document -- and the evidence of Marlowe's literary work -- to write this life? Which comes first for him?

What does Butcher care about? What kind of evidence does he think is important? How does he connect the life and the work?

[see attached articles: section from Brooke and Butcher's introduction to Urry)

Oct.8

Part III. Textual Criticism

In this section of HDYK, we will consider the matter of textual criticism, that is, the practice of editing literary texts. What in fact are we reading when we read a "text" of a play, poem or novel? When we have several different extant versions of a literary work (e.g., a manuscript, a revised typescript, or different printed versions), how do we establish the "correct" text to read? Who decides what is "correct," and on what basis, when there are different versions?

For this section of the course, we will be reading a selection of W.W. Greg's edition of Marlowe's "Doctor Faustus" 1604-1616, Parallel Texts (first published in 1590). This edition reproduces, on facing pages, the two different printed versions of Doctor Faustus, the so-called "A-text" of 1604 and the "B-text" of 1616. We will be discussing the general differences between the two texts' version of the final scenes of the play, so please read them with attention to the significant differences between the two version (your paper assignment will focus on the differences between these two versions of the final scene of the play, on pp. 288-293). If you like, you can also consult the Perseus project web-site's version of the "parallel text" version of Doctor Faustus -- URL http://www.perseus.tufts.edu/Texts/faustus.html : on this site you can also see on facing "pages," the two different printed version of Doctor Faustus,

As you read, you might remember that Marlowe died in 1593, but the first edition we have is dated 1604 (the "A-text"); the earliest recorded performance of the play was in 1594. Another version of the play appeared in 1616 (the "B-text"). The A-text is 1517 lines long, while the B test has 2121 lines (while A has some line that are no in B). The scenes in B that are not in A include many of the comic scenes and the demonic machinery (appearance of Lucifer, etc.)

As you compare the two texts, you should also be comparing these texts with the edited version of Doctor Faustus that you read for the first session -- which resembles neither the A or the B text exactly, but is rather an editorial "collation" and "improvement" on both texts.

Consider this statement by Michael Warren, from his article called "Doctor Faustus: The Old Man and the Text," (published in English Literary Renaissance 11(1981), pp. 11-147). Warren says of the practice of doing editions of early texts like this one:

"The pretensions to scientific accuracy of modern bibliography have often led to misleading statements. True facts are very few in the bibliographical study of plays like Doctor Faustus; for instance, we know few things certain about the origins of these texts, having no certain indication of the dates of the play's composition or first performance; we do not even know whether the 1604 publication of the A-text is the first edition of the play. We merely have reason to be thankful that we know of the A-text's existence, since only one copy of it survives. Yet assumptions are made... and surmises rapidly become "facts." In opposition, I wish to propose that bibliography is not a relatively accurate scientific activity but rather a subjective art in which at all time interpretation precedes analysis and deduction.. It is essential ...to remember that both the A- and B-texts were published as works having their own validity; although some may hypothesize a single common original and postulate mutual relations, all that we know is that each was published as an independent book, presumably as a script that bore some relation to the theater."

[Here: attach the paper topic:]

HDYK: Writing Assignment on Doctor Faustus

In this exercise, you will evaluate the A and B text versions of Marlowe's Doctor Faustus. Your thoughts should be summarized in a well-organized essay of three to five pages.

1. Give an account of the differences between the A and B text versions of the final scene of Doctor Faustus (pp. 288-293)

2. Describe what you think is the significance of the difference: what difference does choosing the A or B text version make for your/our reading of the play?

3. Consider what choices you might make as an editor of the play: would you conflate parts of the A and B version, or choose one or the other

[Here also attached the relevant pages from Greg's edition]

Oct. 10.

Part IV. Literary Interpretation -- in retrospect.

In this final class you will be discussing the play in separate discussion sections. In these classes you will finally turn to the question of literary interpretation. In the discussion section, you will consider the conventions and assumptions of the practice of literary interpretation, and how they might look different in the wake of the questions we have raised about biographical and textual criticism.

How do we know what a text means? How do we test the validity of an interpretation of a text? What counts as evidence, and what does not? How do we use biographical and historical information as evidence? What text do we use to validate our interpretation (and what comes first -- interpretation or choice of text)? How is evidence used in literary interpretation?

Your readings for this class are two very different interpretations of the religious meaning of Doctor Faustus: both address the question of whether we are to understand this play as statement of orthodox Christian morality or a subversion or critique of Christian belief.

The first reading is taken from Douglas Coles' Suffering and Evil in the Plays of Christopher Marlowe (published in 1972). Coles operates on the premise that in this play Marlowe, "whatever his personal views of Christianity might have been, has fashioned a play that is thoroughly Christian in conception and import" (p. 194). Working from the idea that Marlowe must have been very familiar with theology from his studies as Cambridge, Coles reads Faustus as an ironic study of defiance (we will be focusing on the notion of irony, so look for the ways that Coles uses this concept).

The second reading, from Jonathan Dollimore's Radical Tragedy (first published in 1984) originated in a very different critical climate. While Coles wrote at a time when "historical" reading meant seeing Renaissance writers as believing that a Christian world of order and hierarchy, Dollimore was among the first of a new wave of Marxist-influenced British critics who were interested in understanding Renaissance texts as "radical" statements. So Dollimore see Faustus as an "exploration of subversion through transgression," that is, of undermining a world of order, morality, and hierarchy. Where Coles sees irony, Dollimore sees paradox: we will be discussing the difference between these concepts, and what both critics use as evidence for their interpretations. (Note: when Dollimore refers to a conflict as "Manichean," he is referring to the theological opinion that the world is defined by the epic conflict between the Good of God and the Evil of the Devil (as opposed to the view of the world that see Evil only as the absence of Good, thus robbing evil of any real substance or power).

[see attached selection from Coles and Dollimore]

Module 4 (Paul Rozin) Drawing conclusions: Inferential statistics and science and the media

Tuesday, October 16 Introduction to Statistics 2 : Statistical significance and effect size

Friedman, D., Pisani, R., & Purves, R. (1998) Statistics (Third Edition). New York: W. W. Norton.

(The basic concepts will be explained in lecture. This is back up reading, and takes things somewhat further than the lecture)

Chapter 13. What are the chances? (221-232)

Chapter 14. More about chance? 237-245

Chapter 15. The binomial formula 255-258

Thursday, October 18

Introduction to Statistics: 2: Statistical significance and effect size

Rozin: From the laboratory to the media and the public mind: The oat bran controversy

Rozin, P. Background and notes on the oat bran papers.

deGroot, A. P., Luyken, R., & Pikaar, N. A. (1963). Cholesterol lowering effect of rolled oats. The Lancet, 2, 303-304.

Kirby, R. W., Anderson, J. W., Sieling, B., Rees, E. D., Chen, W-J. L., Miller, R. E., & Kay, R. M. (1981). Oat-bran selectively lowers serum low-density lipoprotein cholesterol concentrations of hypercholesterolemic men. The American Journal of Clinical Nutrition, 34, 824-829

Gold, K. V., Davidson, D. M. (1988). Oat bran as a cholesterol-reducing dietary adjunct in a young, healthy population. Western Journal of Medicine, 148, 299-302.

Going gaga over oat cuisine. Time, Sep. 12, 1988

Kowalski, R. E. (1989). The 8-week cholesterol cure. New York: Harper & Row.

Pp. xix-xxii Introduction to the completely revised edition

Pp. 1-7. Introduction

Pp. 78-83. Ch. 5. Getting the scoop on oat bran

Swain, J. F. et al. (1990). Comparison of the effects of oat bran and low-fiber wheat on serum lipoprotein levels and blood pressure. New England Journal of Medicine, 322, 147-152.

Hold the oat bran. Time, Jan. 29, 1990, p. 80.

Are oat bran's claims half baked. U.S. News and World Report, Jan. 29, 1990

Oat bran heartburn.

Is it still the right thing to do? Newsweek, Jan. 29, 1990, pp. 50-52.

Ripsin, C. M. et al. (1992). Oat products and lipid lowering: A meta-analysis. Journal of the American Medical Association, 267, 3317-3325.

Oat bran project

The purpose of the readings is to give you a sense of how a consensus on a particular issue may arise, even in the face of some discordant findings. It is also supposed to give you some insight into the interaction between the science, the personality/ambition of the scientists, the media, and the public.

Your project, having completed these readings, is to construct two one page (double spaced, text only) advertisements, one touting the health benefits of an oat bran product, and one minimizing them, as part of an ad for a wheat bran product. You should try to make the most appealing and convincing case possible, without actually presenting false information. You can be selective, but you cannot let yourself be open to suit for misrepresentation.

Ian Lustick sections

The first of three lecture sessions will explain the historical and current foreign policy background to the readings from Unsettled States, Disputed Lands, which I want the students to have already started reading. I will also explain the logic of the readings for the next two sessions and provide a guide for their effective reading of the assignment for the first Thursday of my module. To complete this students will need to do more reading over the weekend, i.e. for the first meeting, than we will actually cover in that first class.

For the first (Tuesday) meeting students should have read: (October 23)

Unsettled states, Disputed lands

Introduction to Part I: "The Changing Shape of States" (pp. 1-6)

Chapter 1: "Israel and the West Bank and Gaza Strip: Disengagement or Incorporation?" (pp. 7-25)

Chapter 2: "Thresholds of State-Building and State Contraction" (pp. 26-51)

Introduction to Part II: "Wars of Position and the Fate of Hegemonic Projects" (pp. 53-56)

The first part of Chapter 3: "Becoming Problematic: Breakdown of a Hegemonic Conception of Ireland" (pp. 57-70)

During the second meeting (Thursday, October 25) I will analyze the problem of how one might know whether beliefs are really or effectively "inconceivable" or "immutable", i.e. hegemonic and relate this social scientific question to fundamental questions in political philosophy, such as Plato's idea of the "noble lie," Machiavelli's theory of the founding of religions. I will provide the students with a guide to understanding the readings for the following Tuesday, which use the comparative method to decide whether or not we know, in the British-Irish and French-Algerian cases, about the presence or absence of hegemonic beliefs.

For this second meeting the students should have read: (October 25)

The second half of Chapter 3: "Becoming Problematic: Breakdown of a Hegemonic Conception of Ireland" (pp. 70-80)

Chapter 4: "Where and What is France? Three Failures of Hegemonic Construction"

(pp. 81-120)

During my third and last lecture meeting with the students I will pay close attention to the use of multiple and converging streams of data and systematic comparison across time and space to achieve confident judgments about the questions posed in the beginning of the module. We will aim to understand not only how we know whether hegemony existed in these cases but how we could know more generally by applying the techniques developed and applied here. Toward that end I suggest, for the TA led session on the second Thursday of my module, that the students also read an article which used these techniques to successfully predict the failure of a hegemonic project regarding Jerusalem -a prediction likely to be tested by events during the fall semester.

Assigned readings for this second Tuesday class (October 30) are:

Chapter 5: "Patterns of Hegemonic Change: Britain and Ireland, France and Algeria" pp. 121-186

Assigned reading for the second Thursday, TA led session:

Ian S. Lustick, "The Fetish of Jerusalem: A Hegemonic Analysis," in Israel in Comparative Perspective: Challenging the Conventional Wisdom, Michael N. Barnett, ed. (Albany: SUNY Press, 1996) pp. 143-72.

How Do We Know? -- Effects of Nutrition, Hormones, and Social Environment

on Health and Disease

Ingrid Waldron, Biology Department, 203 Leidy Lab, 215-898-8396

This section of the course will discuss the different types of evidence used to evaluate the effects of behavior, physiology, and social environment on health and disease. We will evaluate the strengths and weaknesses of different types of studies and discuss the reasons why these different types of studies often yield apparently conflicting results. We will also discuss why different studies of the same type may yield differing results. These methodological points will be illustrated by considering evidence concerning several specific topics:

(1) the effects of diet on risk of cancer (specifically, the effects of beta-carotene on lung cancer risk),

(2) the effects of sex hormones on risk of heart disease, and

(3) the reasons why low socioeconomic status is associated with higher rates of illness and mortality.

For this section of the course, you should read the assigned readings for each day before coming to class, so you will be able to follow the class presentation and participate in the discussions. I have assigned relatively few pages of reading, but some of the readings will be challenging to understand. Please note any questions you have as you read the assigned readings and ask me your questions in class. Please bring your copy of the assigned readings to class, since this will make it easier for you to follow our discussions of specific tables and figures from the readings. After class, you may find it useful to reread some of the assigned readings for full comprehension.

Please read the articles in the order assigned, since the concepts and definitions introduced in earlier articles and the notes for these articles will assist you in understanding later articles. The notes provide explanations of many of the technical terms in the assigned readings. To link the explanations in these notes to relevant sections of the assigned reading, match the numbers in the margins of the notes with the corresponding numbers in the margins of the assigned readings. Some technical terms are not explained, either because the term is defined within the assigned article, or because the term is not crucial for understanding the main arguments, and explanations of all terms would introduce too much complexity. However, you are invited to ask for definitions and explanations of any terms or concepts you would like to know about.

_________________________________________________________

November 7, 2000 - Does Dietary Intake of Beta-carotene Influence Risk of Lung Cancer?

Reading A briefly describes the different types of scientific studies which provide evidence concerning effects of behavior and environment on health and disease. Readings B and C review evidence from different types of studies concerning the effects of beta-carotene intake on risk of lung cancer; these reviews evaluate the strengths and weaknesses of the various types of studies in an attempt to reconcile apparently contradictory findings. Reading D provides additional commentary on the interpretation and description of scientific findings.

Questions to think about as you do the reading:

-- What conclusions does Albanes draw concerning whether increased beta-carotene intake reduces lung cancer risk? What evidence suggests that higher beta-carotene intake reduces the risk of lung cancer? What evidence indicates that higher beta-carotene intake does not have a protective effect? How does Albanes interpret the apparently contradictory findings?

-- How does Koo's interpretation of the observational studies concerning diet and lung cancer risk differ from Albanes'?

-- What conclusions does Albanes draw concerning the effects of fruit and vegetable consumption on lung cancer risk? What are the strengths and weaknesses of the evidence for this conclusion?

-- What type of study do you think would be the most useful to further our understanding of nutritional effects on lung cancer at this time?

Assigned Readings (and explanations of terms and concepts in the readings):

A. Brody, J. (1998) A Study Guide to Scientific Studies. New York Times, Aug. 11, page F7.

B. Albanes, D. (1999) Beta-carotene and Lung Cancer. American Journal of Clinical Nutrition 69(suppl): 1345S-1350S.

For definitions of types of studies, see Brody article. Randomized intervention trials are described by Brody as clinical trials.

Beta-carotene is a molecule which can be converted to vitamin A and can function as an "antioxidant" which may help to prevent oxidants from damaging DNA; both of these effects may help to prevent cancer (an antineoplastic effect). Beta-carotene is plentiful in dark leafy green vegetables (such as spinach) and rich yellow or deep orange vegetables and fruits (such as carrots and cantaloupe). Beta-carotene is one of the carotenoids, which share similar basic chemical structures, but differ somewhat in structure and function. Beta-carotene intake can be assessed either by evaluating the amount of beta-carotene in foods consumed or by measuring levels of beta-carotene in the blood (or in serum or plasma, the liquid portion of the blood) (referred to as biochemical status or serology). Various forms of vitamin A are known as retinoids and include retinol and retinyl palmitate.

Incidence refers to the rate of developing new cases of a disease such as lung cancer.

Relative risk refers to the ratio of rates of a disease in two groups with different levels of nutritional intake (e.g. the incidence of lung cancer for people with low beta-carotene intake divided by the incidence of lung cancer for people with high beta-carotene intake). A relative risk of 1.0 would indicate that lung cancer risk does not vary with beta-carotene intake, whereas a relative risk greater than 1.0 would indicate that people with lower beta-carotene intake have a higher risk of lung cancer. More generally, relative risk can be used to compare rates of disease for two groups who differ in any preventive or risk exposure. The relative excess refers to the percentage by which the risk in one group exceeds the risk in another group. When statisticians estimate relative risk or relative excess, they often also estimate a 95% CI (confidence interval); statistical calculations indicate that 95 times out of 100 the true value of the relative risk or relative excess will be contained within the range of values designated by the 95% confidence interval.

Correlations which are found in observational studies do not necessarily imply causal effects. For example, a correlation may be due to chance variation in a specific sample, or a correlation may be due to confounding factors which may influence both of the correlated variables. Criteria which are used to infer that an observed correlation is probably due to a causal effect include dose-response gradient (e.g. decreasing lung cancer risk as beta-carotene intake increases), biological plausibility (e.g. identification of biological mechanisms by which beta-carotene intake would be expected to reduce cancer risk), consistency (the same association observed repeatedly in different studies), risk level (a large enough association that it is considered unlikely to be due to confounding factors), and temporal correctness (the proposed causal agent is observed before the proposed effect). Obviously, the strongest evidence to support a causal interpretation of an observed correlation would be an experiment in which manipulation of one of the correlated variables induces the expected change in the other correlated variable.

Etiologic associations refer to causal factors that contribute to the development of a disease such as lung cancer.

Alpha-tocopherol is the most biologically active form of vitamin E. Micronutrients include vitamins and minerals that are required in the diet in small quantities. Phytochemicals are other chemicals found in plants which do not have an identified nutritional function but may influence health and disease risk.

Pharmacologic doses refer to high doses that can be achieved by taking pills, but not by normal dietary intake.

Collinear refers to highly correlated variables such as the different types of dietary carotenoids.

Endpoints refer to the specific measures of health or disease assessed, for example incidence or mortality due to lung cancer or cardiovascular diseases (including heart disease and stroke).

C. Koo, L. (1997) Diet and Lung Cancer 20+ Years Later: More Questions Than Answers? International Journal of Cancer Supplement 10:22-29.

Note: This paper reviews evidence concerning the effects of beta-carotene and other dietary factors on risk of lung cancer; we will focus on the evidence concerning beta-carotene, but the evidence concerning other dietary factors, particularly fruits and vegetables, will also be relevant for our discussion.

Carcinogenesis refers to the development of cancer.

Odds ratio is roughly similar to relative risk. When researchers calculate odds ratios for people with high vs. low consumption of a type of food or a particular nutrient, they often use statistical techniques to correct for differences in age, smoking, etc. that are correlated with nutritional habits; this is referred to as adjustment for confounding factors. Inadequate or imperfect control refers to the fact that these statistical corrections may not fully account for differences in smoking (for example, the researchers may not have complete data on lifetime history of smoking habits).

Histological type refers to the particular type of cells observed in a cancer (for example, squamous cells). Cancers are sometimes called tumors.

D. Kolata, G. (2000) Health advice: A matter of cause, effect and confusion. New York Times, April 25, pp. F1 and 6.

_________________________________________________________

November 9, 2000 - Do Female Sex Hormones Reduce the Risk of Coronary Heart Disease?

Reading E reviews a broad range of evidence concerning the effects of sex hormones on risk of coronary heart disease, with a particular focus on understanding the contributions of endogenous sex hormones to sex differences in coronary heart disease mortality. (Endogenous hormones are the hormones produced by the body, as opposed to exogenous hormones which are medically prescribed hormones from an external source.) Reading F describes a study which tested the effects of women's endogenous sex hormones by assessing whether a woman's risk of coronary heart disease is increased if she has an earlier menopause (when a woman's sex hormone levels fall substantially).

The last two readings focus on the effects of exogenous female sex hormones which are taken by some women after menopause. Reading G summarizes and evaluates observational studies which have compared coronary heart disease risk for women who do vs. don't take hormones after menopause. Reading H reports the results of a clinical trial in which women were randomly assigned to receive sex hormones or placebo. As you will see, the results of these two types of studies appear contradictory, and both readings provide useful interpretations that can suggest how to reconcile these apparently contradictory findings.

Questions to think about as you do the reading:

-- What evidence supports the hypothesis that female sex hormones (estrogens and progesterone) reduce the risk of coronary heart disease?

-- What evidence indicates that female sex hormones do not reduce the risk of coronary heart disease?

-- What explanations could account for these apparently conflicting findings concerning the effects of female sex hormones on risk of coronary heart disease? (Remember that each hormone has multiple physiological effects, and the balance between beneficial and harmful effects may vary depending on the dose and the specific hormone considered.)

Assigned Readings (and explanations of terms and concepts in readings):

E. Waldron, I. (1995) Contributions of biological and behavioral factors to changing sex differences in ischemic heart disease mortality. In Adult Mortality in Developed Countries: From Description to Explanation, Eds. Lopez et al., Oxford University Press. (pp. 161-5 only)

Ischemic heart disease is another name for coronary heart disease (also referred to as CHD). The coronary arteries bring blood, with needed oxygen etc., to the heart muscle. Coronary heart disease results when these coronary arteries are blocked by a buildup of atherosclerosis (deposits of cholesterol etc. in the artery walls) and blood clots. A myocardial infarction is a heart attack, when the heart is damaged and may stop beating due to inadequate blood flow in coronary arteries. Coronary heart disease can produce pain without a heart attack, and this is known as angina pectoris. Cholesterol is carried in the blood in two types of particles: LDL (low-density lipoprotein) and HDL (high-density lipoprotein); the buildup of atherosclerosis is increased by LDL cholesterol and decreased by HDL cholesterol. Serum lipids refer to levels of LDL, HDL, total cholesterol, and triglycerides (fats) in the blood.

A woman's ovaries produce estrogen (=oestrogen) and progesterone. Oophorectomy is the removal of the ovaries. During natural menopause, there is a sharp reduction in the ovaries' production of sex hormones. Postmenopausal hormone replacement therapy or contraceptive pills are referred to as oral estrogens, since they are taken by mouth; this results in very high levels of estrogens reaching the liver, much higher than for endogenous estrogens. Transdermal estrogens diffuse from a patch across the skin into the circulation; these do not result in high estrogen levels in the liver.

F. van der Schouw, Y. T., et al. (1996) Age at menopause as a risk factor for cardiovascular mortality. Lancet 347:714-718.

Note: Skip the Patients and methods section (pp. 714-6, indicated by vertical dashed lines in the margins), since this section is too technical and complex for our purposes.

This study attempts to assess the effects of endogenous female hormones on risk of cardiovascular mortality, by assessing the effects of early menopause. Coronary heart disease constitutes approximately half of cardiovascular mortality.

Cox regression analysis is a statistical procedure that yields an annual hazard which in this context is roughly the risk of dying of cardiovascular disease per year for women who had menopause at a given age. The age-adjusted hazard ratio is approximately the risk of cardiovascular mortality for women who had menopause at a given age divided by the risk for women who had menopause one year younger.

Hysterectomy refers to the removal of the uterus, in this context the removal of uterus without removal of the ovaries.

G. Barrett-Connor, E. and Grady, D. (1998) Hormone replacement therapy, heart disease, and other considerations. Annual Review of Public Health 19, read pp. 55-62 and 66-7

Note: Skip the section on Estrogen Use and Other Risks and Benefits (pp. 63-66, indicated by vertical dashed lines in the margins), since this is not directly relevant for our purposes.

A meta-analysis is a quantitative method for synthesizing results from multiple studies of the same research question.

A progestin is a synthetic molecule used in hormone replacement therapy or oral contraceptives which has some effects similar to those of natural progesterone, and usually some additional effects which are unlike the effects of natural progesterone. Endometrial hyperplasia refers to excessive growth of the lining of the uterus (indicates an increased risk of uterine cancer). Venous thromboembolic events refer to medical problems caused by blood clots in the veins.

Unopposed conjugated equine estrogen refers to the type of estrogen without progestin which was used in the early years of hormone replacement therapy. In the Postmenopausal Estrogens/Progestin Interventions Study, conjugated equine estrogen (CEE) was combined with one of two different progestins (medroxyprogesterone acetate = MPA or cyclic micronized progesterone = MP) which had differing effects on serum lipids. Estrogen may reduce coronary heart disease risk by several effects which reduce the risk of atherosclerosis and increase blood flow in coronary arteries; these potentially beneficial effects of estrogen include decreased LDL and increased HDL lipoproteins, reduction of propensity to blood clotting (beneficial effects on fibrinogen, PAI-1 and antithrombin III), and increased diameter of coronary blood vessels (due to increased vasodilation and decreased vasoconstriction or reactivity).

The primary estrogen produced by ovaries is estradiol.

H. Hulley, S., et al. (1998) Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Journal of the American Medical Association 280: pp. 605-6 and 609-12.

Note: Read p. 605 and the first paragraph on p. 606. Skip the METHODS section and the RESULTS section (pp. 606-610, indicated by vertical dashed lines in the margins), since these involve too much technical terminology and complexity for our purposes. Examine Table 2 and Figure 3 (p. 609), and read COMMENT through the end of the article (pp. 610-612).

"Secondary" is used with two different meanings in this study. Secondary prevention refers to the prevention of additional manifestations of coronary heart disease in women who already have coronary heart disease; these are referred to as recurrent events. Secondary cardiovascular outcomes refers to a variety of manifestations of coronary heart disease or illnesses which results from atherosclerosis in arteries in the brain (stroke or transient ischemic attack) or in the legs, etc. (peripheral artery disease). The primary CHD events are myocardial infarctions and CHD death.

Statisticians calculate power as follows: if an effect of a certain size exists in the population, how likely is a given study to find statistically significant evidence for this effect? If the number of cases in the study is too small, then this study is unlikely to find a statistically significant result even when an effect is present, so the study has low power to detect this effect.

The Cox proportional hazards model used here is the same method as the Cox regression analysis used in the previous study on age at menopause. The relative hazard in this study is the ratio of risk of CHD events for women who were taking estrogen-progestin vs. women who were taking placebo; this is generally similar to the hazard ratio in the study on age at menopause. The survival curves show the cumulative incidence (percent of women who have had a CHD event since entering the study) for women who received estrogen-progestin vs. women who received placebo. Most of the analyses shown are intention-to-treat analyses where women are included in the estrogen-progestin group or the placebo group based on their initial assignment. The as-treated analysis includes only women who actually took the pills they were assigned to take.

Venous thromboembolic events include blood clots in the veins (deep vein thromboses) and in the lungs (pulmonary emboli). The tendency of oral estrogens to increase blood clotting is referred to as a prothrombotic effect.

_________________________________________________________

November 14, 2000 -- What are the Effects of Socioeconomic Status on Health and Mortality? What Causal Mechanisms Link Low Socioeconomic Status to Illness and Mortality?

Reading I reviews evidence concerning the relationship between socioeconomic status and risk of disease and mortality. Reading I also provides an overview of the proposed causal mechanisms which may link low socioeconomic status to increased risk of illness and mortality. Reading J uses data from two national samples to evaluate the importance of several of these hypothesized causal mechanisms. Specifically, this study tests whether lower educational level is associated with poor work and economic conditions, fewer psychosocial resources and more health-damaging behavior. In addition, this study tests whether these variables may be responsible for the poorer health of less educated adults. Reading K briefly summarizes a recent study which found that, although low income adults had poorer health behavior, this was only a minor cause of their poorer health. This reading goes on to introduce the hypothesis that low socioeconomic status environments during both childhood and adult life may contribute to psychological and physiological characteristics that increase adult mortality risk. Reading L describes a study which tests parts of this hypothesis by evaluating the relationship between childhood socioeconomic status and several psychological characteristics, and assessing the contribution of these psychological characteristics to the relationship between childhood socioeconomic status and adult health.

For all these papers, it will be useful to keep in mind the following conceptual framework. If socioeconomic status influences a variable (such as smoking behavior or a person's sense of control) and if this variable in turn influences illness and mortality risk, then this variable is called a mediating variable and is considered to be a causal link between socioeconomic status and health. In order to identify the mediating variables which link socioeconomic status to health, researchers test three relationships. First, the researchers test whether socioeconomic status is associated with the proposed mediating variable(s). Second, the researchers test whether the proposed mediating variable(s) are associated with the health outcome. Third, the researchers compare the strength of the relationship between socioeconomic status and health in statistical analyses which include or omit adjustment for the proposed mediating variable(s). If the strength of the relationship between socioeconomic status and health is weaker in analyses which adjust for the proposed mediating variable(s), then this is considered evidence in support of the hypothesis that these mediating variable(s) provide a causal link between socioeconomic status and health. If the relationship between socioeconomic status and health becomes statistically nonsignificant in analyses which adjust for the mediating variable(s), these mediating variable(s) are said to "account for" or "explain" the effects of socioeconomic status on health. If the relationship between socioeconomic status and health remains statistically significant in analyses which adjust for the mediating variable(s), then researchers use expressions such as the mediating variable(s) "do not eliminate" the effect of socioeconomic status on health, or socioeconomic status has a direct effect on health in addition to acting via the mediating variable(s), or socioeconomic status has an effect on health independent of the mediating variable(s).

In thinking about analyses of this type, it is important to keep in mind the following question:

How do researchers distinguish between mediating variables and confounding variables?

(A confounding variable would be a variable which has causal effects on both

socioeconomic status and health.)

Questions to think about as you do the reading:

-- How is a person's socioeconomic status assessed?

-- What is the relationship between socioeconomic status and health?

-- What are the specific causal mechanisms that are hypothesized to link socioeconomic status to health?

-- What is the evidence for and against each of these hypothesized links?

Assigned Readings (and explanations of terms and concepts in reading):

I. Adler et al., 1993, Socioeconomic Inequalities in Health, Journal of the American Medical Association 269:3140-3145

Morbidity refers to illness, and premature mortality refers to death before old age. Cardiovascular disease includes hypertension, heart disease and stroke. Chronic degenerative diseases include cardiovascular diseases, cancer and other diseases which are the result of long-term changes in the body and are not directly caused by infections.

Linear relationship refers to a gradual increase in health problems as socioeconomic status decreases, whereas a threshold effect would be observed if the relationship between socioeconomic status and health were a step function, with high health problems below a certain level of socioeconomic status and low health problems above that threshold level.

Relative risk refers to the ratio of risk for mortality or disease in two groups with different socioeconomic status. For example, in one study the relative risk of mortality for unskilled laborers relative to top administrators was 2.7, which means that unskilled laborers had nearly three times as high a mortality rate as top administrators.

Primary care refers to checkups, vaccinations and other care to maintain health for people who are not sick. Analogously, primary prevention refers to changes in behavior, vaccination and other actions taken by people who are well to prevent future illness. Secondary prevention refers to actions taken by people who have an illness to prevent recurrence or deterioration of the condition.

Papanicolaou or Pap Tests take small samples from the cervix (the opening to a woman's uterus) to detect early signs of cervical cancer.

Risk factors include physiological parameters which predict increased risk of disease; for example, high LDL cholesterol or blood pressure are risk factors for coronary heart disease. Other risk factors for heart disease include triglyceride levels (levels of fat in the blood), glucose tolerance (a measure of diabetes risk), and body mass index (a measure of obesity).

Malignant melanoma is the most serious form of skin cancer.

Pathogens are bacteria, viruses, etc. which cause infectious disease. Carcinogens are agents which cause cancer.

J. Ross, C.E. and Wu, C. (1995) The links between education and health. American Sociological Review 60: 719-733 and 738-740.

Note: Read from the beginning of the article to the first two paragraphs on page 733, but omit the footnotes (indicated by vertical dashed lines in the margins), since these are excessively detailed and complex for our purposes. Omit the remainder of the RESULTS section (pp. 733-738, indicated by vertical dashed lines in the margins), because the amount of detail may be overwhelming! Read SUMMARY AND CONCLUSIONS (pp. 738-740).

Note that this study focuses on effects of one aspect of socioeconomic status, namely education.

A cross-sectional study measures all variables at the same point in time, whereas a longitudinal study follows individuals over time. A longitudinal study can provide more evidence concerning causal order, i.e. which of two correlated variables has a causal effect on the other. For example, by using a longitudinal study to assess the relationship between initial socioeconomic status and subsequent change in health in a longitudinal study, the researchers have greater confidence that socioeconomic status is influencing health, rather than health having an influence on socioeconomic status.

The stratification system refers to the social system in which different levels of socioeconomic status provide, on average, different in opportunities and harmful exposures.

Standard deviation is a measure of variability, roughly how much variation there was in different subjects. Similarly, standard error is a measure of the variability in regression coefficients (see below).

In this study the relationships between education and health are evaluated using multiple regression analysis. Self-reported health or physical functioning is the dependent variable in each regression equation, and the regression coefficients estimate how much the health variable differs for people with different levels of each independent variable, such as education. (We can ignore the distinction between standardized and unstandardized regression coefficients.) In Table 3, the decrease in magnitude of the regression coefficient for education as more explanatory or independent variables are included in the regression equations indicates that these additional variables (for example, exercise and smoking) can account for part of the association between education and health. This is interpreted as evidence that socioeconomic status influences exercise and smoking which in turn influence health outcomes; i.e. exercise and smoking appear to be mediating variables. Comparing equation 1 with equation 4, we see that the introduction of work and economic conditions variables, social-psychological resource variables, and health lifestyle variables reduces the unstandardized regression coefficient for education from .076 to .031, so these variables are said to explain more than half of the association between education and health.

A false positive is a test which indicates a health problem even when none is actually present.

K. Williams, R. B. (1998) Lower socioeconomic status and increased mortality -- Early childhood routes and the potential for successful interventions. Journal of American Medical Association 279:1745-6.

The sympathetic nervous system is activated during stressful situations and stimulates increased heart rate, release of adrenaline, etc., whereas the parasympathetic nervous system is active during nonstressful times and stimulates digestion, slower heart rate, etc.. The sympathetic and parasympathetic together make up the autonomic nervous system.

Serotonin is a chemical messenger in the brain, including the hippocampus which plays a crucial role in learning.

L. Bosma, H., et al., (1999) Social class in childhood and general health in adulthood: Questionnaire study of contribution of psychological attributes. British Medical Journal 318: 18-22.

External locus of control refers to a person's tendency to perceive that what happens to an individual is generally not controlled by the individual's behavior, but rather by external factors. Neuroticism refers to a person's tendency to experience emotional distress.

Hypertension refers to high blood pressure, and hypercholesterolemia refers to high levels of cholesterol in the blood.

In this study, the association between health and childhood social class was assessed by logistic regression analysis which yields adjusted odds ratios, which are similar to relative risks which have been adjusted for confounding and mediating variables. These odds ratios provide comparable information to the regression coefficients in the previous study.

Childhood socioeconomic status may influence adult health both by influencing adult socioeconomic status and by influencing adult psychological characteristics which may affect health, above and beyond any effect due to adult socioeconomic status. To evaluate the effect of childhood socioeconomic status "independent of adult social class", this study uses regression analyses which include adjustment for adult occupational level.

_____________________________________________________________________________

November 16, 2000 -- Discussion

Readings M and N provide additional cautions concerning the interpretation of results from studies of environmental and medical effects on health and disease. Please also review chapters 1 and 2 from Statistics by Freedman et al. (assigned for September 12).

Questions to think about as you do the reading:

-- What issues do you need to be aware of in order to avoid drawing false conclusions concerning the influence of environmental factors or medical procedures on health?

-- What practices should scientists and journalists adopt in reporting results of epidemiological studies?

Assigned Reading:

M. Brody, J. (1998) Health scares that weren't so scary. New York Times, Aug. 18, p. F7

N. Taubes, G. (1995) Epidemiology faces its limits. Science 269: 164-169.

Radon is a radioactive gas that seeps out of the earth into basements in some regions. The many other postulated causes of health risks are not defined in these notes, since most of these do not have established relationships to health, and the point of this assigned reading is to introduce general methodological problems in the studies; however you're invited to ask about any you would like to know about.

Lymphomas, myelomas and leukemia are different types of cancer.

False positive results occur when a study shows a statistical significant relationship in a particular sample, even though there really isn't a relationship. False negative results occur when a study does not find a statistically significant relationship in a particular sample, even though there really is a relationship.

Paper assignment (Due November 16) -- Write a four-page paper with two sections, approximately two pages each.

The first section should answer one of the following sets of questions, based on the evidence and arguments presented in the readings and class, and your interpretations of this information.

1. Does dietary intake of beta-carotene influence risk of lung cancer?

What evidence suggests that higher beta-carotene intake reduces the risk of lung cancer?

What evidence indicates that higher beta-carotene intake does not have a protective effect?

What interpretations can account for the seemingly contradictory results?

Describe one type of study that you think would be useful to further our understanding of nutritional effects on lung cancer at this time. Explain how this type of study would help to resolve an important unresolved issue.

or

2. Do female sex hormones reduce the risk of coronary heart disease?

What evidence supports the hypothesis that female sex hormones (estrogens and progesterone) reduce the risk of coronary heart disease?

What evidence indicates that female sex hormones do not reduce the risk of coronary heart disease?

What explanations could account for these apparently contradictory findings concerning the effects of female sex hormones on the risk of coronary heart disease?

Describe one type of study that you think would be useful to further our understanding of effects of sex hormones on coronary heart disease risk. Explain how this type of study would help to resolve an important unresolved issue.

or

3. What are the specific causal mechanisms that are hypothesized to link low socioeconomic status to illness and mortality?

What is the evidence for and against each of these hypothesized links?

Describe one type of study that you think would be useful to further our understanding of effects of socioeconomic status on health. Explain how this type of study would help to resolve an important unresolved issue.

The second section of your paper should compare the nature of the evidence related to each of the above three topics. What are the similarities and differences in the types of evidence available for these three sets of questions? What are the relative strengths and weaknesses of the evidence available concerning each of these topics?

Outline for Judith Rodin's section on Body Obsession to follow at a later date