CHAPTER 3 RESEARCH METHODOLOGY
This chapter outlines the methodology that was used for this study, including the research design, setting, target population, sampling techniques, sample size, data collection plan and procedure, validity and reliability of the research instrument, pretesting of the research instrument, data analysis procedures and ethical considerations. This study was conducted in two phases. Phase 1, the quantitative phase, involved the use of structured interviews with women aged 42 and older attending clinics at three health centres in Blantyre; and phase 2, the qualitative phase, involved the use of semi-structured interviews with nurse/midwives providing healthcare services at the selected health centres. The results from a combination of these two phases have provided a comprehensive investigation of factors associated withwomen’s intentions to be screened for cervical cancer among women aged 42 and older in Blantyre, Malawi.
PURPOSE AND OBJECTIVES
The purpose of this study was to identifyand explore factors associated withthe intention to be screened for cervical cancer among Malawian women aged 42 and older in Blantyre, Malawi. This group of women was chosen because they are no longer in the child bearing age; therefore, they only come to the hospitals or clinics when sick. However, the literature review has shown that this group is at high risk of cervical cancer (Tsu& Levin 2008:105; Twinn, Holroyd, Fabrizio, Moore &Dickison 2007:16; Pollack, Balkin& Denny 2006:334). Women in the child bearing age groupget information on cervical cancer and are screened while being attended to for other health ailments. However, older women do not receive this information and might be unable to access this care at the stage of their lives when they are at high risk of cervical cancer.
The specific objectives guiding the study were to:
Identify demographic variables for women aged 42 and older and their association with these women’s intentions to be screened for cervical cancer.
Describe the association between perceived susceptibility to cervical cancer in women aged 42 and older and their intentions to be screened.
Determine the association between perceived seriousness of cervical cancer in women aged 42 and older and their intentions to be screened.
Determine the association between perceived benefits of cervical cancer screening in women aged 42 and older and their intentions to be screened.
Identify the association between perceived barriers to being screened for cervical cancer among women aged 42 and older, and their intentions to be screened.
Analyse the association between cues to action (sources of information) for cervical cancer screening among women aged 42 and older and their intentions to be screened.
Identify, from nurse/midwives’ perspectives, which factors are associated withintentions to be screened for cervical cancer among women aged 42 and older in Malawi.
A researcher faces certain fundamental problems that must be resolved before a study can be started. These include: who is to be studied, what is to be identified, when the identificationsshould be made and how data should be collected and analysed. The research design is the ‘blueprint’ that enables the researcher to come up with solutions to these problems and to guide the researcher in the various stages of the study (Nachmias&Nachmias 2004:99). A cross-sectional exploratory descriptive study, using quantitative and qualitative designs and methodsin different phases of the study, was used for the current study.
Triangulation of research methods and a combination of data sources from women and nurse/midwives allowed the researcher to attain a holistic examination of the research topic and enhanced confidence in the comprehensiveness of the study results. This is because different methods and designs generate different kinds of data. The advantages of integrating designs include: complementing each other, enhanced validity and creation of new information (Polit& Beck 2006:285). The quantitative part (phase 1) was consistent with the the HBM’s concepts,exploring women’s knowledge about factors associated with their intentions to utilise cervical cancer screening services. The qualitative component explored what nurse/midwives considered to be factors associated withtheir intentions to be screened for cervical cancer. Therefore, the combination of the two paradigms in this study allowed aninvestigation of the factors associated with intentions ofwomen aged 42 and older to utilise cervical cancer screening services.
PHASE 1: QUANTITATIVE COMPONENT
For this study a quantitative descriptive research method was used for phase 1. This methodology was implemented in order (for the researcher) to understand the full nature of the phenomenon of interest (women’s intentions to utilise cervical cancer screening services) and the other associated factors at three health facilities(Polit& Beck 2006:22). Structured interviews were conducted to collect data for phase 1 of the study.
A quantitative, exploratory, cross-sectional and descriptive research design involves observing, describing and documenting the behaviour of subjects without influencing it in any way. In a descriptive design the researcher describesthe results without intervening (Polit& Beck 2006:189). Descriptive studies are aimed at providing an accurate portrayal or account of the characteristics of individuals, events, or groups with the purpose of problem identification, justifying actions, making judgments, or finding out what others are doing in similar situations. The design does not involve manipulation of variables and there is no attempt to establish causality.
A cross-sectional, exploratory, descriptive study generates insights about a phenomenon. Data collection takes place during one data collection period from a cross-sectional sample, appropriate for the description of the phenomenon under study (Polit& Beck 2006:239).
Descriptive studies also aim at determining the frequency with which events occur and classifying information (Burns & Grove 2005:26). Phase 1 of this study explored thefactors associated with intentions to be screened for cervical cancer (among women aged 42 and older). A structuredinterview schedule was used to simultaneously collect data from these womenat three health centres.The activities in this phase included: recruitment and training of three research assistants, selection of study settings,selection of study population, sampling, data collection, data management and data analysis.
Study settings are specific places where information is gathered (Polit& Beck 2006:31). This study utilisedthree government health institutions as settings. In Malawi, there were eight government health centres offering cervical cancer screening services in the three major cities during the period of the study.Four of the health centreswere located in the southern region. Three sites in the southern region, which actively provided cervical cancer screening during the period of the study, were conveniently selected for the study. Convenientsampling is one of the non probability sampling methods that allows the researcher to deliberately choose the sources of data, knowing that they would contribute towards the required research information (Polit& Beck 2006:320), based on one or more known characteristics.
A population is an aggregate of all cases that meet specified criteria set for a study. These specific criteria of the population depend on the research problem. The total population in which the researcher is interested and to which he/she wishes to generalise results is called the target population, while an accessible population is part of the target population which the researcher can access (Burns & Grove 2005:341-342; Polit& Beck 2006:511).
The target study population for phase 1 of this study comprised women, aged 42 and older, attending government health service clinics providing cervical cancer screening services in the southern region of Malawi. The accessible population comprised women aged 42 and older attending clinics at Chilomoni, Ndirande and Bangwe health centres.
A sample is a subset of the population. Individual units of a sample and population are called elements. Elements can be in any form such as humans and events. The researcher works with samples rather than the entire population. Samples are smaller groups. Representative samples have all the characteristics of the population. The sample is drawn from the population that the researcher can access. However, findings of the study are initially generalised to the accessible population then to the entire population in similar settings (Polit& Beck 2006:260), provided the sample was representative of the population.
The researcher utilised a convenient sample of women aged 42 and older attending health service clinics at the three selected health centres during the study period. A convenient sample includes respondents that happen to be in the right place at the right time. The respondents were recruited into the study until the desired sample size had been achieved. The utilisation of a convenient sample, implied that the findings of this study might not be generalisable to the population of women aged 42 and older in Malawi.
Sampling refers to the process of selecting units from a population with one or more important attributes relevant to the research. The selected portion has to be representative of the entire population under study. A sample is considered representative if it has similar characteristics of the population from which it has been drawn. If all critical attributes in a population are identical to those in a sample, a sample is representative of the population. If the population is homogeneous, a sample would be sufficient for drawing conclusions about the population because the effect of the attributes on the variables under study in the sample and population would be the same. Sampling is more economical than studying the entire population, though it is a challenge for researchers to obtain representative samples (Burns & Grove 2006:351).
The quantitative research method may utiliseone of the two broad types of sampling: probability and non probability sampling. Non-probability sampling involves the selection of subjects by non random methods while in probability sampling elements are randomly selected from the population (Polit& Beck 2006:264). In this study, the researcher utilised a non probability sampling design to select the sample; a convenient sampling method. The convenient sampling method is used when the researcher selects the most conveniently available subjects of the population in question, or those particularly knowledgeable about the phenomenon of interest. The sample is selected based on the predetermined inclusion criteria (Polit& Beck 2008:309). The women who met the inclusion criteria were recruited to comprise the sample for the study.Convenientsampling impacts negatively on the generalisability of the research findings.
Sample selection procedure
The clinics at the government health centres operate on a non appointment system. Clients begin arriving at the clinics before 07:00 but clinic hours begin at 7:30. The researcher took the opportunity to utilise a few minutes for a “health talk session” to provide information about the study. The explanation included details about the purpose of the study, sampling and study procedure. This aimed tofamiliarise women with the study procedures and allowing them to make informed decisions about whether to participate or not. Interested women met the researcher in a room identified for data collection at the out-patient department after receiving the healthcare they required. In the data collection room, the volunteers were provided with detailed information regarding the study. They were also asked about their willingness to participate and sign an informed consent form, after responding to any questions they might have wished to ask. Those women who could not be interviewedon the sameday returned ata convenient date and time, arranged with them.
Selection of sample size
In quantitative research, the sample size can be determined by using statistical computations. This is based on the assumption that data would be available for calculation of the sample size. Some of the factors utilised in calculating sample size include: effective size, standard error, confidence interval (power) (Burns & Grove 2005:354-355; Sarantakos 2005:171) and the size of the total population from which the sample will be drawn (Saunders, Thornhill, & Lewis 2003:173). For this study the accessible population could not be predetermined because women aged 42 and older rarely attend healthcare clinics. However, the population of women aged 42 and older who had visited the three participating study sites during the previous year, was used as an estimate for the population of such women during the study period.
The researcher obtained a convenience sample of women, aged 42 and older, from those who sought health services from the three health centres during the study period. According to the National Cervical Cancer Coalition in Malawi (2008), the prevalence of cervical cancer among women was 28.0%. Considering this prevalence level nationwide and assuming a precision level of 3% with a confidence level of 95% the sample size required for this study would be 439 women. However, the final population correction factor indicated that 2 711 women, aged 42 and older, had received care for different health problems at Chilomoni, Ndirande and Bangwe health centres from July 2008 to August 2009 (MOH clinic attendance registers, Chilomoni, Ndirande, Bangwe July 2008-August 2009). Based on the calculation of the finite population for the three health centres, the sample size was reduced to 378.Other researchers have allowed for an anticipated non-responses rate of 10.0% (Turner 2003:2-13). However, in the current study, an additional three interviews were recommended by the statistician as the trained interviewers were expected to obtain responses to almost all questions. Therefore the total sample was 381.The sample size was calculated using the following formula:
The sample was distributed among the three health centres proportionally to the catchment area population of women in the child bearing age group. These proportions were used for calculation of the sample size because the estimated population for women aged 42 and older were not available. According to the MOH, the estimated population of women in the child bearing age for the Blantyre district 2010-2011 was 28 317, 45 940 and 17 407 for Bangwe, Ndirande and Chilomoni respectively. Therefore, the samples comprised 126, 190 and 65women respectively.
Inclusion and exclusion criteria for respondents
Inclusion criteria or eligibility criteria is a list of characteristics designated to delimit the study population while exclusion criteria include characteristics that study units must not posses in order to be involved in the study (Polit& Beck 2006:259). These criteria are defined by a researcher to promote the selection of an appropriate sample. According to Burns and Grove (2005:342), sources to determine the criteria include the research problem, the purpose of the study, the conceptual and operational definitions of the study variables and the study design. Inclusion criteria or eligibility characteristics for the study included women:
aged 42 and older (client or guardian to a client attending one of the health centres)
able to speak Chichewa or English
willing to participate and provide consent to participate in the study
The exclusion criteria included women:
aged 41 and younger
unable to speak Chichewa or English
who were critically ill
who had a total hysterectomy (removal of uterus and cervix) or Wertheim’s hysterectomy (surgical removal of uterus cervix and fallopian tubes).
Research instruments are devices used to collect data. These can be in a form of a questionnaire, test and an observation schedule (Polit& Beck 2006:502). Structured interviews, using interview schedules, were used for phase 1 of this study (See annexures E and G). The structured interview schedule may contain closed-ended questions with pre specified response alternatives by the researcher. According to Polit and Beck (2006:294), in structured interview schedules respondents are asked to respond to the same questions in the same order, and they are given the same set of options for their responses. However, structured interview instruments could also include some open-ended questions, which allow subjects to respond in their own words. In this case the interviewer writes down responses verbatim.
Structured interview schedule
The structured interview schedule for phase 1 was developed based on the study objectives, literature review and the HBM constructs. The researcher adapted questions from studies conducted by others, and included sections pertaining to: demographic characteristics, knowledge and awareness in relation to cervical cancer risk factors, symptoms and screening (Yare, Ozkilinc, Guler&Oztop 2008:283); level of knowledge (Gonzalez-Losa, Canto-Perèz& Puerto-Solis 2006:207); and barriers to cervical cancer screening (Holroyd, Twinn&Adab 2004:48). The interview schedule comprised six sections A to F, with the following subsections: demographic, socio-psychological and structural, perceived susceptibility to cervical cancer, perceived severity or seriousness of cervical cancer, perceived benefits of being screened for cervical cancer, perceived barriers to cervical cancer screening and cues to taking a cervical cancer screening test.
This instrument comprised close-ended and open-ended questions (See annexures E andG). One-on-one individual structured interviews were conducted. Closed ended questions required the interviewer to tick available responses and items but where options were provided for the respondent to give ‘other responses’ and ‘specify’ the interviewer was expected to write down the respondent’s answers verbatim. Items accompanied by an option of ‘true’ or ‘false’ or ‘do not know’ response to several declarative statements that express a point of view on a topic were also included. For the open-ended questions respondents were allowed to respond freely and the interviewer recorded the responses verbatim.
Section A: Modifying factors
According to Glanz, Rimer, Marcus and Lewis (2002:500), different modifying factors such as demographic, socio-psychological and structural variables may indirectly influence an individual’s perceptions. This, in turn, affects health-related behaviours such as cervical cancer screening. Some of these demographic factors might include, age, marital status, educational attainment, occupation, number of children, household income, religion and tribe. This section therefore, in questions 1.1 to 1.13 identified respondents’ socio-demographic information.
Section B: Perceived susceptibility to cervical cancer
Section B included questions that assessed respondents’ perceptions of being susceptible to cervical cancer. Questions were designed to determine whether eachrespondent intended usingcervical screening services in the near future; had undergone a cervical screening testpreviously; considered cervical cancer to be a health concern to her and whether she felt she should initiate the idea of having a cervical cancer screening test done. This data established how respondents perceived their risksof having cervical cancer, and their intentions of using cervical screening services. This was addressed in questions 1.10 to 2.21.
Section C: Perceived severity/seriousness of cervical cancer
This section determined how women aged 42 and older perceived the seriousness of cervical cancer. Questions enquired about women’s knowledge on the part of the body affected, signs and symptoms, whether they perceived the condition as a curse and how removal of the uterus would affect them. This was explored through questions: 3.1, 3.2, 3.3, 3.4, 3.5 and 3.6 – 3.15.
Section D: Perceived benefits of being screened for cervical cancer
This section determined perceived benefits of being screened for cervical cancer by women aged 42 and older. Areas assessed included knowledge of available diagnostic procedures, where they could access these services, purpose of the screening procedures, and availability for treatment of early stages of the condition. This was addressed through questions: 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7, 4.8, 4.9 – 4.16, 4.17, 4.18 and 4.19 – 4.21
Section E: Perceived barriers to cervical cancer screening
Section E concentrated on the exploration of perceived barriers to cervical cancer screening. Questions in this section were categorised into sub-sections E-1 to E-3 covering stigma, embarrassment, discrimination and fear; socio-economic factors; and traditional practices and the influence of traditional healers.
Sub-section E–1: Stigma, embarrassment and fear
This sub-section contained questions structured to determine social factors associated with women’s intentions to be screened for cervical cancer such as stigma, embarrassment and fear. These might affect regular attendance for screening and sharing of information for support purposes. This was addressed in questions 5.1 to 5.8.
Sub-section E – 2: Socio-economic factors
Questions 5.9 to 5.15 explored social and financial determinants associated with women’s intentions to utilise cervical screening services. The section covered areas such as discomfort, cost, and accessibility to cervical cancer screening services.
Sub-section E– 3: Traditional practicesandtraditional healers
This section comprised questions 5.16 to 5.19. They explored how respondents incorporated traditional beliefs in addressing cervical cancer-related problems. The information collected from this section assisted in identifying how the “use of traditional healers” is associated with early detection and treatment for cervical cancer.These could act as barriers to using cervical cancer screening services.
Section F: Cues to taking a cervical cancer screening test
This section explored cues to having cervical cancer screening testing done by Malawian women aged 42 and older. Questions were designed to determine respondents’ knowledge about available sources of cervical cancer screening information, preferred venues and methods of sharing cervical cancer screening-related information. This was addressed by items 6.1-6.3 and 11.7.
Table of contents
CHAPTER 1 INTRODUCTION AND BACKGROUND INFORMATION
1.2 BACKGROUNG INFORMATION
1.3 STATEMENT OF THE PROBLEM
1.4 THEORETICAL FRAMEWORK: THE HEALTH BELIEF MODEL(HBM)
1.5 RESEARCH METHODOLOGY
1.5 LAYOUT OF THE THESIS
CHAPTER 2 LITERATURE REVIEW
2.2 CERVICAL CANCER
2.3 THE HEALTH BELIEF MODEL (HBM)
2.4 FACTORS ASSSOCIATED WITH INTENTIONS TO BE SCREENED FOR CERVICAL CANCER IN RELATON TO THE HEALTH BELIEF MODEL’S (HBM’s) COMPONENTS
2.5 Contextualising the literature reviewed within the Health Belief Model
CHAPTER 3 RESEARCH METHODOLOGY
3.2 PURPOSE AND OBJECTIVES
3.3 RESEARCH DESIGN
3.4 PHASE 1: QUANTITATIVE COMPONENT
3.5 VALIDITY AND RELIABILITY
3.6 PHASE 2: QUALITATIVE COMPONENT
3.7 ETHICAL CONSIDERATIONS
3.8 LIMITATIONS OF THE STUDY
CHAPTER 4 DATA ANALYSIS AND DISCUSSION: PHASE 1: STRUCTURED INTERVIEWS CONDUCTED WITH WOMEN AGED 42 AND OLDER ON CERVICAL CANCER SCREENING
4.2 DATA ANALYSIS
4.3 PHASE 1: RESULTS: STRUCTURED INTERVIEWS CONDUCTED WITH WOMEN AGED 42 AND OLDER ON FACTORS ASSOCIATED WITH THEIR INTENTIONS TO USE CERVICAL CANCER SCREENING SERVICES
4.4 PERCEIVED SUSCEPTIBILITY TO CERVICAL CANCER
4.5 PERCEIVED SEVERITY OF CERVICAL CANCER
4.6 PERCEIVED BENEFITS OF CERVICAL CANCER SCREENING
4.7 PERCEIVED BARRIERS OF CERVICAL CANCER SCREENING
4.8 CUES FOR USING CERVICAL CANCER SCREENING SERVICES
4.9 SUMMARY OF DATA ANALYSIS FOR PHASE 1
CHAPTER 5 DATA ANALYSIS AND DISCUSSION: PHASE 2: SEMISTRUCTURED INTERVIEWS CONDUCTED WITH NURSE/MIDWIVES AT THREE HEALTH CANTERS PROVIDING CERVICAL CANCER SCREENING SERVICES IN BLANTYRE, MALAWI
5.2 DATA ANALYSIS FOR PHASE 2
5.3 THEMES AND CATEGORIES
5.4 DISCUSSION OF PRESENTED THEMES AND CATEGORIES
CHAPTER 6 CONCLUSIONS, LIMITATIONS AND RECOMMENDATIONS OF THE STUDY
6.2 CONCLUSION AND RECOMMENDATIONS FOR QUANTITATIVE
6.3 CONCLUSIONS AND RECOMMENDATIONS FOR PHASE 2 OF THE STUDY: RESULTS FROM INTERVIEWS WITH NURSE/MIDWIVES AT BANGWE CHILOMONI AND NDIRANDE HEALTH CENTRES
6.4 RECOMMENDATIONS FOR FUTURE RESEARCH
6.5 RECOMMENDATIONS FOR NURSING EDUCATION
6.6 RECOMMENDATIONS FOR THE HEALTHCARE SERVICES
6.7 RECOMMENDATIONS ON THE USE OF THE HEALTH BELIEF MODEL IN CERVICAL CANCER SCREENING PROGRAMME
6.8 COMPARISON OF STUDY FINDINGS FOR CHAPTER 4 AND 5
6.9 LIMITATIONS OF THE STUDY
6.10 SUMMARY OF THE STUDY.
6.11 SIGNIFICANCE OF THE STUDY
LIST OF REFERENCES
GET THE COMPLETE PROJECT