The Behavior Analyst 1991, 14, 229-239 No. 2 (Fall)

A Behavioral Perspective on College TeachingJack Michael

Western Michigan University

Much ofthe subject matter learned in college, particularly during the first two years, is taught in relativelylarge classes with text and lecture as the primary source of the information being learned. If studentsmake effective and prolonged contact with these sources, they can acquire extensive verbal repertoires inmany areas of knowledge. Unfortunately there are many other activities that compete for the typicalstudent's study time. Several factors are often cited as variables that will support study behavior incompetition with other activities, but the only powerful one available to the teacher is the exam grade,as it is related to the course grade. However, unless exams are clearly related to appropriate study behavior,require thorough and extensive coverage of the subject matter, occur frequently (preferably no less oftenthan once a week), and are also clearly related to the course grade, even this factor will not generateprolonged and effective study behavior. Effective college teaching is essentially a form ofaversive control,but if done properly the aversiveness is quite mild, and such aversive control can be responsible for thedevelopment of large and valuable intellectual repertoires.

There are a number ofcollege studentswho learn very little from their courses.Most of these students spend almost notime studying their textbooks, and iftheyeven attend lectures they typically do notspend much time outside of class study-ing lecture notes. It is customary tolocate the problem within the student asa lack ofintellectual ability, a lack ofmo-tivation, or possibly both. An alternativeapproach would inspect the environmen-tal contingencies that are supposed tomaintain class attendance and effectiveout-of-class study. What follows is ananalysis ofthese contingencies and somesuggestions for improving them.

AN IMPORTANT TYPE OFCOLLEGE TEACHING

This analysis is not directed at labo-ratory courses, seminars, skill coursessuch as in fine arts or in sports training,nor courses that are experiential, or aimedprimarily at such goals as personalgrowth, effective interpersonal skills, etc.;

This is an edited version of a paper presented atthe Western Michigan University DistinguishedFaculty Scholar Colloquium on December 12,1989.A monograph covering the same topics in greaterdetail is currently in preparation, and a preliminaryversion can be obtained from the author upon re-quest, Psychology Department, Western MichiganUniversity, Kalamazoo, MI 49008-5052.

nor is it aimed at courses with very smallenrollments (15 or less) irrespective ofcontent or goal. Iam primarily concernedhere with courses having fairly clear in-structional goals related to a specific con-tent or subject matter area, and with aclass enrollment of 40 or more.At universities with ten thousand or

more students, a large proportion ofmoststudents' formal education takes place insuch content courses. In the sciences andhumanities most ofthe introductory andquite a few ofthe upper level courses areofthis type. Such instruction is also quitecommon in business and educationcourses, less so in engineering and finearts.These content courses are often re-

quired as a part of the major or minor,or semi-required in that they are part ofa small set from which the student mustselect one or more. Text material cov-ering most ofthe content is usually avail-able, and the instructional methodologyconsists largely in assigning such text ma-terial, supplementing it with lectures, andassessing the students' learning with someform of in-class examination, typicallygiven more often than twice a semesterbut seldom as often as once a week.

Students have to learn from the textby abstracting out main points, organiz-ing and outlining, practicing solvingproblems, classifying examples, makingup original examples, making up and an-

229

230 JACK MICHAEL

swering test questions, etc. They mustalso learn during lecture and from study-ing notes taken during lecture, which theyreorganize, outline, memorize, etc. justas with the text material.

I know that some professors and somestudents consider such courses to be nec-essary evils, at best, brought on by masseducation, but I don't share that view.This type of course, when it generateseffective study, is responsible for a greatdeal of learning. It is in such courses thatstudents learn much of what is alreadyknown in a field. They acquire extensiveand often fluent repertoires which permitthem to begin talking and thinking aboutthe subject matter in the same way thatprofessional experts talk and think. Thesecourses are sometimes criticized for notteaching creativity or how to obtain newknowledge, but only how to parrot backold knowledge. However, for most sci-entific and professional areas creativityrequires an extensive familiarity withwhat is already known-an extensiveknowledge base about which one can becreative. Creative and innovative use ofa subject matter can be more easily ad-dressed in later courses with smaller en-rollments or in graduate school.

Digression on Differences BetweenCollege and High School

A heavy reliance on texts and lecturesas source material is in sharp contrastwith teaching practices in many highschool courses where class discussion, in-dividual and group projects, homework,and other activities constitute a signifi-cant basis for the learning that takes place.Related to the dependence on texts andlectures is the use of exam scores as themain basis for the course grade in college.High school course grades are often basedon a variety of activities, including at-tendance and class participation, withexam scores sometimes being less thanhalf of the course grade basis. Some en-tering college students who earned ratherhigh grades in high school may be quiteineffective at studying texts and lecturenotes and at taking exams over such ma-terial. It is important for college coun-

selors to be sensitive to this cause ofpoorperformance in seemingly well educatedentering freshmen. Another importantdifference between high school coursesand the type of college content coursedealt with in this paper is the requirementfor many hours per week of outside classstudy, as described below. In college,homework is not just an occasional in-convenience, it is where most of thelearning takes place. It is my impressionthat these three important differences be-tween college and high school are seldomproperly emphasized in freshman ori-entation programs.

STUDENT MOTIVATIONAt many colleges and universities stu-

dents are advised that to obtain a passinggrade-a grade of C-for typical lecturecourses they will be expected to spendabout two hours outside ofclass for everyhour in class. Thus, for a three-credit-hour lecture course they would be ex-pected to spend nine hours each week,three in class and six in outside study. Idon't think our own undergraduate cat-alog makes any such statement, but I con-sider it a very reasonable expectation, andhave always taught my courses accord-ingly. This outside study requirement issometimes referred to as the two-for-onerule and is based on a full semester loadconsisting of 12 to 15 credit hours. Theresult is a student work week of 36 to 45hours, for what could be called mediocreacademic accomplishment. For grades ofB or A the work week may well become50 hours or more, assuming that all ofone's courses are content lecture coursesas described above. In actuality, moststudents do not take 12 to 15 hours ofsuch courses each semester, and somestudents with favorable educational his-tories are able to achieve at a high levelin less time than described above. How-ever, it is not unreasonable for a studentwho wants high course grades to eitherput in more time or take fewer courses.This is clearly not the nine-to-five workweek considered reasonable in some kindsof employment, but the student is notworking for someone else.

COLLEGE TEACHING 231

Now let us ask what might motivatestudents to spend a large portion of theirwaking hours attending classes andstudying academic materials outside ofclasses. The main problem is that thereare many other activities that competefor the student's time. A good case canbe made (to the student, at least) that thecollege setting is an appropriate place toacquire repertoires other than academicones -leadership, organizational, and in-terpersonal skills, for example. For manystudents it is also a rich source of enter-tainment in terms of concerts, plays,movies, social gatherings, and athleticevents as participant or spectator. Andof course television programs may be asreadily available in college as they werewhen the student was in high school.Social relations involving companion-ship, romance, sex, even marriage com-pete with studying for a course. And forthose students who are already married,the demands and pleasures of family liferequire a moderate amount of mainte-nance time. What variables are availableto motivate the study necessary to masterthree to five college course contents dur-ing a semester instead ofengaging in otherattractive or demanding activities? Themotivational variables usually men-tioned can be grouped into several class-es, as follows.

Intrinsic Interest in the Subject MatterBecause of a favorable social and in-

tellectual background or unusually effec-tive lecture or text material, some stu-dents find contact with a particular topicsufficiently rewarding in itself that theyare happy to have the opportunity tospend the necessary time studying. How-ever, I seriously doubt that this factor canplay much of a motivational role. Thereare two main difficulties. Very few stu-dents will find many of their contentcourses of sufficient intrinsic interest tomaintain the necessary amount of studytime. Many topics are somewhat inter-esting to most students, but such interestis usually satisfied with far less contactthan is required by the instructor for evenbarely adequate accomplishment. In ad-

dition this factor is unfortunately suscep-tible to postponement as contrasted withthe competing activities, where seize themoment is the relevant attitude. Thespontaneous social gathering, the unex-pected dating opportunity, scheduledathletic events, all must be taken advan-tage ofat the time they become available,but the aspects of the text that make itintrinsically interesting will be availableand just as interesting at a later time.

Approval and DisapprovalfromSignificant Others in the Study SettingThe reactions of parents play a signif-

icant role in motivating the study behav-ior ofsome high school students, but thisinfluence is greatly weakened when thestudent moves away from home. Evenfor those still living at home, school workbecomes less subject to effective scrutinyand evaluation by parents as a result ofthe new topics being studied and the par-ents' inability to accurately understandthe contingencies affecting the student atcollege. Roommates might provide ap-propriate approval and disapproval butare not likely to react in such a way as tosubject themselves to any form of aver-sive countercontrol. Thus not much canbe expected from this factor, and in anycase it is not one that the instructor hasany control over.

Social Reactions ofOthers to theRepertoires Acquired Through Study

Praise and admiration for displayingone's newly acquired knowledge may beavailable from the instructor in class,from other students in the same class, orfrom other people in the academic ornonacademic environment. The instruc-tor has some control over this variablewithin the class situation but not muchwhen there are 40 or more students inthe class. The approval of others, in orout of class, is not under instructor con-trol at all, and for most content coursessuch approval is pretty skimpy and nottightly related to mastery of the assignedmaterial. It is hard to imagine a studentgiving up an attractive social activity oran interesting entertainment for fear of

232 JACK MICHAEL

having to temporarily forego this type ofapproval.

Short-Term Advantages to Oneselffromthe Newly Acquired RepertoireSome things learned in college courses

have relatively immediate practical val-ue-as when an automotive engineeringcourse facilitates repair of one's own car.Or the things learned, as in an economicscourse, may permit a better understand-ing of a puzzling political or economicsituation. Many of the content courses,however important for their contributionto further learning in the same field, pro-vide relatively little of this kind of short-term payoff. Even when they do, the pay-off is not likely to support much sacrificeof attractive competing activities, and inany case is not closely related to the de-tails of the study assignment nor undermuch control by the instructor.

Long-Range PayoffsGetting a degree and a good job, going

to graduate school, contributing to theimprovement of the human condition,etc. are events that are sometimes citedas possible sources of motivation forstudying. Future possibilities of this sortmay well have played a role in students'and parents' decisions regarding collegeattendance, and the general importanceof such events would seem to contributeto their strength as possible motivators.However, their temporal remotenessworks in the opposite direction. Evenmore critical is the impossibility of re-lating such events to the details ofa week-ly study assignment -it is easy to believethat one can contribute to the humancondition even if one doesn't really un-derstand a particular graph in the text-book. Also these long-range payoffs sufferfrom the same susceptibility to post-ponement as intrinsic interest-one canprobably get almost as good a job ongraduating if instead of studying tonightone takes advantage of an unusual socialopportunity and studies tomorrow night.Instructors should certainly direct stu-dents' attention to such future possibil-ities when appropriate, but my experi-

ence doesn't suggest that this type ofmotivator plays any appreciable role inmaintaining daily and weekly study.

The Course GradeNow we come to the one motivational

factor over which the instructor has con-siderable control, and which is easily re-lated to the details of the study assign-ment. It is also a factor of considerablestrength, as evidenced by the intensity ofstudy activity occurring immediately be-fore a major exam. Some instructors willassert that grades do not actually serveas effective motivators for many of thecurrent crop of college students, and thisissue will be addressed later. But first afew comments about university gradingpractices.

UNIVERSITY GRADINGPRACTICES

A university is a place where studentsacquire repertoires that are of value tothem and to society. It is generally con-sidered to be the responsibility ofthe uni-versity to assess the extent and quality ofthe student's accomplishment and toprovide a record ofthis accomplishment,the academic transcript. This record be-comes important when the student at-tempts to qualify for a further education-al opportunity or for employment. Insome cases the agency or institution of-fering the further opportunity will per-form some assessment-give a speciallyprepared examination or conduct an in-terview-but most of the assessment isconsidered best done by those who taughtthe student. (This involves some conflictofinterest, ofcourse, since the educator'sassessment that most of the studentsfailed to accomplish anything would cer-tainly raise questions about the adequacyof the university.) It is certainly possiblefor a person to acquire a professional,intellectual, or technical repertoire with-out the aid of an educational institution,but most graduate schools and employersdo not readily accept the task ofassessingthe extent or quality of such self-maderepertoires.Course grades are assigned according

COLLEGE TEACHING 233

to two quite different principles, referredto as norm-referenced and criterion-ref-erenced grading. A grade that reports thestudent's standing in the course relativeto the accomplishments of the other stu-dents is norm referenced. An instructoruses norm-referenced grading when giv-ing the grade ofA to those students whoreceived the top ten percent of the examscores. Criterion-referenced grading in-volves grade assignment according to anabsolute standard, irrespective ofthe per-formance of other students. Giving thegrade ofA to those students who obtain90 percent of the available exam pointsis criterion-referenced grading. In prac-tice these two grading principles are sel-dom seen in a pure form. An instructorusing a norm-referenced approach maystill refuse to give the top ten percent anA if the overall class performance seemsbelow the usual level experienced by thatinstructor, or may exceed the ten percentA assignment ifthe class seems unusuallyeffective. Likewise, an instructor using acriterion-referenced approach may ad-just the absolute criterion downward if itseems that an exam was inappropriatelydifficult when other evidence impliesgood effort and accomplishment by thestudents.At this point it will be useful to con-

sider the often discussed notion that anemphasis on grades fosters an undesir-able form of competition among stu-dents. It is important, first, to distinguishbetween two kinds ofcompetition whichcan be called vicious competition andfriendly competition. Competition is vi-cious when an increase in one student'scourse grade will result in a decrease inanother's; it is friendly when anyone whoachieves at a certain accomplishmentlevel will receive the grade appropriatefor that level. It is quite clear that norm-referenced grading -grading on a curve-is a vicious competition contingency, be-cause if only the top ten percent of theclass receives an A, students improvingtheir performance and moving into thatcategory bump other students out. Cri-terion-referenced grading is a friendlycompetition contingency because eachgrade is assigned to anyone who achieves

at that level (gets a certain number ofpoints related to exams, papers, etc.), ir-respective of how many other studentsachieve at that level. With criterion-ref-erenced grading there is, of course, thecompetition related to having the highestpoint score, or having a higher score thansome other particular student, but this isnot the type of competition that makesstudents rejoice in the misfortune ofoth-er students or refuse to help others forfear of lowering their own course grade.For this reason as well as others, I believethat criterion-referenced grading is theonly appropriate form ofgrading for col-lege academic performance.To return to grading practices, various

scales are used with the most commonbeing a five level A, B, C, D, and E sys-tem, or an eight level scale of A, BA, B,CB, etc. (sometimes recorded as A, B+,B, C+, etc.) In general the larger the num-ber of steps in the scale the more precisethe information being reported, up to thepoint where the precision becomes un-reasonable given the relative crudenessofthe measurement system. I suspect thatan eight level scale is about as precise asis appropriate for most classroom grad-ing practices. When considered as a mo-tivational variable, an eight level scale issuperior to a five level scale near the endof the term. The value of further studyincreases when it leads to an advance tothe next letter-grade, and this remainsavailable longer with an eight level scale.

THE PROCRASTINATIONSCALLOP

Studying for a college course exam islike many tasks where a certain amountofwork must be completed before a rel-atively inflexible deadline, and where thequantity and quality of the work com-pleted determines the value ofsome sub-sequent event-in this case, the gradeearned on the exam. And as with manysuch tasks, the work could be done at anytime during the interval between the an-nouncement ofthe task and the deadlinebut typically does not occur until nearthe end of the interval and then with in-creasing frequency. This pattern of be-

234 JACK MICHAEL

all Studying for an Exam very veryfiished good good

start ) no time left~~o

very ~~very

sm ~~~~~~~~~~~~~~~notime lefttime left to complete the task

Figure 1. The Procrastination Scallop.

havior is sometimes erroneously referredto by behavioral psychologists as aJixed-interval scallop, implying the same func-tional relations between behavior, time,and reinforcement as seen with fixed-in-terval schedules of reinforcement. Thepattern is similar, but the controlling re-lations are quite different.For the students in my courses the task

consists of studying a text and lecturenotes in conjunction with study objec-tives, preparing further study materials(flash cards, notes ofvarious kinds, etc.),and rehearsing or working with these ma-terials until they are well learned. Thediagram above portrays this situation asa task-completion/time-passage space. Atthe beginning of the study interval -im-mediately after finishing the last exam-the student's situation is represented bythe bottom left corner. Working towardcompleting the task moves up; time pas-sage moves to the right. At the end oftheinterval (at the time of actually writingthe exam) the vertical location (how muchof the task was completed) determinesthe value ofthe outcome (the exam score).The importance of the exam score, inturn, is largely a function of its relationto the overall course grade given at theend of the semester.Task completion and time passage to-

gether determine the aversiveness of thesituation at any point in the interval rel-evant to the task. To have completed mostof the task and to have a good deal oftime left before the end of the interval issafe or nonaversive. But to have com-pleted very little of the task and to havevery little time left is a condition that isdangerous or aversive, and which gen-erates escape behavior. Usually the onlyrelevant form of escape under these cir-cumstances is to work on the task, thatis, to study for the exam. The usual re-inforcement for such studying is a changein one's status to a less dangerous con-dition with regard to the task-time re-lation. A number of details are relevantto the dynamics ofthis environment-be-havior relation and these will now beconsidered.

Competing ActivitiesAs mentioned above, we can usually

assume that a number of opportunitiesor demands regarding activities otherthan studying will be superimposed onthis situation. I will also assume that ifstudying has any intrinsic reinforcementvalue, it does not compete effectively withthe reinforcing value of the other behav-iors, either because of its relative weak-

COLLEGE TEACHING 235

ness, or because the other opportunitiesand demands are much less postponable.These assumptions mean that studyingthe subject matter of the college coursewill not ordinarily occur to a sufficientdegree to develop the mastery requiredby the instructor unless the studying isrelevant to course credit and to the coursegrade.

Course Grade ImportanceSecondly, ifthe course grade is oflittle

importance to the student, then the pos-sibility ofreceiving a low exam score willcertainly not function as a basis for aver-sive control, and studying as a form ofescape behavior will certainly not com-pete with behavior related to othersources of reinforcement.

The Relation Between Exam Gradeand Course GradeEven if the course grade is important

to the student, any one exam grade doesnot determine the course grade by itselfbut only in combination with the gradeson the other exams and/or other assignedcourse activities. This is the implicationofthe diverging arrows on Figure 1 lead-ing from the exam outcome to the courseoutcome. The exam outcome influencesthe course outcome but doesn't com-pletely determine it, a factor responsiblefor an unavoidable reduction in the ef-fectiveness ofa low exam grade as a basisfor aversive control. The student can be-lieve, and not incorrectly, that a low scoreon the exam being studied for can pos-sibly be compensated for by a high scoreon a later exam. This source ofvaguenessis enhanced when a large portion of thecourse grade depends on some end-of-semester activity such as a comprehen-sive final exam or a term paper. How-ever, in many college courses there is agood deal of additional vagueness in therelation between exam outcome andcourse outcome that results from the in-structor's general failure to preciselyspecify the basis for the course grade. An-other source of vagueness is the use of anorm-referenced or relativistic grading

system where the exam and course gradesare not only based on the student's ownperformance but also depend to some ex-tent on the performance ofother studentsin the course.

The Relation Between Studyingand the Exam Grade

In order for the potential exam gradeto motivate completing the task beforethe time runs out, the relation of taskcompletion to exam outcome must be astrong one. Iffailure to complete a sizableportion of the task does not result in abad grade, or task completion does notresult in a good grade, then studying asescape behavior will not be strong, norwill task progress function as a form ofreward for the various aspects of studybehavior. The most common difficultyhere is the instructor's failure to be suf-ficiently clear about the relation of textand lecture material to the exam con-tent-usually out offear that with a morespecific assignment the student wouldlearn what was specified but would notlearn all the other things that would haveto be studied if the exam content couldnot be anticipated. Such specification mayalso be withheld on the grounds that partof the scholar's repertoire consists ofbringing order out of chaos, and havingto deal with the instructor's vague as-signments is a good opportunity to ac-quire this skill. Students are often heardto say that in some cases they studiedextensively and received a low grade andin others they studied hardly at all andreceived a high grade. They will say thatthey have no idea where the exam ques-tions came from, or that topics that werebarely touched in lecture and hardlymentioned in the text constituted a sig-nificant portion of the exam. Of course,some ofthis complaint is rationalizationfor poor performance due simply to in-sufficient or ineffective study, but as anex-student I find myself quite sympa-thetic to such complaints. In some casesthe only thing that saved me was the in-structor's strict adherence to a norm-ref-erenced grading system where no onelearned what appeared on the exams, but

236 JACK MICHAEL

those who did least worst received highgrades.

In my courses, and in those ofsome ofmy colleagues, students are provided withweekly study objectives -statementsidentifying essential parts of text and lec-ture material. These function to definethe study task, making it possible for thestudent to know at any point what re-mains to be studied, and when to quitstudying because the task has been ac-complished. Some study objectives iden-tify relatively explicit definitions, facts,principles, etc. but quite a few are relatedto more complex kinds of knowledge,such as being able to apply some conceptor principle to a novel situation, identifyor classify examples, generate original ex-amples, explain the point being made bya text example, etc.

Digression on Lecture Content andClass AttendanceFor large-enrollment content courses,

lecture attendance is closely related tohow essential the lecture material is fordoing well on the exam, not as is widelybelieved, how interesting or inspirationalthe lectures are. Of course, other thingsbeing equal, more interesting lecturerscommand better attendance than thosewho are less interesting, but it is easy tooveremphasize this factor. Because thereare usually pretty good textbooks for suchcourses, the basic problem for many in-structors is how the lecture materialshould be related to that presented in thetext. One popular, but I believe inappro-priate, approach -based on the beliefthatmany of the students will not study thetext, or if they do will not understand it-is to provide in lecture essentially thesame information that is in the assignedtext readings. This renders the lecturesunnecessary for students who can under-stand the text, and the text unnecessaryfor those who attend and make good useof the lectures, and it greatly reduces theamount ofmaterial that can be adequate-ly covered in the course. Much better isto provide detailed study objectives whichguide the students' contact with the text-indicating what is important from the in-

structor's perspective -and which con-tain instructor-prepared supplementarymaterial that clarifies and expands on in-adequate or especially difficult sectionsin the text. With such an approach thelectures can then provide more recentsubject matter developments, elaboratefurther on especially difficult text mate-rial, add material that the instructor be-lieves is important to a thorough under-standing of the subject matter but whichis not covered at all in the text, and soon.However, it is not reasonable for a very

large portion of the course content to beavailable only as lecture material. Ade-quate text material can be found for mostspecific-content courses at the freshmanor sophomore level, and too heavy reli-ance on lecture results in a reduced con-tent and is too dependent on students'note taking skills. If essential material isnot available in text form then the in-structor should write it. It should, ofcourse, be clear by announcement and bypractice that much that is in the text willnot be covered in lecture but will nev-ertheless be covered on exams. (The worldof scholarship depends on the writtenword. College students must be effectivereaders and studiers when they come tocollege, or they must quickly acquire theseskills, or they should not be able to obtaina college education. If they can, the col-lege education doesn't amount to much.)It should also be clear, again by practiceas well as by announcement, that any ma-terial presented in lecture that is not inthe text will be on the examinations. Lec-ture attendance under such conditions isgenerally quite satisfactory, and not forartificial reasons.

Task and Time MonitoringThe student's task monitoring and time

monitoring, ofcourse, also determine theaversiveness ofthe situation and thus thefrequency of the relevant escape behav-ior. Ifthe task is vaguely described by theinstructor, or vaguely understood by thestudent, the current situation will notfunction as an accurate source ofaversivecontrol, with the usual error being that

COLLEGE TEACHING 237

the student underestimates the amountofwork left to be completed. Similarly ifthe student's time assessment is inaccu-rate-the usual error being to overesti-mate available study time-the situationat any point in the assign-study-test cyclewill not be as aversive as it should be.Even with accurate monitoring many ofus have a significant tendency in the di-rection ofunrealistic optimism, resultingin our being slightly to moderately latein completing most tasks. Weakness inthese skills is of importance, however,only at the beginning of exposure to awell managed motivational system: Thesystem, itself, will automatically developadequate task and time monitoring afterexposure to several assign-study-test cy-cles.To summarize, the threat of receiving

a lower grade than is acceptable will usu-ally generate appropriate exam study be-havior, but not if any of the followingconditions prevail: (a) the course gradeis unimportant to the student, (b) the re-lation of exam grade to course grade isunclear, or (c) the relation ofstudying forthe exam to the exam grade earned isunclear.Now back to the claim, often heard,

that many students are simply not mo-tivated by grades. Such a claim is un-doubtedly based on instructors' experi-ences, but before taking this claim tooseriously we should first eliminate evi-dence that is confounded with the secondand third deficiencies described above.My experience has been that very fewstudents are willing to take a low gradewhen the contingencies are properly ar-ranged.

Students do poorly in college for twomain reasons: inadequate intellectualrepertoires and ineffective motivationalcontingencies. The present analysis dealsonly with motivation. Even when thedetrimental conditions listed above arenot present, students with grossly defec-tive intellectual skills will not be easilysuccessful. However an effective moti-vational system can go a long way towardsupporting the automatic developmentof better intellectual skills. The same istrue with respect to self-management-

control by long-range contingencies asopposed to excessive influence by thepleasures, and pains ofthe moment. Self-management skills develop as a result ofpersonal experience and instruction byothers, but this development depends tosome extent on a relatively stable envi-ronment. Most students already havesome tendency to study rather than toplay when it is appropriate, and if doingso generally has good results, and if fail-ing to do so generally has bad results, self-management skills automatically im-prove. Repeated exposure to an effectiveand reasonable assign-study-test cycle canplay an important role in generating andmaintaining effective student behaviorand would seem especially important atthe freshman and sophomore level.

Before drawing some general conclu-sions it is necessary to comment on examcoverage and exam frequency. For mo-tivational purposes any student whomasters only a small proportion of theassigned material should systematicallyand inevitably fail the exam, and stu-dents who master most of it should al-ways receive high exam grades. Examcoverage insufficient to accomplish thesegoals greatly weakens the effectiveness ofthe exam as a motivational basis forstudying rather than doing other things.Ifstudents correctly believe that they canskip studying any ofthe assigned materialand still get an acceptable grade, manywill do so. It may be possible to showthat an exam which covers only a mod-erate sample ofthe assigned material willnevertheless function as an adequate as-sessment device, but this is irrelevant tothe motivational issue. For motivationalpurposes the student must correctly be-lieve that failure to complete even a smallportion of the assigned study task willinevitably lead to an appropriate lower-ing of the exam grade and ultimately thecourse grade.With respect to exam frequency, my

goal is to generate six or more hours ofeffective out-of-class study every week.When I give an exam every week, andwhen the exam covers most of the as-signment, such study does take place. Ihave been teaching this way for about 20

238 JACK MICHAEL

years and there is no question in my mindthat it produces the necessary study bymost of the students. What if I gave anexam every three weeks? I use one-houressay exams and it is quite clear that if Igave such an exam over three times asmuch material it would have to be a muchsmaller sample of the material assigned.This would result in a less accurate mea-sure ofaccomplishment, and would con-stitute an instigation to omit some partof the assignment because of the distinctpossibility that it would not appear onthe exam. But more important, if it takessix or more hours to master a week's as-signment, then it should certainly take18 or more hours of out-of-class study tomaster a three-week assignment. The na-ture of the procrastination scallop de-scribed above strongly suggests, and ex-perience verifies, that sufficient fear togenerate study in competition with otheractivities will only be present during thelast week of a three-week period, and itis absurd to suppose that students willsystematically put in 18 hours studyingfor a single course during any one week.What happens is that instructors lowertheir standards so as not to fail too largea proportion ofthe students-they do notexpect 18 hours of outside study for anexam over a three-week assignment. Evenless do they expect 30 or more hours ofstudy for an exam given after five weeksof instruction with no previous exams.Under such conditions there develops astrong tendency to favor norm-refer-enced grading and give high grades to thebest students, even though they didn'tlearn very much. How about examiningwith only a midterm and a final?

Since exams are seen as the most aver-sive aspect ofcollege learning, it has beensuggested that a system involving weeklyextensive-coverage exams would behighly unpopular with students, but I havenot found this to be the case. The idea isunpopular, but on course evaluationsmost students strongly praise the system,correctly identifying the weekly exams asbeing responsible for generating exten-sive and well-paced study. Students whocan be identified as highly effectivethose who are earning grades of A and

BA- seem especially satisfied with sucha system, saying that they very much ap-preciate knowing exactly what they haveto study (as a result ofhaving the detailedweekly study objectives), being able tothoroughly master small units of work,and being induced to spend much moretime studying than with many of theirother courses.

CONCLUSIONS1. It is unrealistic to consider sources

ofmotivation other than exam and coursegrades to be ofmuch importance in gen-erating study in the typical content course.

2. Exam grades, course grades, andwhat they represent can function to mo-tivate a more than adequate commit-ment to out-of-class study if three im-portant conditions are met: (a) the coursegrade is important to the student, (b) examgrades are closely related to course grades,and (c) studying is closely related to examgrades. (The instructor has little or nocontrol over the first, but is completelyresponsible for the other two.)

3. To meet these conditions: (a) examsmust be extensive and must occur quiteoften; (b) assignments must be quite spe-cific as to the relation between text andlecture material and exam content; and(c) weekly exams must contribute in alarge way to the course grade, that is, finalexams, term papers, etc. cannot contrib-ute too much to the course grade.

4. There are three popular notionsabout college teaching that I believe arequite incorrect for the type of course Ihave been considering, and when takenseriously, impede the development of ef-fective instructional systems. The first isthat if you teach properly the studentswill find learning both fun and easy. Nottrue, for two reasons: intensity and as-sessment. Learning new things is oftenenjoyable but seldom so under condi-tions where a great deal must be learnedin a relatively short time. The college stu-dent taking several content courses willhave to spend over 40 hours a week at-tending classes and studying outside ofclass. But as described above such a stu-dent is also physically, socially, and eco-

COLLEGE TEACHING 239

nomically prepared to participate in agreat many other available activities ofakind that require little or no effort andare either intrinsically interesting orachieve ample extrinsic reward. Studentswill often find themselves saying some-thing like "I can't do it now because Ihave to study." Turning down many at-tractive opportunities in favor of, at best,moderately interesting textbook and lec-ture study is essentially aversive. And inaddition to the intensity, college learningis the kind of activity that is not takenfor granted, but rather is formally as-sessed. Spending time and effort tryingto accomplish something with the dis-tinct possibility that when it is assessedthe result will be disappointing is also anintrinsically aversive situation. The aver-siveness is, of course, tolerated becauseofthe long-range value ofthe repertoiresbeing acquired, but the value of the re-mote end doesn't completely eliminatethe aversiveness of the more immediatemeans to its achievement.The second false notion is that grades

should not be emphasized. Not true!Grades are the primary motivative vari-able, which means that current standingin the course and progress toward a finalgrade should always be clear and fre-quently brought to the student's atten-tion. With the current ready availabilityofmicrocomputers and spread sheet pro-grams an instructor can easily providestudents with a weekly printout showingtheir past and present performance alongwith that ofthe other students, providinga projection to their final course grade,and indicating how they must perform

on remaining course activities to achieveany particular course grade. I provide thistype of weekly information, and it is al-ways well appreciated. On course eval-uations many students write in that theywish all their instructors gave them suchinformation.The third false notion is that good

teaching consists primarily of good lec-turing. Of course it is important for lec-ture material to be interesting and valu-able for learning the course subject matter,but lecture presentation is only a smallpart ofthe total activity. It is much moreimportant for the instructor to have aneffective instructional system involvingclear study objectives over both text andlectures, frequent well-designed examsand an exam environment that rendersany form of cheating relatively useless,appropriate remedial opportunities forstudents who experience occasional poorperformances, and so on.

5. Effective instruction, as impliedabove, is labor intensive. The notion thata content-course system that generates ef-fective intellectual repertoires in 40 ormore students can be managed and con-tinuously updated by spending little morethan the time required to present threeone-hour lectures a week is way off themark.

6. Perhaps in a more general sense itwould be useful to conclude as follows:The world runs on fear. College learningis largely under aversive control, and itis our task to make such control effective,in which case it becomes a form ofgentlepersuasion.