polit chapter 3

Researchers usually work within a paradigmthat is consistent with their world view, andthat gives rise to the types of question that excitetheir curiosity. The maturity of the concept of inter-est also may lead to one or the other paradigm: whenlittle is known about a topic, a qualitative approachis often more fruitful than a quantitative one.

The progression of activities differs for quali-tative and quantitative researchers; we discuss theflow of both in this chapter. First, however, webriefly describe broad categories of quantitativeand qualitative research.

MAJOR CLASSES OFQUANTITATIVE ANDQUALITAT IVE RESEARCH

Experimental and NonexperimentalStudies in Quantitative Research

A basic distinction in quantitative studies is the dif-ference between experimental and nonexperimen-tal research. In experimental research, researchersactively introduce an intervention or treatment. Innonexperimental research, on the other hand, re-searchers collect data without making changes orintroducing treatments. For example, if a re-searcher gave bran flakes to one group of subjectsand prune juice to another to evaluate whichmethod facilitated elimination more effectively, the

Overview of the ResearchProcess in Qualitative andQuantitative Studies

study would be experimental because the re-searcher intervened in the normal course of things.In this example, the researcher created an activevariable involving a dietary intervention. If, on theother hand, a researcher compared elimination pat-terns of two groups of people whose regular eatingpatterns differedfor example, some normallytook foods that stimulated bowel elimination andothers did notthere is no intervention. Such astudy focuses on existing attributes and is nonex-perimental.

Experimental studies are explicitly designed totest causal relationships. Sometimes nonexperi-mental studies also seek to elucidate or detectcausal relationships, but doing so is tricky and usu-ally is less conclusive. Experimental studies offerthe possibility of greater control over extraneousvariables than nonexperimental studies.

Example of experimental research:Johnson (2001) tested the effects of a sub-

maximal exercise protocol, in comparison with a near-maximal voluntary contraction protocol,on continence control and muscle contractionstrength among women with genuine stress urinaryincontinence.

In this example, the researcher intervened bydesignating that some women would receive thesubmaximal exercise protocol and others wouldnot. In other words, the researcher controlled the

3

CHAPTER 3 Overview of the Research Process in Qualitative and Quantitative Studies 47

independent variable, which in this case was thetype of protocol.

Example of nonexperimental research:Wong and her co-researchers (2002)

searched for factors that contributed to hospitalreadmission in a Hong Kong hospital. A readmittedgroup was compared with a nonreadmitted groupof patients in terms of demographic characteristicsand health conditions upon admission.

In this nonexperimental study, the researchersdid not intervene in any way; they observed andmeasured subjects attributes. They exploredwhether there were identifiable characteristics andconditions that distinguished the two groups of pa-tients, with the aim of discovering opportunities toreduce readmissions.

Research Traditions in Qualitative Research

Qualitative studies are often rooted in researchtraditions that originate in the disciplines of an-thropology, sociology, and psychology. Three suchtraditions have had especially strong influences onqualitative nursing research and are briefly describehere. Chapter 11 provides a fuller discussion ofalternative research traditions and the methodsassociated with them.

The grounded theory tradition, which has itsroots in sociology, seeks to describe and understandthe key social psychological and structural processesthat occur in a social setting. Grounded theory wasdeveloped in the 1960s by two sociologists, Glaserand Strauss (1967). The focus of most grounded the-ory studies is on a developing social experiencethesocial and psychological stages and phases that char-acterize a particular event or episode. A major com-ponent of grounded theory is the discovery of a corevariable that is central in explaining what is going onin that social scene. Grounded theory researchersstrive to generate comprehensive explanations ofphenomena that are grounded in reality.

Example of a grounded theory study:Hauck and Irurita (2002) conducted a

grounded theory study to explain the maternal

process of managing late stages of breastfeedingand weaning the child from the breast.

Phenomenology, which has its disciplinaryroots in both philosophy and psychology and isrooted in a philosophical tradition developed byHusserl and Heidegger, is concerned with the livedexperiences of humans. Phenomenology is an ap-proach to thinking about what life experiences ofpeople are like and what they mean. The phenome-nological researcher asks the questions: What is theessence of this phenomenon as experienced bythese people? Or, What is the meaning of the phe-nomena to those who experience it?

Example of a phenomenological study:Sundin, Norberg, and Jansson (2001) con-

ducted a phenomenological study to illuminate thelived experiences of care providers who were highlyskilled communicators in their relationships withpatients with stroke and aphasia.

Ethnography is the primary research traditionwithin anthropology, and provides a framework for studying the meanings, patterns, and experi-ences of a defined cultural group in a holistic fash-ion. Ethnographers typically engage in extensivefieldwork, often participating to the extent possiblein the life of the culture under study. Ethnographicresearch is in some cases concerned with broadlydefined cultures (e.g., Haitian refugee communi-ties), but sometimes focuses on more narrowlydefined cultures (e.g., the culture of emergencydepartments). The aim of ethnographers is to learnfrom (rather than to study) members of a culturalgroup, to understand their world view as they per-ceive and live it.

Example of an ethnographic study:Powers (2001) undertook an ethnographic

analysis of a nursing home residence, focusing onthe ethical issues of daily living affecting nursinghome residents with dementia.

MAJOR STEPS IN AQUANTITATIVE STUDY

In quantitative studies, researchers move from thebeginning point of a study (the posing of a question)

48 PART 1 Foundations of Nursing Research

fascinate you or about which you have a passionateinterest or curiosity.

Step 2: Reviewing the Related LiteratureQuantitative research is typically conducted withinthe context of previous knowledge. To build on ex-isting theory or research, quantitative researchersstrive to understand what is already known about aresearch problem. A thorough literature reviewprovides a foundation on which to base new knowl-edge and usually is conducted well before any dataare collected in quantitative studies. For clinicalproblems, it would likely also be necessary to learnas much as possible about the status quo of cur-rent procedures relating to the topic, and to reviewexisting practice guidelines or protocols.

A familiarization with previous studies can alsobe useful in suggesting research topics or in identify-ing aspects of a problem about which more researchis needed. Thus, a literature review sometimes pre-cedes the delineation of the research problem.

Step 3: Undertaking Clinical FieldworkIn addition to refreshing or updating clinical knowl-edge based on written work, researchers embarkingon a clinical nursing study benefit from spendingtime in clinical settings, discussing the topic withclinicians and health care administrators, and ob-serving current practices. Sterling (2001) notes thatsuch clinical fieldwork can provide perspectives onrecent clinical trends, current diagnostic proce-dures, and relevant health care delivery models; itcan also help researchers better understand affectedclients and the settings in which care is provided. Inaddition to expanding the researchers clinical andconceptual knowledge, such fieldwork can bevaluable in developing methodologic tools forstrengthening the study. For example, in the courseof clinical fieldwork researchers might learn whatextraneous variables need to be controlled, or mightdiscover the need for Spanish-speaking researchassistants.

As with literature reviews, clinical fieldworkmay serve as a stimulus for developing researchquestions and may be the first step in the processfor some researchers.

to the end point (the obtaining of an answer) in afairly linear sequence of steps that is broadly similaracross studies. In some studies, the steps overlap,whereas in others, certain steps are unnecessary.Still, there is a general flow of activities that is typi-cal of a quantitative study. This section describesthat flow, and the next section describes how quali-tative studies differ.

Phase 1: The Conceptual Phase

The early steps in a quantitative research project typ-ically involve activities with a strong conceptual orintellectual element. These activities include reading,conceptualizing, theorizing, reconceptualizing, andreviewing ideas with colleagues or advisers. Duringthis phase, researchers call on such skills as creativ-ity, deductive reasoning, insight, and a firm ground-ing in previous research on the topic of interest.

Step 1: Formulating and Delimiting the ProblemOne of the first things a researcher must do is de-velop a research problem and research questions.Good research depends to a great degree on goodquestions. Without a significant, interesting prob-lem, the most carefully and skillfully designed re-search project is of little value.

Quantitative researchers usually proceed fromthe selection of a broad problem area to the devel-opment of specific questions that are amenable toempirical inquiry. In developing a research ques-tion to be studied, nurse researchers must pay closeattention to substantive issues (Is this researchquestion significant, given the existing base ofknowledge?); clinical issues (Could findings fromthis research be useful in clinical practice?); andmethodologic issues (How can this question bestbe studied to yield high-quality evidence?). Theidentification of research questions must also takeinto consideration practical and ethical concerns.

TIP: A critical ingredient in developinggood research questions is personal interest.

We offer this advice to those of you who plan to un-dertake a research project: Begin with topics that

Step 4: Defining the Framework andDeveloping Conceptual DefinitionsTheory is the ultimate aim of science in that it tran-scends the specifics of a particular time, place, andgroup of people and aims to identify regularities inthe relationships among variables. When quantita-tive research is performed within the context of atheoretical frameworkthat is, when previous the-ory is used as a basis for generating predictions thatcan be tested through empirical researchthe find-ings may have broader significance and utility.

Even when the research question is not em-bedded in a theory, researchers must have a clearsense of the concepts under study. Thus, an impor-tant task in the initial phase of a project is the de-velopment of conceptual definitions.

Step 5: Formulating HypothesesA hypothesis is a statement of the researchers ex-pectations about relationships between the vari-ables under investigation. Hypotheses, in otherwords, are predictions of expected outcomes; theystate the relationships researchers expect to find asa result of the study.

The research question identifies the conceptsunder investigation and asks how the conceptsmight be related; a hypothesis is the predicted an-swer. For example, the initial research questionmight be phrased as follows: Is preeclamptic tox-emia in pregnant women associated with stress fac-tors present during pregnancy? This might be trans-lated into the following hypothesis: Pregnantwomen with a higher incidence of stressful eventsduring pregnancy will be more likely than womenwith a lower incidence of stress to experiencepreeclamptic toxemia. Most quantitative studiesare designed to test hypotheses through statisticalanalysis.

Phase 2: The Design and Planning Phase

In the second major phase of a quantitative re-search project, researchers make decisions aboutthe methods and procedures to be used to addressthe research question, and plan for the actual

collection of data. Sometimes the nature of thequestion dictates the methods to be used, but moreoften than not, researchers have considerable flexi-bility to be creative and make many decisions.These methodologic decisions usually have crucialimplications for the validity and reliability of thestudy findings. If the methods used to collect andanalyze research data are seriously flawed, then theevidence from the study may be of little value.

Step 6: Selecting a Research DesignThe research design is the overall plan for obtaininganswers to the questions being studied and for han-dling some of the difficulties encountered during theresearch process. A wide variety of research designsis available for quantitative studies, including numer-ous experimental and nonexperimental designs.

In designing the study, researchers specifywhich specific design will be adopted and whatcontrols will be used to minimize bias and enhancethe interpretability of results. In quantitative stud-ies, research designs tend to be highly structured,with tight controls over extraneous variables.Research designs also indicate other aspects of theresearchfor example, how often subjects will bemeasured or observed, what types of comparisonswill be made, and where the study will take place.The research design is essentially the architecturalbackbone of the study.

Step 7: Developing Protocols for the InterventionIn experimental research, researchers actively in-tervene and create the independent variable, whichmeans that people in the sample will be exposed todifferent treatments or conditions. For example, ifwe were interested in testing the effect of biofeed-back in treating hypertension, the independent vari-able would be biofeedback compared with eitheran alternative treatment (e.g., relaxation therapy),or with no treatment. The intervention protocolfor the study would need to be developed, specify-ing exactly what the biofeedback treatment wouldentail (e.g., who would administer it, how fre-quently and over how long a period the treatmentwould last, what specific equipment would be used,



probability sampling methods, which use randomprocedures for selecting subjects. In a probabilitysample, every member of the population has anequal probability of being included in the sample.With nonprobability sampling, by contrast, thereis no way of ensuring that each member of the pop-ulation could be selected; consequently, the risk ofa biased (unrepresentative) sample is greater. Thedesign of a sampling plan includes the selection ofa sampling method, the specification of the samplesize (i.e., number of subjects), and the developmentof procedures for recruiting subjects.

Step 10: Specifying Methods to Measure the Research VariablesQuantitative researchers must develop methods toobserve or measure the research variables as accu-rately as possible. Based on the conceptual defini-tions, the researcher selects or designs appropriatemethods of operationalizing the variables andcollecting data. A variety of quantitative data col-lection approaches exist. Biophysiologic mea-surements often play an important role in clinicalnursing research. Through self-reports, anotherpopular method of data collection, subjects areasked directly about their feelings, behaviors,attitudes, and personal traits (for example, in an in-terview with research personnel). Another tech-nique is observation, wherein researchers collectdata by observing and recording aspects ofpeoples behavior.

Data collection methods vary in the degree ofstructure imposed on subjects. Quantitative ap-proaches tend to be fairly structured, involving theuse of a formal instrument that elicits the sameinformation from every subject. Sometimes re-searchers need to develop their own instruments,but more often they use or adapt measuring instru-ments that have been developed by others. Thetask of measuring research variables and develop-ing a data collection plan is a complex and chal-lenging process that permits a great deal ofcreativity and choice. Before finalizing the datacollection plan, researchers must carefully evaluatewhether the chosen methods capture key conceptsaccurately.

and so on) and what the alternative conditionwould be. The goal of well-articulated protocols isto have all subjects in each group treated in thesame way. (In nonexperimental research, of course,this step would not be necessary.)

Step 8: Identifying the Population to be StudiedBefore selecting subjects, quantitative researchersneed to know what characteristics participantsshould possess. Researchers and others using thefindings also need to know to whom study resultscan be generalized. Thus, during the planningphase of quantitative studies, researchers mustidentify the population to be studied. The termpopulation refers to the aggregate or totality ofthose conforming to a set of specifications. For ex-ample, we might specify nurses (RNs) and resi-dence in the United States as attributes of interest;the study population would then consist of all li-censed RNs who reside in the United States. Wecould in a similar fashion define a population con-sisting of all children younger than 10 years of agewith muscular dystrophy in Canada, or all thechange-of-shift reports for the year 2002 inMassachusetts General Hospital.

Step 9: Designing the Sampling PlanResearch studies almost always rely on a sample ofsubjects, who are a subset of the population. It isclearly more practical and less costly to collect datafrom a sample than from an entire population. Therisk, however, is that the sample might not ade-quately reflect the populations behaviors, traits,symptoms, or beliefs.

Various methods of obtaining samples areavailable. These methods vary in cost, effort, andskills required, but their adequacy is assessed bythe same criterion: the representativeness of theselected sample. That is, the quality of the samplefor quantitative studies depends on how typical, orrepresentative, the sample is of the population withrespect to the variables of concern in the study.Sophisticated sampling procedures can producesamples that have a high likelihood of being repre-sentative. The most sophisticated methods are

Step 11: Developing Methods for Safeguarding Human/Animal RightsMost nursing research involves human subjects, al-though some studies involve animals. In eithercase, procedures need to be developed to ensurethat the study adheres to ethical principles. For ex-ample, forms often need to be developed to docu-ment that subjects participation in the study wasvoluntary. Each aspect of the study plan needs to bereviewed to determine whether the rights of sub-jects have been adequately protected. Often that re-view involves a formal presentation to an externalcommittee.

Step 12: Finalizing and Reviewing the Research PlanBefore actually collecting research data, researchersoften perform a number of tests to ensure thatplans will work smoothly. For example, they mayevaluate the readability of any written materials todetermine if people with below-average readingskills can comprehend them, or they may need totest whether technical equipment is functioningproperly. If questionnaires are used, it is importantto know whether respondents understand questionsor find certain ones objectionable; this is usually re-ferred to as pretesting the questionnaire. Duringfinal study preparations, researchers also have to de-termine the type of training to provide to those re-sponsible for collecting data. If researchers haveconcerns about their study plans, they may under-take a pilot study, which is a small-scale version ortrial run of the major study.

Normally, researchers have their researchplan critiqued by peers, consultants, or other re-viewers to obtain substantive, clinical, or method-ologic feedback before implementing the plan.When researchers seek financial support for thestudy, a proposal typically is submitted to a fund-ing source, and reviewers of the proposed planusually suggest improvements. Students conduct-ing a study as part of a course or degree requirementhave their plans reviewed by faculty advisers.Even under other circumstances, however, re-searchers are well advised to ask individualsexternal to the project to assess preliminary plans.

Experienced researchers with fresh perspectivescan often be invaluable in identifying pitfalls andshortcomings that otherwise might not have beenrecognized.

Phase 3: The Empirical Phase

The empirical portion of quantitative studies in-volves collecting research data and preparing thosedata for analysis. In many studies, the empiricalphase is one of the most time-consuming parts ofthe investigation, although the amount of timespent collecting data varies considerably from onestudy to the next. If data are collected by distribut-ing a written questionnaire to intact groups, thistask may be accomplished in a matter of days.More often, however, data collection requires sev-eral weeks, or even months, of work.

Step 13: Collecting the DataThe actual collection of data in a quantitative studyoften proceeds according to a preestablished plan.The researchers plan typically specifies proce-dures for the actual collection of data (e.g., whereand when the data will be gathered); for describingthe study to participants; and for recording infor-mation. Technological advances in the past fewdecades have expanded possibilities for automatingdata collection.

A considerable amount of both clerical and ad-ministrative work is required during data collec-tion. Researchers typically must be sure, for exam-ple, that enough materials are available to completethe study; that participants are informed of the timeand place that their presence may be required; thatresearch personnel (such as interviewers) are con-scientious in keeping their appointments; thatschedules do not conflict; and that a suitable sys-tem of maintaining confidentiality of informationhas been implemented.

Step 14: Preparing the Data for AnalysisAfter data are collected, a few preliminary activitiesmust be performed before data analysis begins. Forinstance, it is normally necessary to look throughquestionnaires to determine if they are usable.



Step 16: Interpreting the ResultsBefore the results of a study can be communicatedeffectively, they must be systematically interpreted.Interpretation is the process of making sense ofthe results and of examining their implications. Theprocess of interpretation begins with an attempt toexplain the findings within the context of the theo-retical framework, prior empirical knowledge, andclinical experience.

If research hypotheses have been supported,an explanation of the results may be straightfor-ward because the findings fit into a previouslyconceived argument. If hypotheses are not sup-ported, researchers must explain why this mightbe so. Is the underlying conceptualization wrong,or was it inappropriate for the research problem?Or do the findings reflect problems with the re-search methods rather than the framework (e.g.,was the measuring tool inappropriate)? To pro-vide sound explanations, researchers not onlymust be familiar with clinical issues, prior re-search, and conceptual underpinnings, but mustbe able to understand methodologic limitations ofthe study. In other words, the interpretation of thefindings must take into account all available evi-dence about the studys reliability and validity.Researchers need to evaluate critically the deci-sions they made in designing the study and to rec-ommend alternatives to others interested in thesame research problem.

Phase 5: The Dissemination Phase

The analytic phase brings researchers full circle: itprovides answers to the questions posed in the firstphase of the project. However, researchers respon-sibilities are not complete until the study results aredisseminated.

Step 17: Communicating the FindingsA study cannot contribute evidence to nursing prac-tice if the results are not communicated. The mostcompelling hypothesis, the most rigorous study, themost dramatic results are of no value to the nursingcommunity if they are unknown. Anotherandoften finaltask of a research project, therefore, is

Sometimes forms are left almost entirely blank orcontain other indications of misinterpretation ornoncompliance. Another step is to assign identifica-tion numbers to the responses or observations ofdifferent subjects, if this was not done previously.

Coding of the data is typically needed at thispoint. As noted in Chapter 2, coding involvesthe translation of verbal data into numeric form,according to a specified plan. This might mean as-signing numeric codes to categorical variablessuch as gender (e.g., 1 for females and 2 formales). Coding might also be needed to categorizenarrative responses to certain questions. For exam-ple, patients verbatim responses to a questionabout the quality of nursing care they received dur-ing hospitalization might be coded into positivereactions (1), negative reactions (2), neutral reac-tions (3), or mixed reactions (4). Another prelimi-nary step involves transferring the data from writ-ten documents onto computer files for subsequentanalysis.

Phase 4: The Analytic Phase

Quantitative data gathered in the empirical phaseare not reported in raw form. They are subjected toanalysis and interpretation, which occurs in thefourth major phase of a project.

Step 15: Analyzing the DataThe data themselves do not provide answers to re-search questions. Ordinarily, the amount of datacollected in a study is rather extensive; researchquestions cannot be answered by a simple perusalof numeric information. Data need to be processedand analyzed in an orderly, coherent fashion.Quantitative information is usually analyzedthrough statistical procedures. Statistical analysescover a broad range of techniques, from simpleprocedures that we all use regularly (e.g., comput-ing an average) to complex and sophisticated meth-ods. Although some methods are computationallyformidable, the underlying logic of statistical testsis relatively easy to grasp, and computers haveeliminated the need to get bogged down with de-tailed mathematic operations.

the preparation of a research report that can beshared with others.

Research reports can take various forms: termpapers, dissertations, journal articles, presentationsat professional conferences, and so on. Journal ar-ticlesreports appearing in such professional jour-nals as Nursing Researchusually are the mostuseful because they are available to a broad, inter-national audience. There is also a growing numberof outlets for research dissemination on the Internet.

Step 18: Utilizing the Findings in PracticeMany interesting studies have been conducted bynurses without having any effect on nursing prac-tice or nursing education. Ideally, the concludingstep of a high-quality study is to plan for its uti-lization in practice settings. Although nurse re-searchers may not themselves be in a position toimplement a plan for utilizing research findings,they can contribute to the process by including intheir research reports recommendations regardinghow the evidence from the study could be incorpo-rated into the practice of nursing and by vigorouslypursuing opportunities to disseminate the findingsto practicing nurses.

Organization of a QuantitativeResearch Project

The steps described in the preceding section repre-sent an idealized conception of what researchersdo. The research process rarely follows a neatlyprescribed pattern of sequential procedures.Developments in one step, for example, may re-quire alterations in a previously completed activity.Nevertheless, for the quantitative researcher, care-ful organization is very important.

Almost all research projects are conductedunder some time pressure. Students in researchcourses have end-of-term deadlines; government-sponsored research involves funds granted for aspecified time. Those who may not have such for-mal time constraints (e.g., graduate students work-ing on theses or dissertations) normally have theirown goals for project completion. Setting up atimetable in advance may be an important means of

meeting such goals. Having deadlines for taskseven tentative oneshelps to impose order and de-limits tasks that might otherwise continue indefi-nitely, such as problem selection and literaturereviews.

It is not possible to give even approximate fig-ures for the relative percentage of time that shouldbe spent on each task in quantitative studies. Someprojects require many months to develop andpretest the measuring instruments, whereas otherstudies use previously existing ones, for example.The write-up of the study may take many months oronly a few days. Clearly, not all steps are equallytime-consuming. It would make little sense simplyto divide the available time by the number of tasks.

Let us suppose a researcher was studying thefollowing problem: Is a womans decision to havean annual mammogram related to her perceivedsusceptibility to breast cancer? Using the organiza-tion of steps outlined earlier, here are some of thetasks that might be undertaken:*

1. The researcher, who lost her mother to breastcancer, is concerned that many older womendo not get a mammogram regularly. Her spe-cific research question is whether mammo-gram practices are different for women whohave different views about their susceptibilityto breast cancer.

2. The researcher reviews the research literatureon mammograms, factors affecting mammog-raphy decisions, and interventions designedto promote it.

3. The researcher does clinical fieldwork by dis-cussing the problem with nurses and otherhealth care professionals in various clinicalsettings (health clinics, private obstetrics andgynecology practices) and by informally dis-cussing the problem with women in a supportgroup for breast cancer victims.

4. The researcher examines frameworks for con-ceptualizing the problem. She finds that the

*This is, of course, only a partial list of tasks and is designed toillustrate the flow of activities; the flow in this example is moreorderly than would ordinarily be true.



13. Data are collected by conducting telephoneinterviews with the research sample.

14. Data are prepared for analysis by codingthem and entering them onto a computer file.

15. Data are analyzed using a statistical softwarepackage.

16. The results indicate that the hypothesis is sup-ported; however, the researchers interpretationmust take into consideration that many womenwho were asked to participate in the study de-clined to do so. Moreover, the analysis revealedthat mammogram use in the sample was sub-stantially higher than had been reported in ear-lier studies.

17. The researcher presents an early report on herfindings and interpretations at a conference ofSigma Theta Tau International. She subse-quently publishes the report in the WesternJournal of Nursing Research.

18. The researcher seeks out clinicians to discusshow the study findings can be utilized inpractice.

The researcher in this study wants to conductthis study over a 2-year period. Figure 3-1 presentsa hypothetical schedule for the research tasks to becompleted. (The selection of the problem is not in-cluded because the research topic has already beenidentified.) Note that many steps overlap or are un-dertaken concurrently. Some steps are projected toinvolve little time, whereas others require monthsof work.

In developing a time schedule of this sort, anumber of considerations should be kept in mind,including researchers level of knowledge andmethodologic competence. Resources available toresearchers, in terms of research funds and person-nel, greatly influence time estimates. In the presentexample, the researcher almost certainly would haverequired funding from a sponsor to help pay for thecost of hiring interviewers, unless she were able todepend on colleagues or students.

It is also important to consider the practical as-pects of performing the study, which were not all enu-merated in the preceding section. Obtaining supplies,securing permissions, getting approval for using

Health Belief Model (see Chapter 6) is relevant, and this helps her to develop a con-ceptual definition of susceptibility to breastcancer.

5. Based on what the researcher has learned, thefollowing hypothesis is developed: Womenwho perceive themselves as not susceptible tobreast cancer are less likely than other womento get an annual mammogram.

6. The researcher adopts a nonexperimental re-search design that involves collecting datafrom subjects at a single point in time. She de-signs the study to control the extraneous vari-ables of age, marital status, and general healthstatus.

7. There is no intervention in this study (the de-sign is nonexperimental) and so this step doesnot need to be undertaken.

8. The researcher designates that the populationof interest is women between the ages of 50and 65 years living in Canada who have notbeen previously diagnosed as having anyform of cancer.

9. The researcher decides to recruit for the re-search sample 200 women living in Toronto;they are identified at random using a tele-phone procedure known as random-digitdialing.

10. The research variables will be measuredthrough self-report; that is, the independentvariable (perceived susceptibility), dependentvariable (mammogram history), and extrane-ous variables will be measured by asking thesubjects a series of questions. The researcherdecides to use existing measures of key vari-ables, rather than developing new ones.

11. A human subjects committee at the re-searchers institution is asked to review theresearch plans to determine whether thestudy adheres to ethical standards.

12. Plans for the study are finalized: the methodsare reviewed and refined by colleagues withclinical and methodologic expertise; the datacollection instruments are pretested; and in-terviewers who will collect the data aretrained.

forms or instruments, hiring staff, and holding meet-ings are all time-consuming, but necessary, activities.

Individuals differ in the kinds of tasks that ap-peal to them. Some people enjoy the preliminaryphase, which has a strong intellectual component,whereas others are more eager to collect the data, atask that is more interpersonal. Researchers should,however, allocate a reasonable amount of time todo justice to each activity.

ACTIVIT IES IN AQUALITAT IVE STUDY

As we have just seen, quantitative research in-volves a fairly linear progression of tasksresearchers plan in advance the steps to be taken tomaximize study integrity and then follow thosesteps as faithfully as possible. In qualitative stud-ies, by contrast, the progression is closer to a circlethan to a straight linequalitative researchers are

continually examining and interpreting data andmaking decisions about how to proceed based onwhat has already been discovered.

Because qualitative researchers have a flexibleapproach to the collection and analysis of data, it is impossible to define the flow of activities preciselythe flow varies from one study to another,and researchers themselves do not know ahead oftime exactly how the study will proceed. The fol-lowing sections provide a sense of how qualitativestudies are conducted by describing some majoractivities and indicating how and when they mightbe performed.

Conceptualizing and Planning aQualitative Study

Identifying the Research ProblemLike quantitative researchers, qualitative researchersusually begin with a broad topic area to be studied.

F I G U R E 3 . 1 Project timetable in calendar months.


0 2 4 6 8 10 12 14 16 18 20 22 24

0 2 4 6 8 10 12 14 16 18 20 22 24

Conceptual Phase

Planning Phase

Empirical Phase

Analytic Phase

Dissemination Phase

Step 2Step 3Step 4

Step 5

Step 6Step 8*

Step 9

Step 10Step 11

Step 12

Step 14

Step 13

Step 15

Step 16

Step 17Step 18

* Note that Step 7 was not necessary because this study did not involve an intervention.


perhaps some clinical fieldwork) to identify the mostsuitable and information-rich environment for theconduct of the study. For a qualitative researcher, anideal site is one in which (1) entry is possible; (2) arich mix of people, interactions, and situations relat-ing to the research question is present; and (3) the re-searcher can adoptand maintainan appropriaterole vis--vis study participants. It is critical to ap-praise the suitability of the site (and the settingswithin the site where data will be collecting) beforeentering the field.

In some cases, researchers may have access tothe site selected for the study. In others, however, re-searchers need to gain entre into the site or settingswithin it. A site may be well suited to the needs of theresearch, but if researchers cannot get in, the studycannot proceed. Gaining entre typically involves ne-gotiations with gatekeepers who have the authorityto permit entry into their world. Gaining entre re-quires strong interpersonal skills, as well as familiar-ity with the customs and language of the site. In ad-dition, certain strategies are more likely to succeedthan others. For example, gatekeepers might be per-suaded to be cooperative if it can be demonstratedthat there will be direct benefits to them or their con-stituentsor if a great humanitarian purpose will beserved. Researchers also need to gain the gatekeep-ers trust, and that can only occur if researchers arecongenial, persuasive, forthright about research re-quirements (e.g., how much time the fieldwork willrequire), andperhaps most importantexpressgenuine interest in and concern for the situations ofthe people in the site. In qualitative research, gainingentre is likely to be an ongoing process of establish-ing relationships and rapport with gatekeepers andothers at the site, including prospective informants.

Research Design in Qualitative StudiesAs we have seen, quantitative researchers do notcollect data until the research design has beenfinalized. In a qualitative study, by contrast, the re-search design is often referred to as an emergentdesigna design that emerges during the course ofdata collection. Certain design features are guidedby the qualitative research tradition within whichthe researcher is working, but nevertheless few

However, qualitative researchers usually focus on anaspect of a topic that is poorly understood and aboutwhich little is known. Therefore, they do not develophypotheses or pose highly refined research questionsbefore going into the field. The general topic area maybe narrowed and clarified on the basis of self-reflection and discussion with colleagues (or clients),but researchers may proceed with a fairly broad re-search question that allows the focus to be sharpenedand delineated more clearly once the study is under-way. (Qualitative researchers may also decide to focuson a topic that has been extensively researched quan-titatively, but has had little qualitative attention.)

Doing Literature ReviewsThere are conflicting opinions among qualitative re-searchers about doing a literature review at the out-set of a study. At one extreme are those who believethat researchers should not consult the literature be-fore collecting data. Their concern is that prior stud-ies or clinical writings might influence researchersconceptualization of the phenomena under study.According to this view, the phenomena should beelucidated based on participants viewpoints ratherthan on any prior information. Those sharing thisviewpoint often do a literature review at the end ofthe study rather than at the beginning. Others feelthat researchers should conduct at least a prelimi-nary up-front literature review to obtain some pos-sible guidance (including guidance in identifyingthe kinds of biases that have emerged in studyingthe topic). Still others believe that a full up-front lit-erature review is appropriate. In any case, qualita-tive researchers typically find a relatively smallbody of relevant previous work because of the typesof question they ask.

Selecting and Gaining Entre Into Research SitesDuring the planning phase, qualitative researchersmust also select a site that is consistent with the topicunder study. For example, if the topic is the healthbeliefs of the urban poor, an inner-city neighborhoodwith a high percentage of low-income residents mustbe identified. In making such a decision, researchersmay need to engage in anticipatory fieldwork (and

qualitative studies have rigidly structured designsthat prohibit changes while in the field. As previ-ously noted, qualitative designs are not concernedwith the control of extraneous variables. The fullcontext of the phenomenon is considered an im-portant factor in understanding how it plays out inthe lives of people experiencing it.

Although qualitative researchers do not alwaysknow in advance exactly how the study will progressin the field, they nevertheless must have some senseof how much time is available for field work andmust also arrange for and test needed equipment,such as tape recorders or videotaping equipment.Other planning activities include such tasks as hiringand training interviewers to assist in the collection ofdata; securing interpreters if the informants speak adifferent language; and hiring appropriate consul-tants, transcribers, and support staff.

Addressing Ethical IssuesQualitative researchers, like quantitative researchers,must also develop plans for addressing ethicalissuesand, indeed, there are special concerns inqualitative studies because of the more intimatenature of the relationship that typically developsbetween researchers and study participants.Chapter 7 describes some of these concerns.

Conducting the Qualitative Study

In qualitative studies, the tasks of sampling, datacollection, data analysis, and interpretation typi-cally take place iteratively. Qualitative researchersbegin by talking with or observing a few peoplewho have first-hand experience with the phenome-non under study. The discussions and observationsare loosely structured, allowing for the expressionof a full range of beliefs, feelings, and behaviors.Analysis and interpretation are ongoing, concur-rent activities that guide choices about the kinds ofpeople to sample next and the types of questions toask or observations to make. The actual process ofdata analysis involves clustering together relatedtypes of narrative information into a coherentscheme. The analysis of qualitative data is anintensive, time-consuming activity.

As analysis and interpretation progress,researchers begin to identify themes and categories,which are used to build a rich description or theoryof the phenomenon. The kinds of data obtained andthe people selected as participants tend to becomeincreasingly focused and purposeful as the concep-tualization is developed and refined. Concept devel-opment and verification shape the samplingprocessas a conceptualization or theory develops,the researcher seeks participants who can confirmand enrich the theoretical understandings, as well asparticipants who can potentially challenge them andlead to further theoretical development.

Quantitative researchers decide in advance howmany subjects to include in the study, but qualitativeresearchers sampling decisions are guided by thedata themselves. Many qualitative researchers usethe principle of data saturation, which occurswhen themes and categories in the data becomerepetitive and redundant, such that no new informa-tion can be gleaned by further data collection.

In quantitative studies, researchers seek to col-lect high-quality data by using measuring instru-ments that have been demonstrated to be accurateand valid. Qualitative researchers, by contrast,must take steps to demonstrate the trustworthinessof the data while in the field. The central feature ofthese efforts is to confirm that the findings accu-rately reflect the experiences and viewpoints ofparticipants, rather than perceptions of the re-searchers. One confirmatory activity, for example,involves going back to participants and sharingpreliminary interpretations with them so that theycan evaluate whether the researchers thematicanalysis is consistent with their experiences.Another strategy is to use triangulation to convergeon a thorough depiction of the target phenomena.

An issue that qualitative researchers sometimesneed to address is the development of appropriatestrategies for leaving the field. Because qualitativeresearchers may develop strong relationships withstudy participants and entire communities, theyneed to be sensitive to the fact that their departurefrom the field might seem like a form of rejection orabandonment. Graceful departures and methods ofachieving closure are important.


AQ4


Disseminating Qualitative Findings

Qualitative nursing researchers also strive to sharetheir findings with others at conferences and injournal articles. Qualitative findings, because oftheir depth and richness, also lend themselves morereadily to book-length manuscripts than do quanti-tative findings. Regardless of researchers positionabout when a literature review should be con-ducted, they usually include a summary of priorresearch in their reports as a means of providingcontext for the study.

Quantitative reports almost never present rawdatathat is, data in the form they were collected,which are numeric values. Qualitative reports, bycontrast, are usually filled with rich verbatim pas-sages directly from participants. The excerpts areused in an evidentiary fashion to support or illus-trate researchers interpretations and theoreticalformulations.

Example of raw data in a qualitative report:Scannell-Desch (2000) studied the hard-

ships and personal strategies of 24 female Vietnamwar nurses. One of the emotional hardships theyexperienced had to do with the youth of the patientsand the severity of their injuries. The researchersupported this with the following quote from anarmy nurse:

I had to amputate the leg of one patient. That was the firsttime I ever had to do that. His leg was hanging by a tis-sue band. I was new here, and the doctor yelled at me toget the damn thing off. Doctors take legs off, nursesdont do that. He yelled at me again and said, You doit. (pp. 533534).

Like quantitative researchers, qualitative nurseresearchers want their findings used in nursing prac-tice and subsequent research. Qualitative findingsoften are the basis for formulating hypotheses thatare tested by quantitative researchers, and for devel-oping measuring instruments for both research andclinical purposes. Qualitative findings can also pro-vide a foundation for designing effective nursing in-terventions. Qualitative studies help to shapenurses perceptions of a problem or situation andtheir conceptualizations of potential solutions.

RESEARCH EXAMPLES

In this section, we illustrate the progression ofactivities and discuss the time schedule of twostudies (one quantitative and the other qualitative)conducted by the second author of this book.

Project Schedule for a Quantitative Study

Beck and Gable (2001) undertook a study to evalu-ate the accuracy of the newly developedPostpartum Depression Screening Scale (PDSS) inscreening new mothers for this mood disorder.

Phase 1. Conceptual Phase: 1 MonthThis phase was the shortest, in large part becausemuch of the conceptual work had been done inBeck and Gables (2000) first study, in which theyactually developed the screening scale. The litera-ture had already been reviewed, so all that wasneeded was to update the review. The same frame-work and conceptual definitions that had been usedin the first study were used in the new study.

Phase 2. Design and Planning Phase: 6 MonthsThe second phase was time-consuming. It includednot only fine-tuning the research design, but gain-ing entre into the hospital where subjects were re-cruited and obtaining approval of the hospitalshuman subjects review committee. During this pe-riod, Beck met with statistical consultants and aninstrument development consultant numeroustimes to finalize the study design.

Phase 3. Empirical Phase: 11 MonthsData collection took almost a year to complete.The design called for administering the PDSS to150 mothers who were 6 weeks postpartum, andthen scheduling a psychiatric diagnostic interviewfor them to determine if they were suffering frompostpartum depression. Women were recruitedinto the study during prepared childbirth classes.Recruitment began 4 months before data collec-tion because the researchers had to wait until 6weeks after delivery to gather data. The nurse

psychotherapist, who had her own clinical practice,was able to come to the hospital (a 2-hour drive forher) only 1 day a week to conduct the diagnostic in-terviews; this contributed to the time required toachieve the desired sample size.

Phase 4. Analytic Phase: 3 MonthsStatistical tests were performed to determine a cut-off score on the PDSS above which mothers wouldbe identified as having screened positive for post-partum depression. Data analysis also was under-taken to determine the accuracy of the PDSS inpredicting diagnosed postpartum depression.During this phase, Beck met with the statisticiansand instrument development consultant to interpretresults.

Phase 5. Dissemination Phase: 18 MonthsThe researchers prepared a research report and sub-mitted the manuscript to the journal NursingResearch for possible publication. Within 4 monthsit was accepted for publication, but it was inpress (awaiting publication) for 14 months beforebeing published. During this period, the authorspresented their findings at regional and interna-tional conferences. The researchers also had to pre-pare a summary report for submission to theagency that funded the research.

Project Schedule for a Qualitative Study

Beck (2002) conducted a grounded theory study onmothering twins during the first year after delivery.Total time from start to finish was approximately 2 years.

Phase 1. Conceptual Phase: 3 MonthsBeck became interested in mothers of multiples asa result of her quantitative studies on postpartumdepression. The findings of these studies had re-vealed a much higher prevalence of postpartum de-pression among mothers of multiples than amongthose of singletons. Beck had never studied multi-ple births before, so she needed to review that liter-ature carefully. Gaining entre into the research site

(a hospital) did not take long, however, because shehad previously conducted a study there and wasknown to the hospitals gatekeepers. The key gate-keeper was a nurse who was in charge of the hos-pitals support group for parents of multiplesanurse with whom Beck had developed an excellentrapport in the previous study (the nurse was one ofthe childbirth educators who had helped recruitmothers for the postpartum depression study).

Phase 2. Design and Planning Phase: 4 MonthsAfter reviewing the literature in the conceptualphase, a grounded theory design was selected. Theresearcher met with the nurse who headed the sup-port group to plan the best approach for recruitingmothers of twins into the study. Plans were alsomade for the researcher to attend the monthlymeetings of the support group. Once the designwas finalized, the research proposal was submittedto and approved by both the hospitals and univer-sitys human subjects review committees.

Phase 3. Empirical/Analytic Phases 10 monthsData collection and data analysis phases occurredsimultaneously in this grounded theory study. Beckattended the parents of multiples support groupfor 10 months. During that period, she conductedin-depth interviews with 16 mothers of twins intheir homes, and analyzed her rich and extensivedata. Becks analysis indicated that life on holdwas the basic problem mothers of twins experi-enced during the first year of their twins lives. Asmothers attempted to resume their own lives, theyprogressed through a four-stage process that Beckcalled releasing the pause button.

Phase 4 Dissemination Phase: 6+ MonthsA manuscript was written describing this study andsubmitted for publication in a journal. The manu-script was published in 2002 in the journalQualitative Health Research. In addition to dis-seminating the results as a journal article, Beckpresented the findings at a regional nursing re-search conference.



steps to safeguard the rights of subjects; and(12) finalizing the research plan, by conferringwith colleagues, pretesting instruments, and, insome cases, conducting a pilot study.

The empirical phase involves (13) collectingdata; and (14) preparing data for analysis.

The analytic phase involves (15) analyzing datathrough statistical analysis; and (16) interpret-ing the results.

The dissemination phase entails (17) commu-nicating the findings through the preparation ofresearch reports that can be presented orally orpublished in written form, most often as jour-nal articles; and (18) efforts to promote the useof the study evidence in nursing practice.

The conduct of quantitative studies requirescareful planning and organization. The prepara-tion of a timetable with expected deadlines fortask completion is recommended.

The flow of activities in a qualitative study ismore flexible and less linear.

Qualitative researchers begin with a broad ques-tion regarding the phenomenon of interest, oftenfocusing on a little-studied aspect.

In the early phase of a qualitative study, re-searchers select a site and seek to gain entreinto it and into the specific settings in whichdata collection will occur. Gaining entre typi-cally involves enlisting the cooperation of gate-keepers within the site.

The research design of qualitative studies is typ-ically an emergent design. Once in the field, re-searchers select informants, collect data, andthen analyze and interpret them in an iterativefashion; field experiences help in an ongoingfashion to shape the design of the study.

Early analysis leads to refinements in samplingand data collection, until saturation (redun-dancy of information) is achieved.

Qualitative researchers conclude by disseminat-ing findings that can subsequently be used to (1)shape the direction of further studies (includingmore highly controlled quantitative studies); (2)guide the development of structured measuringtools for clinical and research purposes; and (3)shape nurses perceptions of a problem or situa-

SUMMARY POINTS

A basic distinction in quantitative studies is between experimental and nonexperimental re-search. In experimental research, researchersactively intervene or introduce a treatment,whereas in nonexperimental research, re-searchers make observations of existing situa-tions and characteristics without intervening.

Qualitative research often is strongly rooted in re-search traditions that originate in the disciplinesof anthropology, sociology, and psychology.Three such traditions have had strong influenceon qualitative nursing research: grounded theory,phenomenology, and ethnography.

Grounded theory seeks to describe and under-stand key social psychological and structuralprocesses that occur in a social setting.

Phenomenology is concerned with the livedexperiences of humans and is an approach tothinking about what the life experiences of peo-ple are like.

Ethnography provides a framework for study-ing the meanings, patterns, and experiences of adefined cultural group in a holistic fashion.

The steps involved in conducting a quantitativestudy are fairly standard; researchers usuallyprogress in a linear fashion from asking re-search questions to answering them.

The main phases and steps in a quantitativestudy are the conceptual, planning, empirical,analytic, and dissemination phases.

The conceptual phase involves (1) defining theproblem to be studied; (2) doing a literature re-view; (3) engaging in clinical fieldwork forclinical studies; (4) developing a framework andconceptual definitions; and (5) formulating hy-potheses to be tested.

The planning phase entails (6) selecting a re-search design; (7) developing interventionprotocols if the study is experimental; (8) spec-ifying the population; (9) developing a sam-pling plan; (10) specifying methods to measurethe research variables, through such approachesas self-report, observation, or the use ofbiophysiologic methods; (11) undertaking

tion and their conceptualizations of potentialsolutions.

STUDY ACTIVIT IES

Chapter 3 of the Study Guide to AccompanyNursing Research: Principles and Methods, 7thedition, offers various exercises and study sugges-tions for reinforcing concepts presented in thischapter. In addition, the following study questionscan be addressed:

1. In quantitative studies, the same measurementsare made of all subjects. What do you thinkresearchers are trying to achieve by this degreeof structure? Why might such structure not beappropriate in qualitative studies?

2. Which type of research do you think is easier toconductqualitative or quantitative research?Defend your response.

3. Suppose you were interested in studying fa-tigue in patients on chemotherapy. (This couldinvolve either a quantitative or a qualitativeapproach.) Suggest some possible clinicalfieldwork activities that would help you con-ceptualize the problem and develop a researchstrategy.

SUGGESTED READINGS

Methodologic References

Creswell, J. W. (1998). Qualitative inquiry and researchdesign: Choosing among five traditions. ThousandOaks, CA: Sage.

Glaser, B. G., & Strauss, A. L. (1967). The discovery ofgrounded theory: Strategies for qualitative research.Chicago: Aldine.

Kerlinger, F. N., & Lee, H. B. (2000). Foundations of be-havioral research. (4th ed.). Orlando, FL: HarcourtCollege Publishers.

Sterling, Y. M. (2001). The clinical imperative in clinicalnursing research. Applied Nursing Research, 14,4447.

Studies Cited in Chapter 3

Beck, C. T. (2002). Releasing the pause button:Mothering twins during the first year of life.Qualitative Health Research, 12, 593608.

Beck, C. T., & Gable, R. K. (2000). PostpartumDepression Screening Scale: Development and psy-chometric testing, Nursing Research, 49, 272282.

Beck, C. T., & Gable, R. K. (2001). Further validation ofthe Postpartum Depression Screening Scale. NursingResearch, 50, 155164.

Hauck, Y. L., & Irurita, V. F. (2002). Constructing com-patibility: Managing breast-feeding and weaningfrom the mothers perspective. Qualitative HealthResearch, 12, 897914.

Johnson, V. Y. (2001). Effects of a submaximal exerciseprotocol to recondition the pelvic floor musculature.Nursing Research, 50, 3341.

Powers, B. A. (2001). Ethnographic analysis of everydayethics in the care of nursing home residents withdementia. Nursing Research, 50, 332339.

Scannell-Desch, E. A. (2000). Hardships and personalstrategies of Vietnam war nurses. Western Journal ofNursing Research, 22, 526550.

Sundin, K., Norberg, A., & Jansson, L. (2001). Themeaning of skilled care providers relationship withstroke and aphasia patients. Qualitative HealthResearch, 11, 308321.

Wong, F., Ho, M., Chiu, I., Lui, W., Chan, C., & Lee, K.(2002). Factors contributing to hospital readmissionin a Hong Kong regional hospital. Nursing Research,51, 4049.


Documents

polit chapter 3