{J.+€rtt

ls

( w'"*"2

Af

l*<l r-v*r+'*z-

/U-t

/-4Vz'14

4,t

r6s

rc-

folt^cz r-To,4

)\1n)

b

Jarausch,Konrad H., and Kenneth A. Hardy. euantirafive Methotls
for
Historians: A Guide to Research,Date, and statistics. chapel Hill,
NC: University of North Carolina press, 1991. xv + 247 pp.
Appendices, index. Hardcover, ISBN 0-907g-lg47_6, $45.00.
Softcover,ISBN 0-8028-4309l, $ 14.95.
Historians have always counted things. Until fairly recently,
however,overt quantificationdid not play a centralrole in historians'data
collection and analysisprocedures.This has not been the casefor many
other types of researchers. Scientific researchbegan with the Ancieni
Greeks,if not earlier,and developedrapidly during the Renaissance.welldefined scientific models and methodscame into common use during the
late nineteenthcentury. Today, the various scientific approaches,which
rely heavily upon quantification,remain in vogue o.ong r"r.archers in a
wide rangeof fields.l
Historical research,or historiography,began even earlicr than
scientificresearch.Generallyspeaking,early historianswrote narrative
accountsof political history that focusedon power elites. l.hey tentlcclto
eschewquantitativeanalysis,and they cameto seetheir reportingmotle as
more literary than scientific. However,at leastas early as the seventeenth
lJere T. Humohrevs,"Precursorsol'Musical
Aptitur.leTesting: From thc Crccks through
the.work of Francisbatton," ttrrriitiii)"ii'inMusic
Educution 4t (winrer r993): 3163t8.

*iliii

e

,

-' '; :.-.."-ji.::1i
'i;
' : 1 - ' *{iF.S*,
:'

t'--11:{j;'

._ i

-.r::'-.','.-*,.:':...,,. ....tjfl,i*=11:1!ffiiffi
.r,
i ; .. ' .' j, .: ..:,l..j3i.i : .. . .
-^

BooxRevrws
I

il
1

,(
I

t4l

ccntury,historiansbeganto adoptsomeattituclesand horrow methodsfrom
lhc scicntificcommunity.
Duringthc nincteenth
century,whenscientists
developed
sophisticated
researchmodels,historianscameunderthe influcnceof the famousGerman
historianand teacherLcopold von Ranke(1795-lgg6). Rankeadopted
morc scientificapproaches
thanprevioushistorians,especiallyhis insistlnce
on meticulousobjectivityin the verificationof sources.However.he also
insistedon retainingthe narrativeaccountand on not employingspecialized
researchproceduresandjargon-ladenreporting language. He reasonedthat
ordinary people should be able to read historical accountseasily, which
would enhancetheir knowledgeand insightsand therebyfacilitatecitizen
participation in the political discourse associatedwith the emerging
Europeandemocracies.2
Ranke's model served reasonablywell during his day, but modernday historiography has become much more diverse in responseto
heightenedsocial sensibilities,so much so that "the universeof historians
hasbeenexpandingat a dizzying rate."3 over time, somehistorianshave
come to value the study of a wide array of subject areas-an enonnous
expansionin scope from the almost exclusive emphasison European
political historyof Ranke'sera.
Different subjectslend themselvesto different researchtechniquesand
reportingstyles. Hence,increasingnumbersof historiansnow experiment
with non-traditionalsources,researchmethodsincluding quantification,and
reportingstyles to help them grapplewith new researchquestions.In short,
historiansas a group no longer confine themselveslargely to male European
political history; they no longer employ the literary-stylenarrativeaccount
exclusively; and they no longer utilize written documentarysourcematerial
to the exclusionof other typesof sources.
.tC9qrg G. lggels and fanrcs M. Powell, Leopold van Ranke and thc Sluping of the
Historical Discipline (Syracuse, NY: Syracuse University Press, 1990), xv, xvi. ite arithors
discuss the inherent tension between what they call qualitative humanistic and quantitative
social science approaches,and the fact that some Enlighienment scholarsroutinely used both
approachcs. For example, a progressive Cdingen historian called statistics "frozen history"
and history "statistics come alive" (p. l0).
lPcter Burke, "Overture: The New History, lts Past and lts Future," in New Pcrspectives
on Historical Writing (Cambridge, UK: Polity Press, l99l), l.

ffi
t42

Boor Rsvnws

Jarauschand Hardy's QuantitativeMethodsfor Historians does not
tell this story of how and why quantificationevolved. The authorsdo state
that quantitativetechniquestransformedeconomichistoriographyduring the
1950sand 1960s,and that historiansnow usequantification extensivelyard
successfullyin historical demography,human health and diseasehistory,
and severalother {ields including political history. They also point out that
quanto-historians
are achievingsomeof their most importantresultsin the
new social history. According to the authors, "this recent and most
dynamic sectorof historical inquiry shifted attenrionaway from political
elites to the often mute massesand addressedthe power of social structures
or customsratherthan of political events" (p. 6).
Background aside, the minimal goal of euantitative Methodsfor
Historians is to help readersunderstandthe growing body of published
quantitativehistorical research;the maximum aim is to inspire readersto
apply statisticalmethodsin their own reseiuch.The authorsclaim they are
responding"to a much lamentedtraining deficit," and that they seek "to
renderresearchdesigntransparent,
to help with establishingdatabases,ancl
to provide enoughstatisticalinformation so that analyticalprocedurescan be
appliedresponsibly"(p. xiii).
In the first and last few chaptersof the book, the authorstackle the
maximum aim of inspirationby addressingtraclitionalhistoriansand some
of their perceivedsensibilities.Thesechaptersdeal with such issuesas
justifications for quantification in history, the use of computers by
historians,conceptsand problemsrelated to the selectionof research
questions,quality of sourcematerial,very large and very small sanrple
sizes,and hypothesistestingand modelbuilding. They even make sonre
practicalsuggestionsfor overcomingtraditionalhistorians'resistanceto
quantification.
The authorsadmitto a numberof inherentweaknesses
in thc nrelho<J.
but insistthat mostof the striclent"criticismsof numericalantiquarianismor
statisticaldehumanization
aregroundless,
sincethey invalidatethe misuseof
quantificationrarherthan its proper application. . )' (p. 2). They note
further that althoughmany of the criticismsof quanto-historyrelate to
inflated claims made by early advocates,the method can accommodate

..,t';,''$i.triitlt:lfr:tlt

n.::ir::l ' rj:n: '
.i
:j'{

Boox Revlews

143

rargcrdatasetsand thereforecan providegreater
precision,identifymore
caus.tivcvariabres,and producemoregenerarizabre
resutts.
As for the quarity of sourcemateriar,the
authorsassertthat ..the
critcriaof traditionarsourcecriticismappryto quantitative
work,perhapsin
an cvcn more stringentway" (p.26). The authors
..as
view quantification
an additionartoor of historicarresearch"(p. 4) that
.un he-rpcorrectcertain
sourccmaterialproblemsin traditionalhistoriography,
suchasmissingand
unrepresentative
data. These inequitiesin sourte-material
ort.n o;.u,
becauseoriginal categorizations
of dataresultedfrom qualitativedecisions
madeby historicalelites,and because
dataon elites",. ',no. likely to have
survived than data on the poor. In a further attempt
to parry a common
source-relatedcriticism of quantificationin history,
the authorsnote that
"[v]irtually
all historicarrecordsare resscompretethan a schorar
might
desire"(p.73'1.
Most of the intervening chaptersconsist of crear,
straightforward
presentationsof some basic conceptsand how to's
of statistiil analy.sis
from descriptionsof mean, median, and mode; through
chi-square,
correlation,regression,analysisof variance,and multiple
regression;and
on to an overview of time seriesanalysis,factor analysis,
clusteranalysis,
LISREL (linear structuralequationmodelling), ani nonlinear
models.
Appropriately for causalhistorical analysis,th-eauthors
discussmultiple
regres-sion
techniquesin some detail, including interaction analysis,
and
providebrief discussionsof importanttechniqueilike
discriminantfunction
and frequently encounteredartifacts like categoricalvariables
as predictors
in multiple regressionanalysis.
This book representsan improvement over the fairry
numerous
quanto-historybooks publishedsincethe early 1960s. It
is much stronger
in advancedstatistical expranationsand eiamples than
earlier books.
Indeed,most of the statisticalexpranationsare crearerthan
thosefound in
the generalrun of non-historicaleducationand psychologystatistics
books,
in part becauseit does not emphasizecomputation methohs. Furthermore,
the authorsgive soundadvice aboutprocedures,such as their admonition
to
keep a record of small decisionsrelative to categorizations
of subjects,
becausenumberstend to mask such decisionsmadeearly in a study.
They

ffi.
r44

Boox Rrvrcws

recommendlisting doubtful cases"so that they can be resolved
[later] in a
uniform way" (p.43).
Despite its strengths,it is difficult to imagine using this
book as a
textbook for music educationclasses,and not only because
graduatemusic
educationdepartmentstypically do not offer an entire
course on a sub_
specialtylike quanto-history. Unfortunately,becausethe
authorsconceived
the book both as an introduction to quantification in history
and as a guide
"to
scholarswho havesomeprior quantitativeexperience,'(p.
xiv), portions
of it are too simple for advancedscholars(e.g.,the
discussionon backing
up computerfiles), while other parts uuetoo advanced
for novices.
Instead,the book could serveas supprementaryreading
material for
studentsand as a referencemanual for more advanced
researchers.Even in
that role' the book would be more useful to music education
researchemif it
had includeda wider rlnge of exampresfrom the quanto-history
research
literature. Most examplesin the book come from
a quantitativestudy of a
former united StatesArmy fort, for which ,"r"or.h"r,
obtainecrdata on
suchphenomenaas propertyownershipby officers
and enlistedpersonnel
stationedat the fort. The authorsreturn repeatedly
to the army data set to
provide examplesof researchquestionsand
anarysistechniques. The
authorsalso discusshow other quanto-historian,*"i"
able to augmentancl
partially overturn a conclusionreachedby
traditional historians: that
supportfor the Nazi movernentin Germany
came in large part from the
lower middle class. Quanto-historians
demonstratedthat, ..while lower
middle class occupationsare somewhat
overrepresented,early Nazi
membershipdependsmore on sex,age,or religion,,
(p.29).
Theseexamplesfrom the quanto-historyresearch
riteratureherpa great
deal, but in someways the book still comes
off as a methodin searchof a
subject' The pedagogical
deviceof returningover andover to the samedata
set resultsin the authorssacriricingwhat
music educationresearchcrs
interestedin quanto-history
ncedmost: ideasaboutwhat to study. sorrrc
examplesof the many successfur
publishetrapprications
of quanto-hist'ry
techniquesfiom otherfierdsmight stimuratethe
fbrmurationof new research
ideason the part of music educationresearchers.
After ail, the music
educationresearchstable is fully stocked
with individuals skilled in

;

t,

I
r

,...,..j,,,

Boor Revuws

t45

statisticalanalysis,but the literaturecontainsonly a few examplesof
quantilicationin history. This is not a criticismof the bookperse, because
thc authorscould not have beenexpecte<l
to proclucea book for music
educationrescarchers
alone. Rather,it is simply an observationaboutthe
book'spotentialusefulness
to the field of musiceducation.4
on the technicalside,a few editing problemsappearto haveresulted
from the fact that two people wrote the book. For example,the authors
statethat it is unnecessary
to assignnumbersto nominaldatacategories,but
in anotherplacethey advisethe readersto do just that. Also, the authors
recommendusing microcomputers,while placing much emphasison
mainframesoftwarecommandsin otherportionsof the book. In fact, much
of the material an word processing,electronicmail, computerizedlibrary
services,datastorageand management,
memoryrequirements,
and statistics
softwarewas not currentwhen the book cameout in 1991,and very little of
it remainsso now.
A curious omission is the authors' failure to discusstextual and
artifactual(including photographic)analysis,an approachmusic education
and other fields haveusedand which shouldbe includedunderthe rubric of
quanto-history. They also fail to treat sampling techniquesadequately,
althoughthey appropriatelyrecommendsamplinginsteadof studyingentire
populationsin somecases.S
As the title suggests,Jarauschand Hardy's book is more for
traditional, non-quantifierhistoriansthan for non-historianquantifiers.
Despite this orientation,however,the authorscaution researchersagainst
spending too much time on mechanicsand not enough on hypothesis
developmentand interpretationof data.

4For some eram-ples of studics from various fields, many of which employed
quantification, see Jim Sharpe, "History from Below," in New Perspectives on Hisiorical
lAriling, cd. Peter Burke (Cambridge,UK: Polity Press, l99l), 28-36.
sA statistical analysis of photographs in textbooks appears in lulia Eklund Koza,
"Femafes
in f 9E8 Middle School Music Ter,tbooks: An Analysis of lllustrations," toumal of
Rcsearchin Music Mucation 42 (Summer 1994): 148-171. In addition, Jere T. Humphreyi,
"Sex
and Demographic Representationin Two Music Education History Book," Bulletin of ihc
Council for Rcsearch in Music Education (in press). uses statistics in an analysis of
photographsand text content.

t46

Boor Rnvnws

The authorsdiscussthe accomplishments
of quanto-historians
in some
fields, but they also note that psycho-history,intellectualhistory, political
biography, and diplomatic history "have hardly been touched by
quantification" (p. 7). with a few exceptions,the same is true of music
educationhistory. Paradoxically,despitethe fact that the authorsof most of
the extant quanto-history studies in music education are primarily
quantitativeresearchers,
they failed to empty inferential statisticsin their
analyses.The neglectof sophisticated
quantitativemethodsin musichistory
and musictheoryresearchis even more remarkabledue to the lengthof the
researchhistoriesin thosefields.6
Quantification will not replace traditional techniquesin music
educationhistory,but quantificationcould help selectivelyenlargeand
enhancepastand future findings. This book, the most recentin a long line
of bookson quanto-historytechniques,could serveas a reminderto music
educationhistoriansto considerthe useof alternativeapproaches,especially
approachesthat might lead to different interpretationsof the past.
Four examplesof such work in music educationinclude statistical
analysesof doctoral dissertationson music educationand music therapy
history,sex and geographicalrepresentation
in music educationhistory
books, the demographicsof a researchjournal's editorial committee
6ln addition to Koza, rcccnl cxamplcs.of publishcdhistoricalstutlicsthat incorporule4
d e s c r i p t i v eb u l n o t i n l c r c n l i a l s t a l i s t i c s i n c l u d e J o h n W . C r a s h e l a r r t l A n n c ' L o w e ,
"Contribution
of School Music Educatorslo lhc ResearchLiteralure as Publishcdin thc Jourtrul
of .Reseurch in Music Education, 1953-1993," The llulletin o1 ttistoriiil
ncst,urclt in Musit:
Education l6 (January 1995): 122-135; Srevcn K. Hcdrien, 'iMusicl e.ru.urru,, Rcscarch:
A
rjozcn convcntions and a D91c1 JRME Yolumcs," The Bulletin o! Historical Research
in Musit:
Educution l5 (Seprcnrbcr1993): l7-30; Steven K. Hcdtlen,i.R"r"arch Scssionsat MIINC
uonlcrcnces: 197(J-1990,".-Th.e
()uurterly Journul of Music Teaching arul Leurning 3
{Spring
N. Hcllcr, "Historical Rcscarih in Music Ertuiarion and Muiic'ti,Jrupy',
19'^2.1:.-.89-lll.Gcorgg
A Quartcr-ccn_luryol Rcscarch, Wlif i!g, and Publication," The
euurterly Journul ,4' Mislc
..Sib1".1,
in Music llducarien
.(Spring tgrtZ):.St-02; John Krarus,
!,e1chiyE"i,!.*"::.i.::.(..3
Kcscilrcn'- f 96l-1990,"'fhe Quurrerlv Jounul of Music Teuching-undLeurning 3 (Wilrcr
1992): 50-55; John Kralus, Emincncc in Music Etlucarion R"icar.tr ii Mcasurcd
in rhc
Haryl.booklf .Reseurchon M_us_i.c
Teuching
und lturning," llullctin ,rf rhi ii,uncil Jitr Rt'stttrt.lt
'2ll-32;
..l'icqucniiy
in Music Education lllt (Fall 1993):
Duanc'Sarnpte,
crrci Srudies as
fndicators of Music Education ResearchInrcrcsls, tCol-1.)d9,'; lrttr,-,-tii'-,,J
R"stttrch itt Musit.
Iiducation 4O (Summcr 1992): t53_157; Charlesp. SchnrirJtantj Srcvcn"F.Zdzinski, ..Circtl
QuantitativeRcscarchArtic'lcsin .Music bducationRcsearchJournals,tgzi-tq90, A Contcnl
<tf Research in Music Dducation 4l (Spring 1993):
fngfysis of_Selectedsrudies," Jgyrnlt
5-18; and cornelia Yarbrough, "A-conrent Analysis ur tlrr- tii'iii-l- uJ i"r"orrh
in Music
Education, 1953-1983," Joumul of Research in Misic Education rz
t9i4l: zl3-222.
t*irir.i

Boor Rrvrws

147

mcmbcrship,and membership
demographics
of a professionar
organization
during the organization'searly years. All four of these
studiesare recent,
all incorporateboth descriptiveanclinferentialstatistics,
and all deal with
mcasurable
changesover time.7
Thus,music educationresearchers
havejust begunto emproyquanto_
history techniques,and only for descriptionsand uniy*",
of datarelatedto
the rcsearchliterature,textbookcontent,and profesiional
demographics.
Nevertheless,music educationprofessorswho wish to
add quantitative
expertiseand insights to their graduatestudents'arsenal
of historical
researchtools could assignthesestudies,selectedstudiesfrom
other fields,
and portionsof Jarauschand Hardy's euantitative Methodsfor
Historians.
Music educationhistorianswho adoptquantitativeanalysisshould
not
faceat leastone of the obstaclesencountered
by quanto-minded
historiansin
other fields: a lack of empathy and knowledge on the part
of other
researchers.The fact that most music educationresearchers
already use
quantitativetechniquesbodes well for the acceptance
of high_quality
quantitativehistorical studies by the researchcommunity.
However,
historiansmust posethe researchquestionsand not rely exclusively
upon
their quantitative colleagues. Before that happens,music education
historianswho haveeschewedquantitativemethodi in the pastwould
have
to embracethem for sometypesof analyses.
It is probably safe to say that most of the important researchquestions
in music educationhistory are not amenableto quantification. However.
sorneare, and Jarauschand Hardy's book provides a starting place for
any
researcherwho wantsto acceptthat challenge.
Appropriately, the authorsmake their strongestcasefor quantification
in history on methodologicalgrounds: "while numbers can seldom
representhuman feelingsadequately,they do help resolvethe bothersome
issue of representativenesswhich undercuts much textually oriented
7JercT. Humphreys,.David
M. Bess,and Martin f. Bergee,..Doctorar
Dissedations
on
"t
uii"iliini'i
karningpress): Hu-mphreys,
irid o".igiapr,i'" iepiJsentarion,,;Jere T.
H-umphreys
andSandraL. Stauffer."DemographicCharacreriitii's.i jirti. mnoriat Commirtee
manuscripttanil iere r. Humfrrriyi r".t'Crt"a"s p. Schmidt,
l!::f_"1_lnpublished
-'uemographrc
characteristicsof Early MENC Members:- l9l2_193g,. unpuullitrei
manuscript.
. __.
the
Hisrory
of MusiceoulatpJr
andMusic...i'terapy:,rn,
eiifrriy t"iiii
..Sex
and
(in

148

Boox REvmws

research"(pp l9o-l9l). Indeed,music educationhistorianshave written a
great deal about the profession's leading individuals, programs, and
organizations,but far less about ordinary music teachersand studentsand
the cultures from which they came. euantitative analysis, applied
appropriatelyand skillfully, could help broadenand deepenthe profession's
knowledgeof its past. Quantitative Methodsfor Historianr could provide
somesupport for that effort.8
-Jere T. Humphreys
ArizonaStateUniversity

{

"Expan-tling
jrLnn1"1t,
thc Horizonsof Music ErJucarionllisrory
.,.^,..,._tJ.:1,"-I-1
)ocfofogy'"'Ihe
QuurterlyJournurof MusicTeuching
und lzarning (in prcss).