A Minimax Principle for Educational Research

Author(s): Cameron Fincher

Source: Research in Higher Education, Vol. 2, No. 2 (1974), pp. 99-107
Published by: Springer
Stable URL: http://www.jstor.org/stable/40194720
Accessed: 05/06/2010 06:46

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=springer.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Springer is collaborating with JSTOR to digitize, preserve and extend access to Research in Higher Education.

http://www.jstor.org
 RESEARCH IN HIGHER EDUCATION, Vol. 2
© 1974 APS Publications, Inc.

A MINIMAX PRINCIPLE FOR

EDUCATIONAL RESEARCH

Cameron Fincher, Institute of Higher Education, University of Georgia,

Athens, Georgia

Educational research has lost considerable prestige in recent years. The optimism of
the 1960's has not been fulfilled and educational research has been relegated to a
defensive position. If the findings of educational research are to be more intuitively
appealing, readily understood, and widely accepted, educational researchers should
give better attention to the reduction of needless methodological complexity and to
the improvement of conceptual clarity. To focus attention on this necessity, a
minimax principle is suggested in which the educational researcher should seek to
minimize the maximum methodological complexity that is necessary to produce a
satisfactory degree of conceptual clarity

There is a pessimism about educational research that requires attention. In the

brief span of a few years educational research has descended from a position of
national prominence and commendable optimism to a lowly state of confusion and
genuine skepticism. An adequate interpretation of that rise and decline is yet to
be made and may require some of our best analytic thinking for several years
to come.
The loss of status for educational research is signified by several reports issued
from 1968 to 1972. The Committee for Economic Development (CED, 1968)
was one of the first critics to become disenchanted with the directions being
taken by educational research during the 1960's. Entering a strong plea for
innovation through technological means, the CED designated both basic and
applied research as an imperative but stressed the need for dissemination and
practical application. In 1969 the National Academy of Education (Cronbach
and Suppes, 1969) issued a call for innovative practice and emphasized the need
for "research that will point out productive paths to follow" (p. xv). The thrust
99
100 Fin cher

of the NAE reportwas that scholarlyinquiryinto the problemsand issuesof

education requiredmore than the traditionalmeansof quantitative,empirical
research.Not only should historical,humanisticmodes of inquirybe en-
couraged,but a cogent distinctionshould be made between conclusion-oriented
researchand decision-orientedresearch.The distinctionhelped emphasizethe
need to approacheducationalresearchin termsof its outcomes as well as its
originsand procedures.
The most ambitiouscritiqueof educationalresearchwas undertakenby the
staff of the NationalCenterfor EducationalResearchand Development(NCERD,
1969). The reportis one of the more commendablepublicationsreleasedby a
federalagency on the subject of education in recent years. It dealswith educa-
tional researchand developmentin unusuallybroad terms and documentsin
good fashion many of the conceptualshifts that have taken place in the last de-
cade or so. Its majorconclusionis that educationalresearchand development
has not been governedby an overallstrategythat would permitconsistent ob-
jectives and effective procedures.Not only has financialsupportbeen inadequate,
but researchmanpowerand informationalneeds have not been met. If research
can providea basis for the improvementof instructionand if such provisioncon-
stitutes a high priority,then an overallstrategyfor the supportof researchand
developmentshould be forthcoming.Educationalresearchmust be improvedin
support,management,and impact.
A more pessimisticview of educationalresearchhas been presentedby staff
membersat the RAND Corporation(Averch,et al., 1972). Although the report
does not deal directly with educationalresearchand development,it contains
numerousimplicationsfor the assessmentof educationalresearchin general.The
reportis explicit in pointing to the limitationsof availableresearchand the
difficultiesof measuringthose aspectsof educationwith which researchis con-
cerned. The majorimplicationof the RAND study is that despite the many fine
efforts of educationalresearch,it does not providea foundationfor the formula-
tion of educationalpolicy.
The extent to which the prevailingpessimismreflects a negativisticattitude
towardeducationalresearchshould be carefullyassessed.There is no doubt that
severalstrong conceptualshifts have taken place in recentyears and that educa-
tional research,as it is traditionallytaughtin our graduateschools, is on the
defensive.The situation is not unlike that describedby Campbelland Stanley
(1963) in discussingthe reactionto experimentationin education duringthe
1920's.

When,in fact, experimentsoften provedto be tedious, equivocal,of unde-

pendablereplicability,and to confirmprescientificwisdom, the overoptimistic
groundsupon which experimentationhas beenjustified were undercut,and a
disillusionedrejectionor neglect took place (p. 3).
 A Minimax Principle for Edu cational R esearch 101

As much as we might agreewith Campbelland Stanley that experimentationis

"the only means for settling disputesregardingeducationalpractice,... of
verifyingeducationalimprovements. . . and establishinga cumulativetraditionin
which improvementscan be introducedwithout the dangerof a faddishdiscard
of old wisdom in favorof inferiornovelties,"it does not follow that experimental
researchis currentlyheld in high esteem by those concernedwith the formula-
tion of public policy in education.
The EducationalAmendmentsAct of 1972 endorses- but does not fund -
educationalresearchas it has not been endorsedpreviously.In establishingthe
NationalInstitute of Education,the legislationstates unequivocallythat it shall
be a policy of the nation to providean equal opportunity for an education of
high quality;that the achievementof that quality will requiremore knowledge
about learningand education than we now have or can expect from presentre-
searchand experimentation;and that the federalgovernmenthas a clear
responsibilityto provideleadershipin the conduct and supportof scientific
inquiryinto the educationalprocess.
Whetherthe National Institute of Educationwill fulfill such a role remainsto
be seen (Glennan, 1973). Whatis not in doubt, however,is Congressionalintent
that the federalgovernmentshould supporteducationalresearchin ways that
were not met by programsand projectsof the 1960's. Congressionalmotives may
or may not be based on disappointmentwith regionallaboratories,researchand
developmentcenters, and instructionalmaterialscenters.Whatis clear is that
Congresshas sought other instrumentalitiesto bringabout a more effective ap-
plication of researchfindingsto the problemsof education.
The reactionof Congressreflects a largershift of attention and emphasisfrom
the creationof knowledgeto its utilization in the solution of societal problems
and issues. Basicresearchhas not been denied by the shift as much as applied,
mission-oriented,problem-solvingformsof researchare being emphasized.A
nationalsense of frustrationand failurehas followed the numerousefforts to
achievethe social, economic, and political goals formingmuch of the revolution
of risingexpectations.Public education, and the school as a social institution,
was a focal concernin the nationalexpectation that minority and disadvantaged
groupscould be fully assimilatedinto the Americanmainstream.The success of
the space programwas a continuous reminderof the failureof researchand de-
velopmentin other areaswith the resultthat the failureof researchand develop-
ment in education may be acutely disappointing.

METHODOLOGICAL IMPERIALISM

The challenge to educational research is to reclaim its authority as a valid

foundation for educational policy - and to reestablish its credibility. There is an
102 Fincher

urgent need to prove that educational research is relevant to the teaching - learn-
ing process, that it does have beneficial effects, and that it deserves to be fostered
and supported in the manner implied by the Educational Amendments Act of
1972. To survive in the current intellectual climate it is not sufficient for educa-
tional research to be valid; the theory, methods, and empirical findings of
educational research must be both effective and creditable.
The strong, continuing shift reflected in the CED, NAE, NCERD, and RAND
publications places a considerable premium on importance, relevance, or ap-
plicational value. This shift involves more than a changing emphasis on develop-
mental research as opposed to experimental research. There is indeed an in-
tensive concern at the present with development, diffusion, and dissemination
models that would convey research findings in terms meaningful to the educa-
tional practitioners, but the substitution of models alone will not satisfy the
demands that are now being placed upon educational research. In addition to
changing tactics in research funding, it will be necessary to alter the perspective
with which educational research is conceptualized and implemented.
It is the contention of this paper that methodological dominance is
responsible for much of the current disenchantment with educational research.
The failure of educational researchers to deal more directly with substantive issues
in education accounts not only for the loss of credibility but affects both the
validity and utility of educational research as it is presently produced, published,
and consumed.
The criticism that too many studies are methodologically tight but sub-
stantively weak will not be regarded as novel. The cirticism has been voiced at
various intervals during the past twenty-five years but the criticism has not
succeeded in having an advantageous effect on the quality, scope, and content of
research as published in many research journals. Graduate programs in educa-
tional research are still suspected of too heavy an emphasis on skillls, techniques,
and procedures as opposed to concepts, values, logical analysis, theoretical de-
velopment, and factual content. The criticism does not contend that the field is
all technique and no substance, but it does imply a failure of too many graduate
programs to relate courses of substance back to skill courses. The techniques
learned in one course are not readily applied in absorbing the content of others.
The NAE Committee on Educational Research has estimated that "not more
than one-tenth of the doctoral dissertations in Education and not more than one-
tenth of the work published in the less well-edited journals ... are respectable
works of serious inquiry" (p. 226). Their is the pessimistic assessment that "much
of what is now published is of modest significance and quality" (p. 232). Yet,
research findings are said to be "hailed and re-echoed" without critical examina-
tion of either the data or the logic supposedly supporting them and with the
outcome that there is a prevalence of self-fulfilling hypotheses. Whatever the
causes of such a condition in educational research, it is unlikely that a remedy
 A MinimaxPrinciplefor EducationalResearch 103

will be found in simply shifting the emphasisfrom basic or fundamentalresearch

to appliedor developmentalstudies.

A MINIMAX PRINCIPLE

If educationalresearchersare to meet the changingdemandsthat are

currentlybeing placedupon them, it would appearthat some modification of
their professionalactivitiesis in order.The needed modificationswould not call
for indepth soul-searchingas much as a better understandingof the purposeand
functions of educationalresearchin the intellectualclimate that currentlypre-
vails.If in the past, educationalresearchershave erredtoo greatlyin emulating
the physicalor biologicalsciences in their researchefforts, there would seem to
be little reasonto continue doing so. The challengewould appeardefinitely to
the contrary.Moreimportant,it would appearthat there is a clearchallengeto
educationalresearchto produceknowledgein a form that is clearlyapplicable
to the problems of education.
As one step in the direction of more usableknowledge,a suggestionmay be
made in the form of a minimaxprinciplewhich educationalresearcherswould
do well to apply. It is suggestedthat:
Educationalresearchbe conducted in such a manneras to minimizethe
maximummethodologicalcomplexity necessaryto produce a satisfactory
degreeof conceptualclarity.
For those who prefera maximinprinciple,the same thought can be suggested
by the statement that educationalresearchshould maximize the minimumcon-
ceptual clarity that is permissiblefor the methodologicalcomplexity that
is necessary.
Regardlessof how it is expressed,the principleis one that stronglyadvocates
better attention to the conceptualaspectsof educationalresearch. Thereis no
intent to make methodology a whippingboy or scapegoatfor the currentdis-
enchantmentwith educationalresearch.There is simply a contention that
needlessmethodologicalcomplexity has contributedto the failureof educational
researchersto communicatetheir researchfindingsin a form or fashion that con-
veyed the gist or substancein termsreadilyunderstoodby practitionersin the
classroom.Even the most casualsurveyof educationalresearchjournalswill
turn up numerousstudies that are methodologicallydominantwith little or no
substantivecontent, immediacy,or relevance. Like the proverbialdoctoraldis-
sertation,the majorconclusion is that furtherinvestigationis needed and that
someone other than the authorought to do it.
Nor should the principlebe interpretedas antimethodology.It is merely a
protest againstresearchthat turnsout to be nothing other than a methodological
104 Fincher

tour de force. Certaindoctoral dissertationsmay well be defended on the grounds

that the doctoralstudentsneed a vehicle to demonstratemethodologicalcom-
petence or accomplishment.Such dissertationsmake no pretenseto an original
contributionbut are designedto demonstrateprofessionalcompetence with the
tools of the trade.The absenceof substance,relevance,or applicabilityin such
studies need not be seriouslyregretted.The purposeof such dissertationsshould
be explicit, however, and their publicationshould be discourageduntil such
time that a higherorderor relevanceor importanceis more probable.
Nothing in the principledenies the need for methodologicalstudies as such.
The continued, systematicinvestigationof method is clearlyneeded, and every
encouragementshould be given to studies that will contributenew insightsto
method itself or develop new techniquesfor the actual conduct of research.
The purposeof such studies, however,is usually designatedas such and we need
protest only those studies that seek a coat-taileffect of providingtrivialsub-
stance with a greatdisplayof methodologicalfireworks.Neither shotgunsnor
heavy artilleryshould be denied as a suitableinclusionin the educationalre-
searcher'sarmory.A strong case could be made, however,for the inclusionof
flyswattersthat could be used with properdexterity.
Educationalresearchershave indeed been guilty of using cannonson house-
hold pests. This tendency seems to have been intensifiedby the availabilityof
packagedcomputerprogramsand other forms of cannedanalyticprocedures.
Yet, neitherthe advancesin factor analytic,multivariate,and covariancetech-
niques nor the Bayesianinfluence would suggestthe complete demise of Chi-
square,t-tests, or Pearsonr derivativesas useful analyticmethods. A law of
parsimonythat is often appliedto the hypothesis-makingphasesof research
may be ignoredtoo frequentlyin the applicationof methodology.

CONCEPTUAL CLARITY VS. METHODOLOGICAL COMPLEXITY

Applicationof a minimaxprincipleto educationalresearchwill necessarily

requirethat the relationshipbetween conceptualclarity and methodological
complexity be better understood.To gain that understanding,it may be helpful
to postulate a curvilinearrelationshipbetween the two. This relationshipwould
imply that the conceptualclarity of a researcheffort should be improvedwith
an increasein methodologicalcomplexity. More variablescould be handledin
the researchdesign, their interactioneffects studiedwith a better degreeof pre-
cision, and a better understandingof their dynamic- as opposed to static -
potentialitiesgained.
But the relationshipwould also imply that a point would eventuallybe
reachedwhere furtherincrementsin methodologicalcomplexity would not be
accompaniedby a correspondingimprovementof conceptualclarity but may
 A Minimax Principle for Educa tional R esearch 1 05

actuallyresultin a decrementof such. This is to say that a point of diminishing

returnswould be reachedsomewherealong the continuumof methodological
complexity. Furtherelaborationof experimentaldesign,more intensiveanalysis
of experimentalvariables,and more detailed statisticaltreatmentof experi-
mental data would not be followed by a better understandingof the relationships
being investigatedor establishedin the process.
If educationalresearchdoes indeed suffer from methodologicaldominance
and if the relationshipwith conceptualclarity is curvilinear,it would follow
that some kind of trade-offbetween the two would frequentlybe necessary.
Hopefully, all educationalresearcherswould desireto produce researchfindings
that are (1) intuitively appealing,(2) readilyunderstood,and (3) widely accept-
able. They would not be expected, however,to optimize any of these three at
the riskof debasingmethodologicalcurrency.As criteria,it could well be that
intuitive appeal,readyunderstanding,and wide acceptancecould not be optimized
equallywell or coterminously.As criteria,they are not necessarilyexhaustive
or mutuallyexclusive.
Yet, intuitive appeal,understanding,and acceptanceare preciselywhat more
educationalresearchersneed to considerin the selection and applicationof a
methodology to researchproblems.Most researcherswould be genuinelypleased
if their researchefforts, both method and findings,were intuitivelyappealing;
produceda readyor generalunderstandingof theory, method, and findings
amongeducationalpractitioners;and found a wide acceptance,application,or
use. Even in a less-than-idealisticworld these outcomes should not be scorned.
Yet, it would be foolish to regardthem as anythingother than a highly infrequent
event, as educationalresearchis currentlytaught, conducted, and published.
Severalimplicationsof the relationshipbetween conceptual clarity and
methodologicalcomplexity may be directly drawn.One concernsthe quality
and content of graduateprogramsin educationalresearchwhile another pertains
to the publicationpolicies of researchjournals.The first implicationis based on
the reasonablysound assumptionthat most graduateprogramsdo a creditablejob
of teachingmethodology but may fail to expose students properlyto the sub-
stantiveissues and problemsof education. The difficulties of teachingproblem
sensitivityarewidely recognizedbut not easily overcome.It is too easy to fall
back on the instructionof techniqueswith the rationalizationthat the better
studentswill recognizea "realproblem"once they encounterit. It is even easy
to fantasizethat most graduateswill, after all, serveas consultantson research
projectswhere otherswill be requiredto handle the substantiveor conceptual
issues.This places the graduatein a role of mechanicwith no acknowledgment
of responsibilityfor the actual operationof the machinery.
To continue in this manneris to encouragethe conditionswe should seek to
remove.Few professorsof educationalresearchare happy to be cast in the role
of "methodologist"when the role insinuatesa lack of realisticcontact with the
 106 Fin cher

substantive issues of education. Yet, it produces an unhappy schism in the

student's graduate program when all discussion of educational matters is left to
other courses and other professors. There is no neat and simple division between
method and substance that easily fits the procrustean implications of most
graduate courses. Five hours in the "philosophy of education" will not suffice
for a meaningful exposure to educational issues, problems, trends, and dilemmas.
The second implication will not delight journal editors but it is nonetheless
important. More editors should be prepared to ask the question: "Is all this
methodology necessary?" If the elaborateness of design or complexity of
analysis does not contribute to a better understanding of the problem the
author himself has chosen to study, can the paper be justified on the basis of
methodology alone? Given the triviality of a problem or the insignificance of a
question, can any amount of methodological legerdemain redeem the publish-
ability of the paper? Turner (1973) has suggested otherwise. His analysis indi-
cates that manuscripts are far more likely to be rejected for a lack of "signifi-
cance" than for any other single reason. Yet, judging from what is published
in many educational journals, a cynic can wonder why the criterion of
"significance" has not been applied more frequently.
The consideration of more substantive issues in research journals may or may
not increase the readership of such journals. There is good reason to believe,
however, that as a journal article is more widely understood and as it definitely
throws light on problems, issues, or questions that are meaningful to the reader,
the choice and application of methodology would be directed to the solution,
answer, or clarification of the subject ostensibly investigated. In any event, it
should help eliminate some of the distortion, weakening, or undermining of
concept by method. Other opportunities could be provided for methodological
arabesques that frequently obscure the intelligent discussion of educational issues.

SUMMARY AND CONCLUSIONS

The conceptual shift from fundamental, basic, theoretical, conclusion-

oriented research to mission-oriented, applied, product/output, problem-solving,
policy-oriented, decision research suggests strongly that educational research
would benefit from a realistic application of a minimax principle that would
produce a better balance between conceptual clarity and methodological com-
plexity. Irrespective of whether the research emphasis is on general theory de-
velopment, product development, or policy analysis, there should be advantages
to reducing the methodological ambiguities that are involved in so much educa-
tional research. If this should enhance the overall concern for substantive issues
in education, the benefits may not be easily costed out but they could be
immeasurable and they could be obvious to a public that eventually bears all
elements of that cost.
 A Minimax Principle for Educational Research 107

In game theory a minimax approach is usually regarded as a highly conservative

approach to the solution of problems. Not only is the worst expected, but actions
may be taken as if the worst was inevitable. The suggestion to minimax meth-
odological complexity need not be regarded as a pessimistic view of educational
research, its current status, or its potential development. Nor should the sugges-
tion be regarded as a specious plea for theory development, model construction,
or other conceptual phases in the larger process of research. The plea is simply
to restore a more suitable balance and to reintroduce, where it is so frequently
absent, a genuine concern for educational issues, problems, and trends.
The advocacy of a concern for educational issues rests on the belief that the
proper audience for educational research must be larger than the group of edu-
cational researchers at work in a particular problem area. There are numerous
other groups that do indeed have a vested interest in educational research. The
loss of their interest in educational research as it has been conducted during the
past twenty years is undoubtedly related to the disenchantment that has fallen
upon educational research as a federally sponsored endeavor. The accusation
that most educational research is irrelevant is a harsh one but nonetheless one
that must be heeded. A concern for policy issues in education means quite
simply that educational research must regain the authority it has lost for the
formulation of policy. That authority can be regained only if educational re-
searchers are willing and able to deal with educational matters.

REFERENCES
Averch, H. A., Carroll, Stephen J., Donaldson, Theodore S., Kiesling, Herbert J.,
and Pincus, John (1972). "How Effective Is Schooling? A Critical Review
and Synthesis of Research Findings." Santa Monica, Calif.: Rand Corporation.
Campbell, D. T., and Stanley, J. C. (1963). "Experimental and Quasi-Experi-
mental Designs for Research." Chicago: Rand McNally.
Committee for Economic Development (1968). "Innovation in Education: New
Directions for the American School." New York: Author.
Cronback, L. J., and Suppes, Patrick (Eds.) (1969). "Research for Tomorrow's
Schools: Disciplined Inquiry for Education." New York: Macmillan.
Glennan, Thomas (1973). National institute of education: A personal view.
Educational Researcher, 13-16.
National Center for Educational Research and Development (1969). Educa-
tional Research and Development in the United States." Washington, D.C.:
U.S. Government Printing Office.
Turner, Richard L. (1973). A report from the retiring editor. American
Educational Research Journal, 10:1-3.