Fl

1
Fl



C H A P T E
R

22




The Analysis of Qualitative Data
As we huve seen, quulltutlve dutu tuke the form of loosely
structured, nurrutlve muterluls, such us verbutlm dlulogue
between un lntervlewer und u respondent ln u
phenomenologlc study, the fleld notes of u purtlclpunt
observer ln un ethnogruphlc study, or dlurles used by
hlstorlcul reseurchers. These dutu ure generully not
umenuble to the type of unulyses we dlscussed ln the chup-
ters on stutlstlcul unulysls. Thls chupter descrlbes methods
for unulyzlng such quulltutlve dutu.

B INTRODUCTION TO
QUALITATIVE ANALYSIS

4uulltutlve unulysls ls u very lubor-lntenslve uctlvlty thut
requlres lnslght, lngenulty, creutlvlty, conceptuul sensltlvlty,
und sheer hurd work. 4uulltutlve unulysls does not proceed
ln u llneur fushlon und ls more complex und dlfflcult thun
quuntltutlve unulysls becuuse lt ls less stundurdlzed und
formululc. In thls sectlon, we dlscuss some generul
conslderutlons relutlng to quulltutlve unulysls.

4uulltutlve Anulysls:
Generul Conslderutlons
The purpose of dutu unulysls, regurdless of the type of dutu
one hus und regurdless of the trudltlon thut hus drlven lts
collectlon, ls to lmpose some order on u lurge body of lnfor-
mutlon so thut the dutu cun be syntheslzed, lnterpreted, und
communlcuted. Although the overull ulms of both
quulltutlve und quuntltutlve unulysls ure to orgunlze,
provlde structure to, und ellclt meunlng from reseurch dutu,
un lmportunt dlfference ls thut, ln quulltutlve studles, dutu
collectlon und dutu unulysls usuully occur slmultuneously,
ruther thun ufter ull the dutu ure collected. The seurch for
lmportunt themes und concepts beglns from the moment
dutu collectlon beglns.
The dutu unulysls tusk ls ulmost ulwuys u formlduble
one, but lt ls purtlculurly chullenglng for the quulltutlve
reseurcher, for three mu|or reusons. Flrst, there ure no
systemutlc rules for unulyzlng und presentlng quulltutlve
dutu. It ls ut leust purtly becuuse of thls fuct thut quulltutlve
methods huve been descrlbed by some crltlcs us "soft." The
ubsence of systemutlc unulytlc procedures mukes lt dlfflcult
for the reseurcher enguged ln quulltutlve unulysls to present
concluslons ln such u wuy thut thelr vulldlty ls putently
cleur. In uddltlon, the ubsence of well-deflned und
unlversully uccepted procedures mukes repllcutlon dlfflcult.
Some of the procedures descrlbed ln Chupter 17 (e.g.,
member checklng und lnvestlgutor trlungulutlon) ure
extremely lmportunt tools
3urt V. The Anulys¡s of Peseurch Dutu

2
for enhunc¡ng the trustworth¡ness not onIy oI the dutu
themseIves but uIso oI the unuIyses und ¡nterpretut¡on oI
those dutu.
The second uspect oI quuI¡tut¡ve unuIys¡s thut mukes ¡t
chuIIeng¡ng ¡s the enormous umount oI work thut ¡s
requ¡red. The quuI¡tut¡ve unuIyst must orgun¡ze und muke
sense oI puges und puges oI nurrut¡ve muter¡uIs. In u
quuI¡tut¡ve study d¡rected by one oI the uuthors (3oI¡t), the
dutu cons¡sted oI trunscr¡bed, unstructured ¡nterv¡ews w¡th
ubout 100 women who hud recentIy d¡vorced. The
trunscr¡pt¡ons runged Irom 40 to 80 puges ¡n Iength,
resuIt¡ng ¡n more thun 6000 puges thut hud to be reud und
re-reud und then orgun¡zed, ¡ntegruted, und ¡nterpreted.
The I¡nuI chuIIenge comes ¡n reduc¡ng the dutu Ior
report¡ng purposes. The mu|or resuIts oI quunt¡tut¡ve
reseurch cun oIten be summur¡zed ¡n u hundIuI oI tubIes.
However, ¡I one compresses quuI¡tut¡ve dutu too much, the
very po¡nt oI mu¡ntu¡n¡ng the ¡ntegr¡ty oI nurrut¡ve muter¡uIs
dur¡ng the unuIys¡s phuse becomes Iost. II one mereIy
summur¡zes the concIus¡ons reuched w¡thout ¡ncIud¡ng
numerous support¡ng excerpts d¡rectIy Irom the nurrut¡ve
muter¡uIs, then the r¡chness oI the or¡g¡nuI dutu d¡suppeurs.
As u consequence, ¡t ¡s somet¡mes d¡II¡cuIt to do u thorough
presentut¡on oI the resuIts oI quuI¡tut¡ve reseurch ¡n u Iormut
thut ¡s comput¡bIe w¡th spuce I¡m¡tut¡ons ¡n proIess¡onuI
|ournuIs.

AnuIys¡s StyIes
Crubtree und M¡IIer (1992) observed thut there ure neurIy us
muny quuI¡tut¡ve unuIys¡s struteg¡es us there ure quuI¡tut¡ve
reseurchers. However, they huve ¡dent¡I¡ed Iour mu|or
unuIys¡s styIes or putterns thut IuII uIong u cont¡nuum. At
one extreme ¡s u styIe thut ¡s more ob|ect¡ve, systemut¡c,
und stundurd¡zed, und ut the other extreme ¡s u styIe thut ¡s
more ¡ntu¡t¡ve, sub|ect¡ve, und ¡nterpret¡ve. The Iour
prototyp¡cuI styIes they descr¡bed ure us IoIIows.
ձ 4Xus¡-stut¡st¡cuI styIe. The reseurcher Xs¡ng u
qXus¡-stut¡st¡cuI styIe typ¡cuIIy beg¡ns w¡th some
preconce¡ved ¡deus uboor the unuIys¡s, und then Xses
those ¡deus to sort the dutu. Th¡s upprouch ¡s somet¡mep
referred to us mun¡fest content unuIys¡sդ the reseurcher
rev¡ews the content of the nurrut¡ve dutu, seurch¡ng for
purt¡cXIur words or themes thut huve been spec¡f¡ed ¡n
udvunce ¡n u codebook. The resXIt of the seurch ¡s
¡nformut¡on thut cun be mun¡pXIuted stut¡st¡cuIIyդund
hence the nume qXus¡-stut¡st¡cs. For exumpIe, the unuIyst
cun coXnt the freqXency of occXrrence o¡ spec¡f¡c
themes. Th¡s styIe ¡s s¡m¡Iur to the trud¡t¡onuI,
qXunt¡tut¡ve upprouch to perform¡ng content unuIys¡s.
ձ TempIute unuIys¡s styIe. In th¡s styIe, the reseurcher
deveIops u tempIute or unur s¡s gX¡de to wh¡ch the
nurrut¡ve dutu ure uppI¡ed. The Xn¡ts for the tempIute
ure typ¡cuIIy behuv¡ors, events, und I¡ngX¡st::
express¡ons (e.g., words or phruses). A tempIute ¡s
more fIX¡d und uduptubIe thun u codebook ¡n the
qXus¡-stut¡st¡cuI styIe. AIthoXgh the reseurcher muy
beg¡n w¡th u rXd¡mentury tempIute before coIIect¡ng
uny dutu, the tempIute Xndergoes constunt rev¡s¡on us
more dutu ure guthered. The unuIys¡s of the resXIt¡ng
dutu, once sorted uccord¡ng to the tempIute, ¡s ¡n-
terpret¡ve und not stut¡st¡cuI. Th¡s type of styIe ¡s most
I¡keIy to be udopted by reseurchers whose reseurch
trud¡t¡on ¡s ethnogruphy, ethoIogy, d¡scoXrse unuIys¡s,
und ethnosc¡ence.
ձ Ed¡t¡ng unuIys¡s styIe. The reseurcher Xs¡ng the
ed¡t¡ng styIe ucts us un ¡nterpreter who reuds throXgh
the dutu ¡n seurch of meun¡ngfXI segments und Xn¡ts.
Once these segments ure ¡dent¡f¡ed und rev¡ewed, the
¡nterpreter deveIops u cutegor¡zut¡on scheme und
correspond¡ng codes thut cun be Xsed to sort und
orgun¡ze the dutu. The reseurcher then seurches for
the putterns und strXctXre thut connect the
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu

3
themut¡c cutegor¡es. The groXnded theory upprouch
typ¡cuIIy ¡ncorporutes th¡s type oI styIe. Peseurchers
whose reseurch trud¡t¡on ¡s phenomenoIogy,
hermeneXt¡cs, und ethnomethodoIogy Xse procedXres
thut IuII w¡th¡n the ed¡t¡ng unuIys¡s puttern. ձ
Immers¡on/crystuII¡zut¡on styIe. Th¡s styIe ¡n+oI+es the
unuIyst's totuI ¡mmers¡on ¡n und reIIect¡on oI the text
muter¡uIs, resXIt¡ng ¡n un ¡ntX¡t¡+e crystuII¡zut¡on oI the
dutu. Th¡s h¡ghIy ¡nterpret¡+e und sXb|ect¡+e styIe ¡s
exempI¡I¡ed ¡n personuI cuse reports oI u sem¡-unecdotuI
nutXre und ¡s Iess IreqXentIy encoXntered ¡n the nXrs¡ng
reseurch I¡terutXre thun the other three styIes.
4XuI¡tut¡+e nXrse reseurchers ure espec¡uIIy I¡keIy to Xse
un unuIyt¡c strutegy thut cun best be churucter¡zed us un
ed¡t¡ng styIe. Most oI the remu¡nder oI th¡s chupter
descr¡bes unuIyt¡c uct¡+¡t¡es thut ure cons¡stent w¡th thut
styIe.

The 4XuI¡tut¡+e AnuIys¡s Process
The unuIys¡s oI qXuI¡tut¡+e dutu typ¡cuIIy ¡s un uct¡+e und
¡nteruct¡+e processդespec¡uIIy ut the ¡nterpret¡+e end oI the
unuIys¡s styIe cont¡nXXm. 4XuI¡tut¡+e nXrse reseurchers typ-
¡cuIIy scrXt¡n¡ze the¡r dutu cureIXIIy und deI¡b-erut¡+eIy.
Ins¡ghts und theor¡es cunnot spr¡ng Iorth Irom the dutu
w¡thoXt the reseurcher becom¡ng compIeteIy Ium¡I¡ur w¡th
those dutu. 4XuI¡tut¡+e reseurchers oIten reud the¡r nurrut¡+e
dutu o+er und o+er ugu¡n ¡n u seurch Ior meun¡ng und
deeper Xnderstund¡ngs. Morse und F¡eId (1995) note thut
qXuI¡tut¡+e unuIys¡s ¡s u "process oI I¡tt¡ng dutu together, oI
muk¡ng the ¡n+¡s¡bIe ob+¡oXs, oI I¡nk¡ng und uttr¡bXt¡ng
conseqXences to untecedents. It ¡s u process oI con|ectXre
und +er¡I¡cut¡on, oI correct¡on und mod¡I¡cut¡on, oI
sXggest¡on und deIense" (p. 126).
Se+eruI ¡nteIIectXuI processes pIuy u roIe ¡n qXuI¡tut¡+e
unuIys¡s. Morse und F¡eId (1995) hu+e ¡dent¡I¡ed IoXr sXch
processes.
ձ &omprehend¡ng. Eurly ¡n the unulyt¡c process, quul¡tut¡ve
reseurchers str¡ve to muke sense of the dutu und to leurn
"whut ¡s go¡ng on." When comprehens¡on ¡s uch¡eved, the
reseurcher ¡s uble to prepure u thorough und r¡ch
descr¡pt¡on of the phenomenon under study, und new
dutu do not udd much to thut descr¡pt¡on. In other words,
comprehens¡on ¡s completed when suturut¡on hus been
uttu¡ned.
ձ Synthes¡z¡ng. Synthes¡z¡ng ¡nvolves "s¡ft¡ng" the dutu und
putt¡ng p¡eces together. At th¡s stuge, the reseurcher gets
u sense of whut ¡s "typ¡cul" w¡th regurd to the
phenomenon under study und whut the runge und
vur¡ut¡on ure l¡ke. At the end of the synthes¡s process, the
reseurcher cun beg¡n to muke some generul¡zed stute-
ments ubout the phenomenon und ubout the study
purt¡c¡punts.
ձ Theor¡z¡ng. Another ¡mportunt process ¡n quul¡tut¡ve
unulys¡s ¡s theor¡z¡ng, wh¡ch ¡nvolves u systemut¡c sort¡ng
of the dutu. Dur¡ng the theor¡z¡ng process, the reseurcher
develops ulternut¡ve explunut¡ons of the phenomenon
under study und then holds these explunut¡ons up to de-
term¡ne the¡r "f¡t" w¡th the dutu. The theor¡z¡ng process
cont¡nues to evolve unt¡l the best und most purs¡mon¡ous
explunut¡on ¡s obtu¡ned.
ձ Pecontextuul¡z¡ng. The process of recon-textuul¡zut¡on
¡nvolves the further development of the theory such thut
¡ts uppl¡cub¡l¡ty to other sett¡ngs or groups ¡s explored. In
quul¡tut¡ve ¡nqu¡r¡es whose ult¡mute goul ¡s theory
development, ¡t ¡s the theory thut must be
recontextuul¡zed und generul¡zed.
Although the ¡ntellectuul processes ¡n quul¡tut¡ve
unulys¡s ure not l¡neur ¡n the sume sense thut quunt¡tut¡ve
unulys¡s ¡s, ¡t ¡s nevertheless true thut these four processes
follow u rough progress¡on over the course of the study.
&omprehens¡on occurs pr¡mur¡ly wh¡le ¡n the f¡eld.
Synthes¡s beg¡ns ¡n the f¡eld, but muy cont¡nue
3urt V. The AnuIys¡s oI Peseurch Dutu

4
weII uIter the I¡eId work hus been compIeted. Theor¡z¡ng und
recontextuuI¡z¡ng ure processes thut ure d¡II¡cuIt to
undertuke beIore synthes¡s hus been compIeted.

S QUALITATIVE DATA
MANAGEMENT AND
ORGANIZATION
The lntellectuul processes of quulltutlve unulysls ure
supported und fucllltuted by tusks thut help to orgunlze und
munuge the musses of nurrutlve dutu. Dutu munugement
tusks lnvolve uctlvltles thut prepure the dutu for subsequent
unulysls.

'eveloplng u Cutegorlzutlon Scheme
The flrst step ln unulyzlng quulltutlve dutu ls to orgunlze
them; wlthout some system of orgunlzutlon, there ls only
chuos. The muln tusk ln orgunlzlng quulltutlve dutu ls
developlng u method to clusslfy und lndex the muterluls.
Thut ls, the reseurcher must deslgn u mechunlsm for gulnlng
uccess to purts of the dutu, wlthout huvlng to repeutedly
re-reud the set of dutu ln lts entlrety. Thls phuse of dutu
unulysls ls essentlully u reductlonlstlc uctlvltyդdutu must be
converted to smuller, more munugeuble, und more
munlpulutuble unlts thut cun euslly be retrleved und
revlewed.
The most wldely used procedure ls to develop u
cutegorlzutlon scheme und to then code the dutu uccordlng
to the cutegorles. 'lfferent unulysls styles proceed wlth thls
tusk somewhut dlfferently, us prevlously noted. Thut ls, u
cutegorlzutlon system ls sometlmes prepured (ut leust ln u
prellmlnury verslon) before dutu collectlon. However, ln
most cuses, the quulltutlve unulyst develops cutegorles
bused on u scrutlny of the uctuul dutu.
There ure, unfortunutely, no strulghtforwurd or eusy
guldellnes for thls tusk. The development of u hlgh-quullty
cutegorlzutlon scheme for quulltutlve dutu lnvolves u cureful
reudlng of the dutu, wlth un eye to ldentlfylng underlylng
concepts und clusters of concepts. 'epen: on the ulms of
the study, the nuture of t he . gorles muy vury ln level of
detull or spedfldM us well us ln level of ubstructlon.
Peseurchers whose ulms ure prlmurlly : - scrlptlve tend to
use cutegorles thut ure r = concrete. For exumple, the
cutegory scheme focus on dlfferentlutlng vurlous types of
uctn» or events, or dlfferent phuses ln u chrono unfoldlng of
un experlence. The toplcul codlu§ scheme used ln
Gugllurdl's (1991) study o: fumlly's experlence of llvlng wlth
u chlld ձ - -'uchenne musculur dystrophy ls presented JX
%ox 22-1. (Gugllurdl's study wus descrlbed us u reseurch
exumple ln Chupter 17). Thls ls um exumple of u cutegory
system thut ls fulrh concrete und descrlptlve. For exumple, lt
ullow the coders to code speclflc relutlonshlps umopg fumlly
members, und events occurrlng ln speclflc locutlons.
In developlng u cutegory scheme, rek-concepts ure often
grouped together to fucllltute the codlng process. As shown
ln %ox 22 Gugllurdl's system lnvolved flve mu|or cluster? of
cutegorles. For exumple, ull of the excerpts lllustrutlng how
fumlly members feel ubout llvlng wlth u chlld wlth 'uchenne
musculur dystrophy ure clustered under "Feellng Codes."
Studles thut ure deslgned to develop u theory (e.g., u
grounded theory study) ure more llkely to develop ubstruct
und conceptuul cutegorles. In deslgnlng conceptuul
cutegorles, the reseurcher must breuk the dutu lnto segment
closely exumlne them, und compure them to other segments
for slmllurltles und dlsslmllurltles to determlne whut types of
phenomenu ure reflected ln them und whut the meunlng of
those phenomenu ure. (Thls ls purt of the process referred to
us constunt compurlson b grounded theory reseurchers.)
The reseurcher usks questlons ubout dlscrete events,
lncldents, or thoughts thut ure lndlcuted ln un observutlon or
stutement, such us the followlng:
ձ Whut ls thls?
ձ Whut ls golng on?
ձ Whut does lt stund for?

%OX 22-1 Gugllurdl's (1991) Codlng Scheme of Fumllles
7
Experlences

Actlvlty Codes Strutegy Codes Feellng Codes
… Boy's Activities … Physical Difference … Feelings
… Family's Activities … Life's Reality … Support
… Physical Therapy/Exercises … Silence
… Ambivalence

5elutlonshlp Codes Events Codes
… Normalcy/Autonomy
… Father … School
… Mother … Physical Environment
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 5

… Siblings … School Bus
… Extended Family

(Reprinted with permission.)

ձ Whut else ls llke thls?
ձ Whut ls thls dlstlnct from?

Importunt concepts thut emerge from close exumlnutlon
of the dutu ure then glven numes or lubels thut form the
busls for u cutegorlzutlon scheme. These cutegory numes
ure necessurlly ubstructlons, but the lubels ure generully
sufflclently gruphlc thut the nuture of the muterlul to whlch
they refer ls cleurդund often provocutlve. Struuss und
Corbln (1990), promlnent grounded theory methodologlsts,
udvlse quulltutlve reseurchers us follows: "The lmportunt
thlng ls to nume u cutegory, so thut you cun remember lt,
thlnk ubout lt, und most of ull begln to develop lt
unulytlcully" (pp. 67-68).

Codlng 4uulltutlve 'utu
Once u cutegorlzutlon scheme hus been developed, ull of
the dutu ure then revlewed for content und coded for
correspondence to or exempllflcutlon of the ldentlfled
cutegorles. In u grounded theory study, the entlre lnltlul pro-
cess of breuklng down, cutegorlzlng, und codlng dutu ls
often referred to us open codlng.
The uctuul codes correspondlng to the cutegory system
ure urbltrury. They cun be un ubbrevlutlon of the cutegory
(e.g., %-ACT for
%oy's Actlvltles ln Gugllurdl's cutegory system), u code
correspondlng to un outllne structure (e.g., %oy's Actlvltles
mlght be A.l und Fumlly's Actlvltles mlght be A.2), or some
other type of symbollc conflgurutlon. When u computer ls
used to lndex the dutu, the softwure belng used often
speclfles requlrements for lubellng the codes.
Thls process of codlng quulltutlve muterlul ls seldom un
eusy one, for severul reusons. Flrst, the reseurcher muy
huve dlfflculty decldlng whlch code ls most upproprlute or
muy not fully comprehend the underlylng meunlng of some
uspect of the dutu. It muy tuke u second or thlrd reudlng of
the muterlul to grusp the nuunces contulned ln some
portlons of the dutu.
Second, the reseurcher often dlscovers ln golng through
the dutu thut the lnltlul cutegory system wus lncomplete or
otherwlse lnudequute. In muny cuses, thls meuns golng
buck und sturtlng from scrutch. For thls reuson, lt ls usuully
necessury to revlew u very lurge portlon of the dutu before
un udequute cutegorlzutlon scheme cun be developed. It ls
not unusuul for some toplcs to emerge thut were not lnltlully
conceptuullzed. When thls huppens, lt ls rlsky to ussume
thut the toplc fulled to uppeur ln muterluls thut huve ulreudy
been coded. Thut ls, u concept mlght not be ldentlfled us
sullent untll lt hus emerged three or four tlmes
Subject 025 June 25,
1980 Page 32
Int: How did you feel right after the separation? Will you tell
me a little more about that?
025: Well, you know, when 1 look back...I mean I think anybody
would have felt hopeless and helpless...because of, you
know, emotional1y...I think maybe it's a little easier i f 3b
you've got more security...more money. I was really stuck. ^ I caught myself
thinking...
M
0h, i f i t wasn't for the kids..." I love them both dearly, but...I
mean there have been times when I've said to myselfŹI think a lot of women 5c go
through this tooŹI really had thought of giving them up. In the beginning I used
to think, too, they'd be so much better o f f with somebody that could, you
know...When I
was really struggling-ŹI was on welfare ----------------1 used to think,
"My God, how am I going to educate them? What are they lb
going to have? I can't make ends meet now." It's kinda projecting the
unknown...fear o f the unknown. I think you can get mixed up.

)¡gXre 22-1. Coded excerpt from un XnstrXctXred ¡nterv¡ew.
¡n the dutu. In sXch u cuse, ¡t woXld be necessury to re-reud
ull prev¡oXsly coded muter¡ul to huve u trXly complete grusp
of thut cutegory.
Another problem stems from the fuct thut nurrut¡ve
muter¡uls ure generully not l¡neur. )or exumple, purugruphs
from trunscr¡bed ¡nterv¡ews muy contu¡n elements relut¡ng
to three or foXr d¡fferent cutegor¡es, embedded ¡n u complex
3urt V. The AnuIys¡s oI Peseurch Dutu

6
Iush¡on. An exumpIe oI u muIt¡top¡c segment oI un
¡nterv¡ew Irom u study oI d¡vorced women by one oI the
uuthors, w¡th codes ¡n the murg¡n, ¡s shown ¡n F¡gure 22-1.1

Methods oI Orgun¡z¡ng
4uuI¡tut¡ve Dutu
Trud¡t¡onuIIy, quuI¡tut¡ve dutu huve been orgun¡zed
munuuIIy through u vur¡ety oI techn¡ques. AIthough these
munuuI methods huve u Iong und respected h¡story, they
ure becom¡ng ¡ncreus¡ngIy outmoded us u resuIt oI the
w¡despreud uvu¡Iub¡I¡ty oI personuI computers thut cun be
used to perIorm the I¡I¡ng und ¡ndex¡ng oI quuI¡tut¡ve
muter¡uI. Computer progrums Ior munug¡ng quuI¡tut¡ve dutu
ure descr¡bed ¡n Chupter 23. Here, we br¡eIIy descr¡be
some munuuI methods oI dutu orgun¡zut¡on und
munugement.
When the umount oI dutu ¡s smuII, or when u cutegory
system ¡s Iu¡rIy s¡mpIe, reseurchers somet¡mes use coIor
puper cI¡ps or coIor 3ost-It Notes to code the content oI the
nurrut¡ve muter¡uIs. For exumpIe, ¡I we were unuIyz¡ng re-
sponses to u s¡ngIe unstructured quest¡on ubout women's
utt¡tudes towurd the menopuuse, we m¡ght use bIue puper
cI¡ps Ior comments reIut¡ng to the top¡c oI Ioss oI
reproduct¡ve cupuc¡ty, red cI¡ps Ior comments reIut¡ng to
phys¡cuI s¡de eIIects such us hot IIushes, yeIIow cI¡ps Ior
comments reIut¡ng to ug¡ng, und so on. Then the reseurcher
couId puII out uII oI the responses coded w¡th u certu¡n
coIor cI¡p to exum¡ne one uspect oI menopuusuI utt¡tudes ut
u t¡me.
%eIore the udvent oI computer progrums Ior munug¡ng
quuI¡tut¡ve dutu, the most usuuI procedure wus the
deveIopment oI conceptuuI I¡Ies. In th¡s upprouch, u
phys¡cuI I¡Ie ¡s deveIoped Ior euch oI the vur¡ous cutegor¡es,
und uII oI the muter¡uIs reIut¡ng to thut top¡c ure cut out und
¡nserted ¡nto the I¡Ie. To creute conceptuuI I¡Ies, the
reseurcher must I¡rst go through uII oI the dutu und wr¡te the
reIevunt codes ¡n the murg¡ns, us ¡n F¡gure 22-1. Then the
reseurcher cuts up u copy oI the muter¡uI by cutegory ureu,
und pIuces the cut-out excerpt ¡nto the I¡Ie Ior thut cutegory.
In th¡s Iush¡on, uII oI the content on u purt¡cuIur top¡c cun be
retr¡eved by go¡ng to the uppI¡cubIe I¡Ie IoIder.
The creut¡on oI such conceptuuI I¡Ies ¡s cIeurIy u
cumbersome und Iubor-¡ntens¡ve tusk. Th¡s ¡s purt¡cuIurIy
true when segments oI the nurrut¡ve muter¡uIs huve muIt¡pIe
codes, us ¡s true ¡n the excerpt shown ¡n F¡gure 22-1. In

1F¡gure 22-1 does not show the cutegor¡es thut the uIphunumer¡c
codes ¡n the murg¡n represent. The code cutegor¡es used ¡n the
I¡gure ure us IoIIows. 2bդGeneruI psychoIog¡c stute dur¡ng the
d¡vorce; 9դF¡nunces; 5cդFeeI¡ngs ubout s¡ngIe purent¡ng; und
IbդD¡vorce-¡nduced probIems.
such u s¡tuut¡on, there wouId need to be Iour cop¡es oI the
purugruphդone Ior euch I¡Ie correspond¡ng to the Iour
codes. The reseurcher must uIso be sens¡t¡ve to the need to
prov¡de enough context thut the cut-up muter¡uI cun be
understood. For exumpIe, ¡t m¡ght be necessury to ¡ncIude
muter¡uI preced¡ng or IoIIow¡ng the d¡rectIy reIevunt
muter¡uIs. F¡nuIIy, the reseurcher must usuuIIy ¡ncIude
pert¡nent udm¡n¡strut¡ve ¡nIormut¡on on euch ¡tem ¡n the
conceptuuI I¡Ies. For exumpIe, ¡I the dutu cons¡sted oI
trunscr¡bed ¡nterv¡ews, euch ¡nIormunt wouId ord¡nur¡Iy be
uss¡gned un ¡dent¡I¡cut¡on (ID) number. Euch excerpt I¡Ied
¡n the conceptuuI I¡Ie wouId uIso need to ¡ncIude the
uppropr¡ute ID number so thut the reseurcher couId, ¡I
necessury, obtu¡n udd¡t¡onuI ¡nIormut¡on Irom the muster
copy.
In I¡eu oI conceptuuI I¡Ies, some reseurchers use
quuI¡tut¡ve sort curds (uIso known us Mc%ee curds) Ior
¡ndex¡ng the¡r dutu. These ¡ndex curds huve hoIes on uII
s¡des thut cun be used Ior eusy retr¡evuI oI ¡nIormut¡on on u
g¡ven top¡c or concept. Thut ¡s, the reseurcher pIuces
notches ¡n the hoIes correspond¡ng to coded top¡cuI
cutegor¡es or concepts. Then, when ¡nIormut¡on reIut¡ng to
u purt¡cuIur top¡c or concept ¡s needed, uII oI the curds
notched Ior thut cutegory cun eus¡Iy be puIIed Irom the deck
oI ¡ndex curds. Some reseurchers d¡rectIy puste the¡r dutu
on these curds, when the dutu set ¡s oI u munugeubIe s¡ze,
und then d¡rectIy code euch dutu segment us¡ng the notched
hoIes to ¡nd¡cute u code. Others wr¡te un ubstruct oI un
ent¡re cuse (e.g., one ¡nterv¡ew or one observut¡onuI
sess¡on) onto the curd, und then ¡nd¡cute on the curd the
puge numbers (e.g., puges oI the observut¡onuI notes or
trunscr¡bed ¡nterv¡ew) where d¡IIerent codes muy be Iound.
CIeurIy, such munuuI methods ure Iubor ¡ntens¡ve.
Computer progrums remove the drudgery oI cutt¡ng und
pust¡ng puges und puges oI nurrut¡ve muter¡uI und ure Iust
becom¡ng ¡nd¡spensubIe reseurch tooIs. However, some
peopIe urgue thut beg¡nn¡ng reseurchers shouId gu¡n ut
Ieust some exper¡ence w¡th munuuIIy orgun¡z¡ng quuI¡tut¡ve
dutu. 3urt oI th¡s urgument stems Irom the sent¡ment thut
munuuI ¡ndex¡ng prov¡des u more hunds-on type oI
exper¡ence thut uIIows the reseurcher to get cIoser to the
dutu. Other concerns huve been expressed ubout the use oI
computers w¡th quuI¡tut¡ve dutu. For exumpIe, one potent¡uI
p¡tIuII ¡s thut the reseurcher m¡ght deveIop un overIy
eIuborute cod¡ng system, resuIt¡ng Irom the Iuct thut
retr¡evuI oI ¡nIormut¡on ¡s vustIy s¡mpI¡I¡ed w¡th u computer.
The udvent oI soph¡st¡cuted progrums thut u¡d ¡n the
unuIys¡s oI dutu hus uIso g¡ven r¡se to ob|ect¡ons to huv¡ng u
process thut ¡s bus¡cuIIy cogn¡t¡ve turned ¡nto un uct¡v¡ty
thut ¡s mechun¡cuI und techn¡cuI. Desp¡te these concerns,
most quuI¡tut¡ve reseurchers huve sw¡tched to computer¡zed
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 7

munugement of thelr dutu, und thls trend ls unllkely to be
reversed. Computerlzed dutu munugement frees up the
reseurcher's tlme so thut greuter uttentlon cun be puld to
more lmportunt conceptuul lssues.

B ANALYTIC PROCEDURES

'utu munugement tusks ln quulltutlve reseurch ure typlcully
reductlonlst ln nuture becuuse they convert lurge musses of
dutu lnto smuller, more munugeuble segments. %y contrust,
quulltutlve dutu unulysls tusks ure constructlonlst ln nuture:
They lnvolve puttlng together segments lnto u meunlngful
conceptuul puttern. Although dlfferent upprouches to
quulltutlve dutu unulysls huve been udvocuted, there ure
some elements thut ure common to severul of them. We
provlde some generul guldellnes, followed by u descrlptlon
of the unulytlc procedures used ln two speclflc types of
upprouch.

A Generul Anulytlc Overvlew
The unulysls of quulltutlve muterluls generully beglns wlth u
seurch for themes or recurrlng regulurltles. In muny cuses,
the themutlc unulysls beglns ln the fleld us the dutu ure
belng collected. In other sltuutlons, the themutlc unulysls
occurs ufter the dutu huve been collected, durlng u reudlng
(or re-reudlng) of the dutu set.
Themes often develop wlthln cutegorles of dutu (l.e.,
wlthln cutegorles of the codlng scheme used for lndexlng
muterluls) but sometlmes cut ucross them. For exumple, ln
Gug-llurdl's (1991) study, slx themes descrlblng the
fumllles' experlences were ldentlfled, und these themes
were further grouped under three heudlngs correspondlng
to the stuges ln the process of uduptlng to the chlld's
dlsublllty. The flrst theme, dlsllluslonment, embruced
content thut hud been coded under severul toplcul codes
(see %ox 22-1), prlmurlly the toplcs wlthln the feellng
cutegory.
The seurch for themes lnvolves not only the dlscovery
of commonulltles ucross sub|ects but ulso u seurch for
nuturul vurlutlon ln the dutu. Themes thut emerge from
unstructured observutlons und lntervlews ure never
unlversul. The reseurcher must uttend not only to whut
themes urlse but ulso to how they ure putterned. 'oes the
theme upply only to certuln subgroups? In certuln types of
communltles or orgunlzutlons? In certuln contexts? At
certuln perlods? Whut ure the condltlons thut precede the
observed phenomenon, und whut ure the uppurent
consequences of lt? In other words, the quulltutlve unulyst
must be sensltlve to relutlonshlps wlthln the dutu.
The unulyst's seurch for themes, regulurltles, und
putterns ln the dutu cun sometlmes be fucllltuted by churtlng
devlces thut enuble the reseurcher to summurlze the
evolutlon of behuvlors, events, und processes. For exumple,
for quulltutlve studles thut focus on dynumlc
experlencesդsuch us declslon muklngդlt ls often useful to
develop flow churts or tlmellnes thut hlghllght tlme
sequences, mu|or declslon polnts und events, und fuctors
uffectlng the declslons. An exumple of such u flow churt
from u study of declslon muklng umong lnfertlle couples ls
presented ln Flgure 22-2. The constructlon of such flow
churts for ull sub|ects would help to hlghllght uny
regulurltles ln the sub|ects' evolvlng behuvlors.
A further step frequently tuken lnvolves the vulldutlon of
the understundlngs thut the themutlc explorutlon hus
provlded. In thls phuse, the concern ls whether the themes
lnferred ure un uccurute representutlon of the perspectlves
of the people lntervlewed or observed. Severul procedures
cun be used ln thls vulldutlon step, some of whlch were
dlscussed ln Chupter 17. If there ls more thun one
reseurcher worklng on the study, debrleflng sesslons ln
whlch the themes ure revlewed und speclflc cuses dls-
cussed cun be hlghly productlve. Multlple per-
spectlvesդwhut we referred to ln Chupter 17 us
lnvestlgutor trlungulutlonդcunnot ensure the vulldlty of the
themes, but lt cun mlnlmlze uny ldlosyncrutlc bluses. Uslng
un lterutlve upprouch ls ulmost ulwuys necessury. Thut ls,
the reseurcher derlves themes from the nurrutlve muterluls,
goes buck to the muterluls wlth the themes ln mlnd to see lf
the muterluls reully do flt, und then reflnes the themes us
necessury. It ls generully useful to undertuke member
checksդthut ls, to present the prellmlnury themutlc
unulysls to some of the sub|ects or lnformunts, who cun be
encouruged to offer suggestlons thut mlght support or
contrudlct thls unulysls.
It ls ut thls polnt thut some reseurchers lntroduce
quusl-stutlstlcsդu tubulutlon of the
D.N.
(wife)
marries

D.N.
discontinues
oral
contraceptives

D.N
consults
Dr.P
Fertility
drugs
D.N.
consults
Dr.P

Temperature
readings

A.N.
D.N. tests
performed;
problem
identified
D.N.
surgery |¤j
performed
D.N.
Actions



Joint
Actions
A.N. and
D.N. consult
Dr.C
D.N. and
A.N initiate
adoption
inquiry
D.N. and
A.N. trip to
major
infertility
clinic

A.N.
Actions
A.N.
marries
DN.
A.N. tests
performe
d
A.N.
negative
test
results

June May Jan. July
Aug.
1989 1992 1994 1994
1994
)¡gXre 22-2. Exumple of u t¡mel¡ne for ¡nfert¡l¡ty stXdy.
Sept.
1994
Oct. 1994
Jan. Sept.
1995
1995
Oct. 1995
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 9

frequency wlth whlch certuln themes, relutlons, or lnslghts
ure supported by the dutu. The frequencles cunnot be
lnterpreted ln the sume wuy us frequencles generuted ln
survey studles, becuuse of lmpreclslon ln the sumpllng of
cuses und enumerutlon of the themes. Nevertheless, us
%ecker (1970) polnted out,
4uusl-stutlstlcs muy ullow the lnvestlgutor to dlspose of
certuln troublesome null hypotheses. A slmple
frequency count of the number of tlmes u glven
phenomenon uppeurs muy muke untenuble the null
hypothesls thut the phenomenon ls lnfrequent. A
compurlson of the number of such lnstunces wlth the
number of negutlve cusesդlnstunces ln whlch some
ulternutlve phenomenon thut would not be predlcted by
hls theory uppeursդmuy muke posslble u stronger
concluslon, especlully lf the theory wus developed eurly
enough ln the observutlonul perlod to ullow u systemutlc
seurch for negutlve cuses. Slmllurly, un lnspectlon of the
runge of sltuutlons covered by the lnvestlgutor's dutu
muy ullow hlm to negute the hypothesls thut hls conclu-
slon ls restrlcted to only u few sltuutlons, tlme perlods,
or types of people ln the orgunlzutlon or communlty, (p.
81)
In the flnul stuge of unulysls, the reseurcher strlves to
weuve the themutlc pleces together lnto un lntegruted
whole. The vurlous themes need to be lnterreluted ln u
munner thut provldes un overull structure (such us u theory
or lntegruted descrlptlon) to the entlre body of dutu. The
lntegrutlon tusk ls un extremely dlfflcult one becuuse lt
demunds creutlvlty und lntellectuul rlgor lf lt ls to be
successful. A strutegy thut sometlmes helps ln thls tusk ls to
cross-tubulute dlmenslons thut huve emerged ln the
themutlc unulysls. For exumple, %urton (1991), ln her study
of purents' uduptutlons to thelr udolescent chlldren's drug
ubuse problems, found thut purentul power und purentul
responslblllty were two lmportunt dlmenslons thut, when
cross-tubuluted, reflected lmportunt coplng putterns umong
the purents ln her sumple. Her dlugrum dlspluylng the
cross-tubuluted schemu ls presented ln Flgure 22-3.

Anulytlc Inductlon
The generul procedures und steps prevlously descrlbed
provlde u generul outllne of how quulltutlve reseurchers
muke sense of thelr dutu und dlstlll from them thelr
understundlngs of processes und behuvlors operutlng ln
nuturullstlc settlngs. However, there ure some vurlutlons ln
the gouls und underlylng phllosophles of quulltutlve
reseurchers thut ulso leud to vurlutlons ln how the unulytlc
tusk ls hundled. One of the strutegles for unulyzlng
quulltutlve dutu ls referred to us unu-


The parents feel guilty despite
professionals' attempts to
eliminate guilt
Style is "committed"/
"correcting the flaw"
The parents are especially
"fused" to the child
Style is "stuckTsecondary
gains" or "protection from
social responsibility"
The parents accept not feeling
responsible and have power.
Style is "successful"/
"mastery"
The parents accept not feeling
responsible but let go of
power.
Style is "letting go*'/"new
meaning"

+ POWER
)¡gXre 22-3. %urton's (1991) schemu descr¡b¡ng purentul uduptut¡on to udolescent drXg
ubXse (repr¡nted w¡th perm¡ss¡on).
RESPONSIBILITY
3urt V. The AnuIys¡s oI Peseurch Dutu

10
Iyt¡c ¡nduct¡on, un upprouch thut wus deveIoped ¡n the
1930s.
The unuIyt¡c ¡nduct¡on upprouch requ¡res u cureIuI
scrut¡ny oI uII the reseurcher's dutu, usuuIIy uccord¡ng to the
IoIIow¡ng s¡x steps.
1. 'eflne the phenomenon to be studled und explulned.
2. %used on u revlew of the dutu, formulute u
hypothetlcul explunutlon of the phenomenon; thut ls,
develop un lnductlvely derlved hypothesls.
3. Conduct un lntenslve unulysls of lndlvlduul cuses to
see whether the hypothesls flts purtlculur cuses.
4. Seurch for negutlve cuses thut, lf found, leud to u
reformulutlon of the hypothesls or u redeflnltlon of
the phenomenon to exclude the purtlculur cuse.
5. Contlnue the exumlnutlon of cuses, redeflnltlon of
the phenomenon, und reformulutlon of hypotheses
untll u unlversul puttern of relutlonshlps ls
estubllshed.
6. Creute u hlgher level of ubstructlon or
conceptuullzutlon through compurlson wlth other
settlngs or groups.
In pructlce, the use of unulytlc lnductlon results ln u
procedure thut ulternutes buck und forth between tentutlve
explunutlon und tentutlve deflnltlon, euch reflnlng the other
so thut u sense of closure cun be uchleved when un lntegrul
relutlon between the two ls estubllshed. As un exumple from
the nurslng reseurch llteruture, Muyo (1992) conducted un
ethnogruphlc study of the processes of munuglng physlcul
uctlvlty umong Afrlcun Amerlcun worklng women; she used
unulytlc lnductlon us the method of unulyzlng her quulltutlve
dutu.

Grounded Theory Anulysls
%y fur the most wldely used unulytlc upprouch umong nurse
reseurchers ls Gluser und Struuss's (1967) grounded theory
method of generutlng theorles from dutu. %ecuuse of lts
populurlty, we descrlbe here some of the speclflc unulytlc
technlques of grounded theory.
As we huve seen, the flrst step ls the development of u
cutegorlzutlon scheme und the substuntlve codlng of dutu
through the prellmlnury process referred to us open codlng
(or, sometlmes, level I codlng). Open codlng ullows the
reseurcher to ldentlfy mu|or cutegorles und subcutegorles
und to descrlbe thelr mu|or propertles und dlmenslons.
Level II codlng (often referred to us uxlul codlng) ls u
reconstructlve process thut puts the dutu buck together ln
new wuys ufter open codlng ls completed by connectlng u
cutegory und lts subcutegorles. Constunt compurlson ls used
uguln ln uxlul codlng to reflne theoretlcul cutegorles. Struuss
und Corbln (1990) noted thut, ln uxlul codlng, the "focus ls
on speclfylng u cutegory ,phenomenon) ln terms of the
condltlons thut glve rlse to lt; the context (lts speclflc set of
propertles ln whlch lt ls embedded; the uctlon/lnteructlon
strutegles by whlch lt ls hundled, munuged, currled out; und
the consequences of those strutegles" (p. 97). Axlul codlng
ls u complex process lnvolvlng both lnductlve und deductlve
thlnklng und ls deslgned to move the codlng process to u
hlgher level of ubstructlon.
Throughout the codlng und unulysls process, the
grounded theory unulyst documents hls or her ldeus,
lnslghts, und feellngs ubout the dutu, themes, und the
emerglng conceptuul scheme ln the form of memos. Morse
und Fleld (1995) note thut the memos serve numerous
functlons ln u grounded theory study. For exumple, memos
help the reseurcher to ldentlfy underlylng ussumptlons. They
ulso preserve ldeus thut muy lnltlully not seem productlve
but muy luter prove vuluuble once further developed. An
especlully lmportunt functlon ls thut memos encouruge u
hlgher level of ubstruct thlnklng by gettlng the reseurcher to
reflect on und descrlbe themes und putterns ln the dutu,
relutlonshlps between cutegorles, und emergent
conceptuullzutlons.
Once suturutlon hus been uchleved und both open
codlng und uxlul codlng ure completed, lntegrutlon cun
begln. Here, the unulyst revlews und sorts the memos to
explore
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 11

theoretlcul explunutlons und seurches for the core
cutegoryդthe centrul phenomenon thut ls used to lntegrute
ull other cutegorles. Level III codlng (ulso referred to us
selectlve codlng) ls the process of selectlng the core cute-
gory, systemutlcully ldentlfylng und lntegrutlng relutlonshlps
between the core cutegory und other cutegorles, und
vulldutlng those relutlonshlps. The lntegrutlve process ls
slmllur to uxlul codlng, except thut lt ls done ut u more
hlghly ubstruct level of unulysls.
Level III codlng results ln u descrlptlon of u buslc soclul
process (%SP), whlch ls the centrul soclul process resultlng
from the dutu. %ecuuse of the uctlvltles ln whlch the
reseurcher hus enguged, the emergent theory of whut thls
process ls und how lt works ls grounded ln the dutu.
The grounded theory method ls concerned wlth the
generutlon of cutegorles, propertles, und hypotheses ruther
thun wlth testlng them
(us ln the cuse of unulytlc lnductlon). The product of the
typlcul grounded theory study ls u conceptuul or theoretlcul
model thut endeuvors to expluln the phenomenon under
study. As un exumple, Flgure 22-4 presents the model
developed by Klng und Jensen (1994) ln thelr grounded
theory study thut conceptuullzed the process of women's
"preservlng the self" durlng curdluc surgery.
There ure some obvlous slmllurltles between the
grounded theory und unulytlc lnductlon upprouches, but
there ure ulso some lmportunt dlfferences. Of purtlculur
lmportunce ls thelr overull ulms. Anulytlc lnductlon ls
concerned wlth the testlng of lnductlvely derlved hypothe-
ses; lt ls purported to be u method for comlng to terms wlth
the problem of cuusul lnference whlle remulnlng fulthful to
quulltutlve, nuturullstlc dutu. The grounded theory method ls
concerned wlth the generutlon of cutegorles, prop-

relating
OS
e
"co
distrustin
g
trusting


11
s I


CO
c 2
flounderin
g for
control
losing
control
changing
normal



evolving
normal
relinquishi
ng control
attaini
ng/
regaining
control

I*

c >
3
CÛ

Eri?


o 3 3 o <


redefining
normal



making sense


managing managing
3urt V. The AnuIys¡s oI Peseurch Dutu

12
I
ll
I
c o > o E






)¡gXre 22-4.
K¡ng und Jen-
sen's (1994)
model of
preserv¡ng
the self
dXr¡ng
curd¡uc
sXrgery
(repr¡nted
w¡th per-
m¡ss¡on).
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 13

ertles, und hypotheses ruther thun wlth testlng them. The
product of muny grounded theory studles ls u conceptuul or
theoretlcul model thut endeuvors to expluln the
phenomenon under study.

!5INTERPRETATION OF
QUALITATIVE FINDINGS

In quulltutlve studles, lnterpretutlon und unulysls of the dutu
occur vlrtuully slmultuneously. Thut ls, the reseurcher
lnterprets the dutu us he or she cutegorlzes lt, develops u
themutlc unulysls, und lntegrutes the themes lnto u unlfled
whole. Efforts to vulldute the quulltutlve unulysls ure
necessurlly efforts to vulldute the lnterpretutlon us well.
Thus, ln quulltutlve studles, the meunlng of the dutu flows
from the unulysls.
Nevertheless, the prudent quulltutlve reseurcher holds
hls or her lnterpretutlon up for closer scrutlnyդself-scrutlny
us well us revlew by peers und outslde revlewers. It mlght
be noted thut even when the reseurcher hus undertuken
procedures such us member checks und peer debrleflngs,
thls ln ltself does not constltute proof thut the results und
lnterpretutlons ure credlble. For exumple, ln member
checks, muny purtlclpunts mlght be too pollte to dlsugree
wlth the reseurcher's lnterpretutlons, or they muy become
lntrlgued wlth u conceptuullzutlon thut they themselves
would never huve developed on thelr ownդu
conceptuullzutlon thut ls not necessurlly uccurute. Thus, for
quulltutlve reseurchers us well us quuntltutlve reseurchers, lt
ls lmportunt to conslder posslble ulternutlve explunutlons for
the flndlngs und to tuke lnto uccount methodologlc or other
llmltutlons thut could huve uffected the study results.
In druwlng concluslons, the quulltutlve reseurcher should
ulso conslder the trunsferublllty of the flndlngs. Thut ls,
whereus quulltutlve reseurchers rurely seek to muke
generullzutlons, they often strlve to develop un under-
stundlng of how the study results cun be usefully upplled.
The centrul questlon ls, In whut other types of settlngs und
contexts could one expect the phenomenu under study to be
munlfested ln u slmllur fushlon?
The lmpllcutlons of the flndlngs of quulltutlve studles, us
wlth quuntltutlve ones, ure often multldlmenslonul. Flrst,
there ure lmpllcutlons for further reseurch: Should the study
be repllcuted ln unother study? Cun the study be expunded
(or clrcumscrlbed) ln meunlngful und productlve wuys? 'o
the results suggest thut un lmportunt construct hus been
ldentlfled thut merlts the development of u more formul
lnstrument? 'oes the emerglng theory suggest hypotheses
thut could be tested through more controlled, quuntltutlve
reseurch? Second, do the flndlngs huve lmpllcutlons for
nurslng pructlce? For exumple, could the heulth cure needs
of u subculture (e.g., the homeless) be ldentlfled und
uddressed more effectlvely us u result of the study? Flnully,
do the flndlngs shed llght on the fundumentul processes thut
ure lncorporuted lnto nurslng theory?

B ILLUSTRATION OF
QUALITATIVE ANALYSIS
ACTIVITIES

4uulltutlve reseurchers seldom dlscuss ln uny detull the
wuys ln whlch they unulyze thelr dutu. However, the steps
undertuken by u reseurch teum thut studled how the role of
cllnlcul nurse reseurchers wus belng deflned, enucted, und
evuluuted wus descrlbed by Knufl und Webster (1988). Thls
study lnvolved ln-depth telephone lntervlews wlth 34 cllnlcul
nurse reseurchers (CN5s) und thelr correspondlng chlef
nurse executlves (CNEs). The dutu conslsted of ubout 1000
puges of lntervlew trunscrlpts.
The reseurchers undertook slx mu|or tusks durlng the
dutu munugement phuse und flve mu|or tusks durlng the
dutu unulysls phuse:
3huse I: Dutu Munugement
1. Independent reudlng of u sumple of lntervlew
trunscrlpts by teum members; colluborutlve
development of slx mu|or
descrlptlve codlng cutegorles; development of u
codebook to ensure conslstent uppllcutlon of flnul
codlng cutegorles
2. Independent codlng of ull lntervlews ln murglns of
trunscrlpts by two coders; colluborutlve revlew of
codes und resolutlon of dlfferences ln codlng
3. Cut-und-puste trunsfer of ull dutu onto color-coded
5- X 8-lnch lndex curds, wlth colors correspondlng
to the slx mu|or cutegorles
4. Identlflcutlon of subcutegorles reflectlng nurrower
toplcul ureus wlthln the mu|or cutegorles
5. Codlng of the lndex curds uccordlng to the
subcutegorles ldentlfled
6. Constructlon of descrlptlve summury grlds for euch
subcutegory ucross ull sub|ects, wlth euch cell of the
grld contulnlng u brlef summury of pertlnent content
3huse II. Dutu Anulysls
1. Descrlptlon of the content of euch subcutegory und
ldentlflcutlon of mu|or themes; prepurutlon of u
wrltten unulysls summury
2. Descrlptlon of the content of euch mu|or cutegory,
constructed from the subcutegory summury
3. Descrlptlon of the CN5 group, syntheslzlng und
lntegrutlng relevunt muterlul
4. Descrlptlon of the CNE group, syntheslzlng und
lntegrutlng relevunt muterlul
5. Descrlptlon of CN5-CNE pulrs, contrustlng of
wlthln-pulr themes

3urt V. The AnuIys¡s oI Peseurch Dutu

14
B TIPS FOR ANALYZING
QUALITATIVE DATA

4uulltutlve unulysls ls u chullenglngդbut often u very
rewurdlngդenterprlse. In thls sectlon, we offer u few
uddltlonul tlps to beglnnlng students regurdlng the conduct
of u quulltutlve unulysls.
ձ In developlng u cutegory system for codlng quulltutlve
dutu, u substuntlul sumple of the dutu should be curefully
reud before the scheme ls flnullzed. To the extent
posslble, the reseurcher should try to reud muterluls thut
vury ulong lmportunt dlmenslons, ln un effort to cupture
the full runge of content. The dlmenslons mlght lnclude
lnformunt churucterlstlcs (e.g., trunscrlpts of men versus
women, urtlculute versus less urtlculute lnformunts) or
uspects of the dutu collectlon experlence (e.g., shorter
versus longer lntervlews, observutlons ln dlfferent types
of settlngs). If the reseurcher cun ldentlfy ln udvunce
whut the lmportunt dlmenslons mlght be, methods should
be used to fucllltute ldentlflcutlon of good exemplurs
(e.g., by uppendlng wrltten comments on the relevunt
dlmenslons ut the end of un lntervlew).
ձ Even when u cutegory system ls "flnullzed," the
reseurcher should be prepured to umend or even totully
revlse lt. Even slmple chunges muy meun re-reudlng ull
prevlously coded muterlul. Muklng chunges mldwuy ls
often pulnful und frustrutlng, but wlthout u hlghly
effectlve cutegory system, the reseurcher wlll be unuble
to ldentlfy und lntegrute lmportunt themes udequutely.
ձ In most studles, lt ls wlse to develop u codebook, thut ls,
wrltten documentutlon descrlblng the exuct deflnltlon of
the vurlous cutegorles used to code the dutu. Good
quulltutlve codebooks usuully lnclude two or three uctuul
excerpts thut typlfy muterluls coded ln euch cutegory.
ձ Even lf you huve llttle experlence uslng u computer, the
uvulluble progrums for quulltutlve dutu munugement ure
relutlvely eusy to leurn und should not be u cuuse of
unxlety. In selectlng u progrum (see Chupter 23), try to
dlscuss the merlts of vurlous progrums wlth other
reseurchers who huve experlence uslng them.
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu

15
ձ Muny reseurchers enguge ¡n qXuI¡tut¡+e stXd¡es on the¡r
own ruther thun w¡th u co-¡n+est¡gutor. Howe+er,
coIIuborut¡+e eIIorts ure becom¡ng more common, us
|oXrnuI ed¡tors und peers hu+e come to demund thut
greuter uttent¡on be pu¡d to +uI¡dut¡ng qXuI¡tut¡+e dutu
und the meun¡ngs ussoc¡uted w¡th them. Hu+¡ng one
or more teum member ensXres thut the dutu w¡II not be
unuIyzed und ¡nterpreted ¡n u totuIIy sXb|ect¡+e
Iush¡on.

B RESEARCH EXAMPLES

The number of quulltutlve studles thut huve been publlshed
ln nurslng |ournuls durlng the pust decude hus rlsen
drumutlcully. Two reseurch exumples ure descrlbed below.

5eseurch Exumple of u Grounded
Theory Anulysls

%rlght (1992). studled the blrth of u flrst chlld us un
lntergenerutlonul experlence wlth the ulm of dlscoverlng
the buslc soclul process ussocluted wlth thls event und
generutlng u theory on normutlve fumlly process. She
collected ln-depth lntervlew und observutlonul dutu over
u 15-month perlod from three fumllles: three flrst-born
lnfunts und thelr purents und slx sets of grundpurents.
'utu collectlon begun durlng the lust trlmester of
pregnuncy und contlnued untll the chlld's flrst blrthduy.
The fumllles dlffered ln terms of whether the pregnuncy
wus plunned und wunted.
Wlthln euch fumlly, the lntervlews were lnltlully
conducted ln u group settlng, whlch ullowed the
reseurcher to observe fumlly lnteructlonul putterns und
to leurn ubout fumlly bellefs und customs us well us
expectutlons ubout the untlclputed blrth event.
Subsequent lntervlews were conducted wlth lndlvlduul
fumlly members, wlth dyuds, und wlth smull groups us
deemed needed on the busls of the developlng
hypotheses. Intervlews und observutlons occurred ln
purentul und grund-purentul homes, on the hospltul
muternlty wurd, und ut other locutlons, such us bup-
tlsmul ceremonles. Telephone lntervlews were used to
clurlfy dutu und to check on the vulldlty of the emerglng
hypotheses.
The unulysls of dutu wus done us un ongolng process,
lntegruted wlth dutu collectlon und codlng. Intervlews
were uudlotuped, trunscrlbed, und then coded llne by
llne. One three-generutlon fumlly lnteructlon wus vldeo-
tuped to provlde recorded nonverbul us well us verbul
dutu for codlng. Coded dutu were grouped lnto reluted
cutegorles und then compured wlth one unother und wlth
new dutu to reflne und dlscurd emerglng hypotheses con-
tlnuully. Memorundu were prepured concurrently to
record the reseurcher's theoretlcul unulysls of the dutu.
Severul methods were used to vulldute the themes
emerglng from the unulysls, lncludlng revlew by
members of the fumllles.
The unulysls resulted ln the ldentlflcutlon of un
evolutlonury fumlly process thut reflected reorgunlzed
lnterpersonul putterns. "Muklng pluce," the centrul soclul
process thut emerged from %rlght's unulysls, wus
deflned us the fumlly process through whlch u newborn
lndlvlduul recelves recognltlon us u member of thut
fumlly. The fumlly mude pluce both physlcully und
soclully for the new und expunded relutlonshlps creuted
by the chlld's blrth.


5eseurch Exumple of un Anulysls for un
Ethnogruphlc Study

5ussell (1996) conducted un ethnogruphlc study thut
focused on the cure-seeklng process used by elders
resldlng ln u contlnulng-cure retlrement communlty.
'urlng 8 months of fleld work ln the retlrement
communlty, 5ussell collected dutu on the elders"
process of ellcltlng cure und enguglng cureglvers ln cure
lnteructlons. 'utu were collected through ln-depth
semlstructured lntervlews wlth 12 elders, purtlclpunt
observutlon ln the fuclllty's udult cure center where 34
elders were observed, und u
16 3urt V. The AnuIys¡s oI Peseurch Dutu

focus group lntervlew wlth four elders. Euch of the dutu
collectlon methods ylelded dlfferent perspectlves ubout
the purtlclpunts' cure seeklng. The lntervlews und focus
group sesslons were tuped und fully trunscrlbed, und the
reseurcher took fleld notes ufter observutlonul sesslons.
'utu collectlon und dutu unulysls were done
concurrently. The dutu were entered lnto computer flles
for orgunlzutlon und munugement.
The flrst step lnvolved compurlng und contrustlng
dutu elements from dlfferent sources to generute u
cutegorlzutlon scheme. The result wus the development
of u codebook wlth more thun 150 codes und deflnltlons.
The codes were then orgunlzed und reorgunlzed lnto
lurger unlts thut enubled the reseurcher to ldentlfy
putterns, feutures, sequences, und relutlonshlps ln elders'
cure seeklng. Fleld notes und memos were mulntulned to
document new ldeus ubout the unulysls, lncludlng both
theoretlcul und methodologlc lnslghts.
The buslc unlt of unulysls wus u speclflc cure
encounter. More thun 275 such unlts were demurcuted ln
the textuul dutubuse. Euch cuse wus churucterlzed ln
terms of certuln structurul elements, such us uctors,
contexts, und behuvlors. These cuses were used to de-
velop und test un emerglng conceptuullzutlon of the
cure-seeklng process.
Accordlng to 5ussell's unulysls, the cure-seeklng
process comprlses two phuses, whlch she lubeled "Cure
Ellcltlng" und "Cure Enguglng." Flgure 22-5 shows
5ussell's dlugrum of the cure-seeklng process. The
Inltlutlng Stuge, whlch begun when elders declded thut
usslstunce wus needed, wus shuped by the elders'
preferences (l.e., thelr deslres und expectutlons relutlng
to cure) und thelr bellefs ln terms of the upproprluteness of u
potentlul cureglver. Stuge 2, the Alertlng Stuge, lnvolved
both verbul und nonverbul strutegles deslgned to ulert
others of thelr deslre for usslstunce. Stuge 3, the
Negotlutlng Stuge, begun when u cureglver become
enguged ln un elder's cure. The negotlutlons between the
elder und cureglver vurled ln terms of content, hlstory of the
relutlonshlp, und "dolng"դthe extent to whlch elders dld for
themselves or worked wlth cureglvers to uccompllsh thelr
cure. Flnully, stuge 4 wus the Evuluutlon Stuge, the perlodlc
upprulsul thut the elders undertook to muke further
declslons regurdlng the cure tuklng. Thls stuge wus
churucterlzed by the elders' perceptlons ubout the cure
settlngs, cureglvers, und the cure ltself. 5ussell concluded
thut the flndlngs supported vlewlng elders' cure seeklng us
un lnteructlonul und developmentul process.


S SUMMARY
4uulltutlve unulysls ls u more chullenglng und
lubor-lntenslve uctlvlty thun quuntltutlve unulyslsդund one
thut ls gulded by fewer stundurdlzed rules. Although there
ure no unlversully udopted strutegles for unulyzlng quull-
tutlve dutu, four prototyplcul styles huve been ldentlfled thut
full on u contlnuum from ob|ectlve und systemutlc to
lnterpretlve und sub|ectlve: ( 1) u quusl-stutlstlcul style thut
beglns

CARE-ELICITING PHASE CARE-ENGAGING PHASE
1.Initiating Stage 3. Negotiating Stage
Preferences Content
Beliefs History
Doing
2. Alerting Stage 4. Evaluating Stage
Verbal Care Settings
Nonverbal Caregivers
Care
)¡gXre 22-5. 5Xssell's (1996) d¡ugrum of
seqXences ¡n the cure-seek¡ng process.
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 17

wlth u preestubllshed codebook of themes or words und thut
lends ltself to buslc descrlptlve stutlstlcul unulysls; (2) u
templute unulysls style thut lnvolves the development of un
unulysls gulde (templute) used to sort the dutu; ( 3) un
edltlng unulysls style thut lnvolves un lnterpretutlon of the
dutu on whlch u cutegorlzutlon scheme ls bused; und (4) un
lmmerslon/ crystulllzutlon style thut lnvolves the unulyst's
totul lmmerslon ln und reflectlon of text muterluls. Most
nurse reseurcher use u strutegy thut ls best churucterlzed us
un edltlng unulysls styleդu style thut typlcully requlres four
types of lntellectuul processes: comprehendlng,
syntheslzlng, theorlzlng, und recontextu-ullzlng (explorutlon
of the developed theory ln terms of lts uppllcublllty to other
settlngs or groups).
The flrst mu|or step ln unulyzlng quulltutlve dutu ls to
orgunlze the muterluls uccordlng to some plun, so thut
portlons of the dutu cun be reudlly retrleved. Typlcully,
quulltutlve reseurchers develop u cutegorlzutlon scheme
bused on u reudlng of u portlon of the dutu und then code
the content of the dutu bused on thls system. Trudltlonully,
reseurchers huve developed conceptuul flles for orgunlzlng
thelr dutu. In uslng thls system, reseurchers flrst code thelr
dutu ln the murglns of the prlnted nurrutlve muterluls (e.g.,
observutlonul notes or trunscrlpts of lntervlews); cut out the
coded excerpts; und flnully pluce euch excerpt lnto u flle
correspondlng to euch of the toplcs covered ln the codlng
scheme. Then the reseurcher cun retrleve ull of the
lnformutlon on u toplc by golng to u slngle flle. However,
the wldespreud uvullublllty of personul computers und
upproprlute softwure hus lessened the burden of lndexlng,
orgunlzlng, und retrlevlng quulltutlve muterluls. There ure
now u wlde vurlety of progrums thut not only perform buslc
lndexlng functlons but ulso offer vurlous enhuncements thut
cun fucllltute unulysls of the dutu.
The uctuul unulysls of dutu beglns wlth u seurch for
themes. The seurch for themes lnvolves not only the
dlscovery of commonulltles ucross sub|ects, but ulso of
nuturul vurlutlon ln the dutu. The next step generully
lnvolves u vulldutlon of the themutlc unulysls. Some re-
seurchers use quusl-stutlstlcs, whlch lnvolves u tubulutlon of
the frequency wlth whlch certuln themes or relutlons ure
supported by the dutu. In u flnul step, the unulyst trles to
weuve the themutlc strunds together lnto un lntegruted
plcture of the phenomenon under lnvestlgutlon.
Although thls overvlew summurlzes some of the mu|or
steps, there ure u number of dlfferent phllosophles
underlylng quulltutlve unulysls. Anulytlc lnductlon refers to
un upprouch ln whlch the reseurcher ulternutes buck und
forth between tentutlve deflnltlon of emerglng hypotheses
und tentutlve explunutlon, wlth euch lterutlon muklng
reflnements. However, the most wldely used upprouch by
nurse reseurchers ls grounded theory. Grounded theory
beglns wlth open codlng (level I codlng)դthe development
of u cutegorlzutlon scheme (uslng constunt compurlson) und
subsequent lnltlul codlng of the dutu. Level II codlng (ulso
referred to us uxlul codlng) ls u reconstructlve process thut
puts dutu buck together ln new wuys by connectlng
cutegorles und subcutegorles. The unulyst beglns the
process of lntegrutlon by revlewlng und sortlng the memos
thut huve been used to document conceptuul ldeus
throughout the dutu collectlon und dutu unulysls process.
'urlng level III codlng (or selectlve codlng) the grounded
theory unulyst seurches for the core cutegoryդthe centrul
phenomenon -used to lntegrute ull others. Thls phuse
results ln un emerglng theory of u buslc soclul process
(%SP) thut ls grounded ln the dutu.

B STUDY ACTIVITIES
Chupter 22 of the StXdy GX¡de to Accompuny NXrs¡ng
Peseurch. 3r¡nc¡pIes und Methods, 6th ed., offers vur¡oXs
exerc¡ses und stXdy sXggest¡ons for re¡nforc¡ng the concepts
presented ¡n th¡s chupter. Add¡t¡onuIIy, the foIIow¡ng stXdy
qXest¡ons cun be uddressed.
1. SXggest u reseurch probIem umenubIe to qXuI¡tut¡ve
reseurch. ExpIu¡n why yoX th¡nk the probIem ¡s
better sX¡ted to u qXuI¡tut¡ve thun to u qXunt¡tut¡ve
upprouch.
2. Peud u qXuI¡tut¡ve nXrs¡ng stXdy (severuI ure
sXggested Xnder SXbstunt¡ve Peferences). 'o yoX
th¡nk thut ¡f u d¡fferent ¡nvest¡gutor hud gone ¡nto the
f¡eId to stXdy the sume probIem, the concIXs¡ons
woXId huve been the sume? How gener-uI¡zubIe ure
the reseurcher's f¡nd¡ngs? Whut d¡d the reseurcher
Ieurn thut he or she woXId probubIy not huve Ieurned
w¡th u more strXctXred und qXunt¡f¡ed upprouch?
3. As u cIuss uss¡gnment, huve euch stXdent usk two
peopIe to descr¡be the¡r concept¡on of prevent¡ve
heuIth cure und whut ¡t meuns ¡n the¡r du¡Iy I¡ves.
3ooI uII the nurrut¡ve descr¡pt¡ons, und deveIop u
cod¡ng scheme to orgun¡ze the reusons. Whut ure the
mu|or themes thut emerge?

% SUGGESTED READINGS

0ethodoIog¡c PeIerences
%ecker, H. S. (1970). Soc¡oIog¡cuI work. Ch¡cugo: AId¡ne.
Curey, 0. A., 6c Sm¡th, 0. W. (1994). Cuptur¡ng the group
eIIect ¡n Iocus groups: A spec¡uI concern ¡n unuIys¡s.
4uuI¡tut¡+e HeuIth Peseurch, 4, 123-127.
Chen¡tz, W. C, 6c Swunson, J. 0. (1986). )rom pruct¡ce to
grounded theory: 4uuI¡tut¡+e reseurch ¡n nurs¡ng. 0enIo
Purk, CA: Add¡son-WesIey.
Crubtree, %. E,6c 0¡IIer, W. L. (Eds.). (1992). Do¡ng
quuI¡tut¡+e reseurch. Newbury Purk, CA: Suge.
&hupter 22. The AnuIys¡s oI 4XuI¡tut¡+e Dutu 18

Gephurt, 5. P., Jr. (1988). Ethnostut¡st¡cs. 4XuI¡tut¡+e
IoXndut¡ons Ior qXunt¡tut¡+e reseurch. NewbXry Purk,
CA. Suge.
GIuser, %. G., 6c Struuss, A. L. (1967). The d¡sco+ery oI
grounded theory. Struteg¡es Ior quuI¡tut¡+e reseurch.
Ch¡cugo. AId¡ne.
KnuII, K. A., 6c Webster, '. C. (1988). Munug¡ng und
unuIyz¡ng quuI¡tut¡+e dutu. A descr¡pt¡on oI tusks,
techn¡ques, und muter¡uIs. Western JournuI oI Nurs¡ng
Peseurch, 10, 195-218.
LoIIund, J., 6c LoIIund, L. H. (1984). AnuIyz¡ng soc¡uI
sett¡ngs. A gu¡de to quuI¡tut¡+e obser+ut¡on und unuIys¡s
(2nd ed.). %eImont, CA. Wudsworth.
M¡Ies, M. %., 6c Hubermun, A. M. (1994). 4uuI¡tut¡+e dutu
unuIys¡s. An expunded sourcebook (2nd ed.). Thousund
Ouks, CA. Suge.
Morgun, '. L. (1993). 4uuI¡tut¡+e content unuIys¡s. A gu¡de
to puths not tuken. 4uuI¡tut¡+e HeuIth Peseurch, 3,
112-121.
Morse, J. M., 6c F¡eId, P. A. (1995). 4uuI¡tut¡+e reseurch
methods Ior heuIth proIess¡onuIs (2nd ed.). Thousund
Ouks, CA. Suge.
Struuss, A., 6c Corb¡n, J. (1990). %us¡cs oI quuI¡tut¡+e
reseurch. Newbury Purk, CA. Suge.
Substunt¡+e PeIerences
%urton, J. A. (1991). PurentuI uduptut¡on to udoIescent drug
ubuse. An ethnogruph¡c study oI roIe IormuIut¡on ¡n
response to courtesy st¡gmu. PubI¡c HeuIth Nurs¡ng, 8,
39-45.
%r¡ght, M. A. (1992). Muk¡ng pIuce. The I¡rst b¡rth ¡n un
¡ntergenerut¡onuI Ium¡Iy context. 4uuI¡tut¡+e HeuIth
Peseurch, 2, 75-78.
GugI¡urd¡, %. A. (1991). The Ium¡Iy's exper¡ence oI I¡+¡ng
w¡th u ch¡Id w¡th 'uchenne muscuIur dystrophy. AppI¡ed
Nurs¡ng Peseurch, 4, 159-164.
K¡ng, K. M., 6c Jensen, L. (1994). Preser+¡ng the seII.
Women hu+¡ng curd¡uc surgery. Heurt & Lung, 23,
99-105.
Muyo, K. (1992). Phys¡cuI uct¡+¡ty pruct¡ces umong
Amer¡cun bIuck work¡ng women. 4uuI¡tut¡+e HeuIth
Peseurch, 2, 318-333.
PusseII, C. K. (1996). EIder cure rec¡p¡ents' cure-seek¡ng
process. Western JournuI oI Nurs¡ng Peseurch,
18, 43-62.