Data Collection Methods CHAPTER OVERVIEW Chapter 13 introduces the student to the various ways of collecting information about and from samples in order to measure variables of interest. Such methods are viewed as the process of operationalization (refer to Chapter 13 regarding operational definitions), and information should be obtained in both an objective and systematic manner. Four categories of data collection methods are presented: physiological or biological measurements, observational methods, interviews and questionnaires, and records or available data. The purposes, nature, use, and advantages and disadvantages of each method are thoroughly discussed, with published studies cited to illustrate their use in research. Of critical importance to the research design are the selection and evaluation of available instruments that measure the variables of interest. Selection is based on the operationalization of the variables and the literature review, because these steps in the research process provide the clues necessary to identify suitable instruments. New instruments may also be constructed in cases in which researchers cannot locate an acceptable instrument or method to measure the variable of interest. The authors point out that the evaluation of data-collection methods is often difficult for the new research consumer because in written research reports the instrument itself is usually not included. Despite this problem, judgements can be made about instruments, and several questions are presented for the student to consider. The aim of the critique is to determine whether the data-collection method is appropriate to the problem and whether the procedures are appropriate to the population studied. LEARNING OUTCOMES After reading this chapter, the student should be able to do the following: Define the types of data-collection methods used in nursing research. List the advantages and disadvantages of each of these methods. Critically evaluate the data-collection methods used in published nursing research studies.
TEACHING STRATEGIES It is important for students to develop an appreciation of the various ways that information is gathered in the context of carrying out a research design. This involves defi ning and operationalizing the variables, locating and selecting appropriate instruments to measure them, and assessing the overall “fi t” among the method, problem, and population studied. Since Chapter 13 of the Course Material addresses strategies for defi ning and operationalizing the variables, the following teaching strategies address instrument location, selection, and methodological fi t. 1. Critical Thinking Challenges The following questions can facilitate student comprehension and critical thinking about this chapter’s content: RECALL AND UNDERSTANDING What are the fi ve types of data-collection methods used in nursing research? What are the advantages and disadvantages of the fi ve types of data-collection methods used in nursing research? On what is the selection of data-collection instruments based? Explain the concept of consistency in data collection. ANALYSIS AND SYNTHESIS How should a nurse collecting data respond to a participant’s request for a specifi c nursing intervention? In what situations would interviews be more appropriate than questionnaires? How can bias in data collection be reduced? Can it be eliminated entirely? Explain. How can ethical concerns in data collection be addressed? 2. Using the Learner’s Previous Experience Students can evaluate the survey method from several vantage points based on their own research experience. For example, students can examine the process of questioning subjects in terms of use of an interview versus a questionnaire, use of open-ended versus close-ended questions, response style versus social desirability, interviewer bias, controls employed, and advantages and disadvantages of this method. This teaching strategy can be used for any data-collection method with which the student has had experience. 3. Using a Visual Summary Summarizing several concepts and principles that have been presented to students is often helpful to review and reinforce content. A visual summary not only serves this purpose but also shows the relationship between concepts and principles at a glance. 4. Differentiating Concepts: Measuring Variables It may be helpful to incorporate learning experiences that broaden students’ awareness of different ways to collect data, even if the same variable is being studied. Defi nitions and measurements can be developed differently, according to differing conceptual frameworks or even the same framework. Figure 13-1 lists the variable “self-effi cacy” and selected research reports that examine this variable. Assign students to read the indicated research reports in preparation for a subsequent class. In class, ask students to identify, compare, and contrast the conceptual and operational defi nitions of the variable, as well as the data- collection method used, for each report. You may fi nd it useful to list their responses in columns to more easily illustrate the similarities and differences (refer to Table 13-1 below) FIGURE 13-1. Variables and Selected Research Reports a. Conceptualization of the Variable Variable: Self-effi cacy: “Self-effi cacy refers to personal judgements of how well a person believes they can perform specifi c behaviours in particular situations” ( Bandura, 1997, p. 484). “Self-effi cacy, confi dence in one’s ability to perform relevant behaviours in a particular situation” (Bandura, 1997, p. 171). “General self-effi cacy is the broad and stable sense of personal beliefs about competence to deal effectively with a variety of situations or tasks, without reference to any specifi c behavioural domain” (Bandura, 1997, p. 348). b. Operationalization of the Variable Chronic Pain Self-Effi cacy Scale (Anderson, 1995), measuring pain management, coping, and physical function Parenting self-effi cacy can be defi ned as a parent’s belief that he or she is capable of organizing and executing tasks related to parenting a child (de Montigny & Lacharite, 2005). Sherer et al. (1982) defi ned general self-effi cacy as a composite of a person’s past experiences with success and failure across a variety of situations, which results in a general set of expectations that a person carries into a new situation. c. Method Questionnaire (close-ended) Self-report inventory (scale) Discussion should focus initially on the variable and the conceptual and operational definitions to reinforce prior learning and set the stage for comparisons of data-collection methods. Then focus on the overall considerations for evaluating data-collection methods. These c riteria appear at the end of Chapter 13 (see textbook, p. 303).
5. Providing Experiential Learning Activities a. Locating Measurement Instruments The purposes of the following learning activity are (1) to give students an opportunity to experience the process and become knowledgeable about the rationale for selecting data-collection instruments and (2) to increase students’ facility with using library resources. Student objectives for the learning activity: Locate research articles/books describing measurement instruments that can be used for research in nursing. Identify the purpose, design, validity, and reliability of selected measurement instruments. Directions for implementation. Students should be directed to use online databases to locate one or more instruments. Encourage them to use measurements books or psychometric articles. Some examples follow: Bowling A. (2005). Measuring health: A review of quality of life measurement scales. Milton Keynes: Open University Press, 1991. Waltz, C.F.; Strickland, O.L.; Lenz, E.R. (2010). Measurement in nursing and health research (4th editions). Springer; New York. Lovibond, P. F., & Lovibond, S. H. (1995). The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behaviour Research and Therapy, 33, 335–343. Retrieved from http:// search.proquest.com.ezproxy.library.yorku.ca/ docview/65148490?accountid=15182 Hambrick, J. P., Turk, C. L., Heimberg, R. G., Schneier, F. R., & Liebowitz, M. R. (2004). Psychometric properties of disability measures among patients with social anxiety disorder. Anxiety Disorders, 18, 825–839. Retrieved from http://search.proquest.com.ezproxy.library. yorku.ca/docview/64616248?accountid=15182. In fact, you may wish to have students look at more than one source for a particular instrument to get a sense of how this material is presented by different authors. In examining these sources, students can identify the purpose(s) and design/ method of the instrument—for example, whether it is a questionnaire or scale—and also identify the statements about instrument validity and reliability. Students should be prepared to discuss their fi ndings in a subsequent class. Guidelines for discussion. Follow-up in the classroom can consist of discussing students’ fi ndings, including the stated purpose, design/method, validity, and reliability of the instrument(s). General impressions should be shared about the various sources for locating instruments—for example, clarity, usefulness, and other impressions—and the experience of actually locating them. b. Appraising Data-Collection Methods This learning activity is designed to help students identify the appropriateness, advantages, and disadvantages of various types of data-collection methods. Student objectives for the learning activity: Critically evaluate the use of various types of data-collection methods. Determine the appropriateness of the methods used in a published research study. Identify problems of reactivity, informed consent, and bias in data-collection methods. Directions for implementation. In addition to r eviewing Chapter 13 in the textbook, students should read at least one of the following nursing studies: Bryanton, J., Gagnon, A. J., Hatem, M., & Johnston, C. (2008). Predictors of early parenting selfeffi cacy. Nursing Research, 57(4), 252–259. Mayer, C., Andrusyszyn, M-A., & Iwasiw, C. (2005). Self-effi cacy of staff nurses for health promotion counseling of patients at risk for stroke. AXON, 26(4), 14–21. Zhenfeng, M., Faber, A., & Dubé, L. (2007). Exploring women’s psychoneuroendocrine responses to cancer threat: Insights from a computer-based guided imagery task. Canadian Journal of Nursing Research, 39(1), 98–115. After identifying the problem, hypothesis statements, and design used, students should do the following: Find the operational and conceptual definitions of each variable. Identify the method employed in the study. Determine whether the data-collection method is appropriate to the stated relation ship between variables and the population studied. Identify the advantages and disadvantages of the method used. Determine whether there are any problems with reactivity, informed consent, or bias. Suggest alternative methods for gathering similar information. Guidelines for discussion. Engage students in a discussion of the data-collection methods of the study and their appropriateness. In addition, assist students to determine alternative methods for gathering similar information from similar subjects, for example, through questionnaires, observation, physiological measurement, and other means identifi ed in the text.
Rigour in Research CHAPTER OVERVIEW Chapter 14 focuses on defining and clarifying the concepts of rigour in research for both qualitative and quantitative research. Rigour is the quality, believability or trustworthiness of the study findings. Rigour in quantitative research is determined by measurement instruments that validly and reliably reflect the concepts of the theory being tested, so that conclusions drawn from a study will be valid and will advance the development of nursing theory and evidence-informed practice. In qualitative research, rigour is ascertained by credibility, auditability, and fittingness. The chapter begins with a brief discussion of measurement error, defining the terms error variance, random error, and systematic error. The term instrument validity is defined as the accuracy with which a measurement tool measures the concept it is intended to measure. Several types of validity—content, face, criterion-related, and construct—are described. The authors present four major approaches to testing construct validity: hypothesis testing, convergent and divergent, contrasted groups, and factor analytical. Reliability is defined as “the extent to which the instrument yields the same results on repeated measures,” and the major forms of testing reliability are presented. These are test–retest reliability, internal consistency or homogeneity, interrater reliability, and alternate or parallel forms. Qualitative researchers seek to achieve two goals: to account for the method and the data, which must stand independently so that another researcher can analyze the same data in the same way and make the same conclusions, and to produce a credible and reasoned explanation of the phenomenon under study. Thus, this rigour in qualitative methodology is judged by unique criteria appropriate to the research approach. Credibility, auditability, and fittingness are the scientific criteria proposed for qualitative research studies by Guba (1981). Although these criteria were proposed two decades ago, they still capture the rigorous spirit of qualitative inquiry and persist as reasonable criteria for evaluation. The chapter ends with a discussion of how to critique the rigour in research described in a research report. Criteria to guide such a critique are presented. LEARNING OUTCOMES After reading this chapter, the student should be able to do the following: Discuss the purposes of reliability and validity. Define reliability. Discuss the concepts of stability, equivalence, and homogeneity as they relate to reliability. Compare the estimates of reliability. Define validity. Compare content, criterion-related, and construct validity. Discuss how measurement error can affect the outcomes of a research study. Discuss the purpose of credibility, auditability, and fittingness. Identify the criteria for critiquing the reliability and validity of measurement tools. Use the critiquing criteria to evaluate the reliability and validity of measurement tools. Discuss how evidence related to research rigour contributes to clinical decision making. CHAPTER 14 Rigour in Research 77
TEACHING STRATEGIES We have found that students have diffi culty grasping reliability and validity as concepts and applying their knowledge of these concepts to c ritiquing the instrument section of research reports. It is therefore advisable to approach this content area in an organized and unhurried manner. The following teaching strategies have helped our students understand and apply principles of reliability and validity. 1. Critical Thinking Challenges The following questions are suggested for review of this chapter’s concepts and as a stimulant to critical thinking about their application: RECALL AND UNDERSTANDING • What is the meaning of validity? How would you defi ne the following types of validity? Content Criterion-related Concurrent Construct Predictive What is the meaning of reliability? How would you defi ne the following types of reliability? • Internal consistency Test–retest Interrater Parallel forms What is the meaning of credibility, auditability, and fi ttingness? ANALYSIS AND SYNTHESIS How do concepts of validity and reliability apply to the assessment and diagnosis of patient health problems? Conduct of quality assurance studies? If the author of an instrument states only that the instrument has content validity, what additional information would you need to determine the truth of that statement? How would the fi ndings of a study be affected if the researcher used an instrument that was found to produce systematic error? What is the relationship between validity and reliability? 2. Clarifying Concepts and Principles: The Appropriate Use of Reliability Tests Students often misunderstand the indications for using different tests of reliability to support an instrument’s stability, accuracy, or equivalence. We have found that the appropriate use of test–retest and interrater reliability procedures particularly needs clarifi cation. Because test–retest reliability is intended to demonstrate the stability or consistency of a measure, it is inappropriate for use as an instrument intended to measure a nonstable construct. In clarifying this concept for students, we often use the State-Trait Anxiety Inventory (Spielberger, et al., 1970) as an example. According to the definitions presented by Spielberger and colleagues, “State anxiety (A-State) is … a transitory emotional state or condition of the human organism that is characterized by subjective, consciously perceived feelings of tension and apprehension, and heightened autonomic nervous system activity. A-States may vary in intensity and fl uctuate over time.” On the other hand, “Trait anxiety (A-Trait) refers to relatively stable individual differences in anxiety proneness. …” Because A-State is transitory, test–retest reliability would be expected to be low. But because A-Trait is conceptualized as a stable characteristic of an individual, test–retest procedures should yield a high correlation coeffi cient. (Results of such testing with this instrument are as follows: for the A-State portion of the inventory, correlations ranged from 0.16 to 0.54; for the A-Trait, they ranged from 0.73 to 0.86.) Interrater reliability is intended to demonstrate the equivalence or agreement among raters who are collecting data. The misconception that arises in relation to this type of reliability is that it is appropriate for self-administered tests or questionnaires. Therefore in a discussion of types of reliability, it is important to emphasize the types of instruments or measures for which various reliability tests are appropriate. It may be helpful to distinguish between subjects and data collectors for the purpose of clarifi cation. Interrater reliability is appropriate when the subjects in a study are being observed or rated by a researcher or assistant. It is not appropriate when the research subjects are rating their own behaviour, perceptions, opinions, or attitudes. An additional problem with the concept of interrater reliability is dealt with in strategy 3 below. 3. Using Case Studies Another potential problem area is the development of an appreciation for the need to establish interrater reliability in studies using more than one individual to observe or rate a dependent variable or outcome behaviour. One of the most interesting phenomena we have encountered is that students, as well as colleagues with similar backgrounds and interests, fi nd it diffi cult to believe that sometimes they can see things differently. An understanding of the individuality or subjectivity of perceptions is needed to appreciate the need for establishing objectivity in measurement, particularly if observational or rating methods that involve more than one data collector are used. In our experience, internalization of this principle is facilitated by experiential learning. One strategy for achieving this goal is to have students assess the nursing care problems of a patient—that is, formulate nursing diagnoses and compare their assessments. Two case studies for possible use are presented here. There will probably be disagreements among students in the class. Encourage students to discuss their differences and attempt to reach consensus. Bear in mind that “right” answers are not as important as reaching consensus. At the completion of this exercise, relate the process the class has experienced to the need for and establishment of objectivity in measurement, especially in relation to interrater reliability. CASE STUDY 1 Mrs. R., a slim 66-year-old woman, lives alone and works full-time in a bookstore. She has been a widow for 10 years. She has no children. Mrs. R. was admitted to the hospital on February 22 with a diagnosis of rectal cancer. Her physician told her before admission that she had cancer and would need a colostomy. An initial assessment of Mrs. R.’s understanding of the nature of her illness r eveals that she perceives surgery as a lifesaving p rocedure. She expresses her desire to continue leading an independent life. She took care of both her mother and her husband before they died. She keeps referring to their dependence on her. She says, “I don’t want to be dependent like they were. I want to be able to take care of myself.” She expresses her goals as wanting to return to her own home after discharge and continue at her job after recuperation. Preoperatively, Mrs. R had several misconceptions about what a colostomy was and the effect it would have on her life. For example, she loved to swim, but stated initially that it would be out of the question after surgery. She asked many questions during the preoperative teaching sessions and learned quickly. Thus before surgery, Mrs. R. had thorough understanding of the anatomy and physiology of the colon, how the colostomy would alter her anatomy, and the self-care she would need to learn. She read a pamphlet and expressed great excitement at the fact that she would still be able to swim. Diet teaching was begun but did not seem to be of immediate concern to her. She would refocus on the care of the stoma itself.
CHAPTER 14 Rigour in Research 79
On February 26, Mrs. R. had an abdominoperineal resection and sigmoid colostomy. No evidence of metastasis was found. As soon as she was able to move around comfortably, she began asking the nurses what they were doing when they changed her temporary appliance. She was able to look at her stoma in a mirror and watch. At the last teaching session on February 29, she was able to clean the skin around her stoma and apply a skin protective. She had some diffi culty centering her stoma in the middle of the faceplate opening. She stated, “I can’t seem to get this. Will I ever be able to do this? How can I go home if I can’t put this thing on?” Mrs. R. also began wondering about the gas that was frequently b eing eliminated from her colostomy. She had not yet begun to eat solid foods. She expressed concern over others noticing “something unusual” about her. Her perineal wound was healing. There was no evidence of infection or unusual drainage (Levin, 1981). CASE STUDY 2 Sally S. is a 7-year-old white girl who lives with her parents and 5-year-old sister in a semisuburban middle-class community. She is recuperating from a femoral leg-lengthening procedure used to correct unilateral congenital shortening of the femur. After 8 weeks of hospitalization, characterized by signifi cant pain, she is discharged home with pins and bar still in place. She is able to bear weight on the unaffected side and ambulate with a walker. Care is to be provided by her parents with periodic visits by the pediatric nurse practitioner (PNP). At the fi rst home visit (5 days after discharge) the PNP notes an untidy home. Mrs. S. appears unkempt and moderately fl at of affect. Sally is lying on a cluttered but clean couch watching television. She appears thin. A 2.5-kilogram weight loss in 6 weeks is reported. (Sally was 120 cm tall and weighed 22 kg before surgery.) When questioned as to Sally’s food intake, Mrs. S. states that the child “barely touches her meals,” but drinks a lot of juice and soda during the day. Although Sally is initially friendly, she is soon distracted and resumes watching television. Dressings at the pin sites are changed by Mrs. S., who meticulously observes clean technique. The PNP observes Sally demand a r itualistic progression of steps: “Now pour the peroxide. No, not that way. Don’t rub the pin yet.” She is essentially willing to endure the procedure but is rather “whiny.” The skin surrounding the pin sites is clean and intact. The PNP cultures the pin site. As the PNP is preparing to leave, Mrs. S. states, “It’s no fun to be with Sally anymore. I always have to be the heavy; my husband can just play with her. I can’t get her to eat, and she always wants to watch TV.” Mrs. S. also confi des that she doesn’t seem to have time for her younger daughter, who since Sally returned from the hospital, has remained in her room most of the time complaining of stomachaches and headaches and has missed several days of school. She also has begun to bite her nails and wet her bed on occasion. 4. Applying Concepts and Principles: Measurement of Variables The ultimate objective of this chapter is for students to apply critiquing criteria to the evaluation of the measurement aspects of a research report. Your demonstration of the application of these criteria in class can reinforce learning of concepts and principles, as well as provide students with an example of how this task is accomplished. In carrying out this strategy, it may be helpful to critique a well-developed discussion of the measurement aspects of a research report, as well as one that has obvious weaknesses. You may choose to present this material in didactic form or provide for a s tructured class discussion in which you take an active role. Suggested studies for your use follow: Anderson, E. H. (2000). Self-esteem and optimism in men and women infected with HIV. Nursing Research, 49(5), 262–271. McCurren, C., et al. (1999). Depression among nursing home elders: Testing an intervention strategy. Applied Nursing Research, 12(4), 185–195. 5. Providing Experiential Learning Activities a. Critique of a Measurement Tool The purpose of this activity is to give students experience in critiquing a tool that has been developed to measure a construct of concern to nurses. Student objectives for the learning activity: Evaluate the conceptual basis for tool development. Assess the operational defi nition of the construct to be measured. Determine whether the items refl ect the defi nitions of the construct. Identify the kind of reliability that was established and assess its adequacy. Assess the strengths and weaknesses of the tool in relation to its intended use. Directions for implementation. Students should review Chapter 14 of the textbook, particularly the “Critiquing Criteria” section (p. 328). Using these criteria, direct students to evaluate a tool that has been developed for measuring constructs in nursing research. A number of instrument- development articles can be found in issues of Nursing Research, Research in Nursing and Health, Advances in Nursing Science, and Applied Nursing Research. The following studies are suggested: Clarke, P. (2008). Validation of two postpartum depression screening scales with First Nations and Métis women. Canadian Journal of Nursing Research, 40(1), 112–125. (Quantitative) Fillion, L., Gélinas, C., Simard, S., Savard, J., & Gagnon, P. (2003). Validation evidence for the French-Canadian adaptation of the Multidimensional Fatigue Inventory as a measure of cancer-related fatigue. Cancer Nursing, 26(2), 143–154. (Quantitative) Harwood, L., Ridley, J., Lawrence-Murphy, J. A., White, S., Spence-Laschinger, H. K., Bevan, J., & O’Brien, K. (2007). The CANNT Journal, 17(2), 35–43. (Quantitative) Heaman, M., Chalmers, K., Woodgate, R., & Brown, J. (2007). Relationship work in an early childhood home visiting program. Journal of Pediatric Nursing, 22(4), 319–330. (Qualitative) Tobin, G. A., & Begley, C. M. (2004). Methodological rigour within a qualitative framework. Journal of Advanced Nursing, 48(4), 388–396. (Qualitative) Wharf- Higgins, J., Young, L., Cunningham, S., & Naylor, P-J. ( 2006). Out of the main stream: Low-income, lone mothers’ life experiences and perspectives on heart health. Health Promotion Practice, 7(2), 221–233. (Qualitative) This learning activity may be approached in two ways—either by having the entire class participate in a critique of the study(ies) in seminar fashion or by dividing the class into groups. With the latter approach, each group would present its fi ndings to the class. Class discussion should be stimulated by you, as well as by the presenters.
Qualitative Data Analysis CHAPTER OVERVIEW Chapter 15 examines how analysis is completed in qualitative research. As explained early in the chapter, there is no specific canon for qualitative analysis; however, there are some established methods that are important for the research consumer to be aware of. The author begins with a discussion about the data focusing on interviewing and data management. An overview of the process of analysis follows, including data reduction, coding, data display, and conclusion drawing. Specific methods for phenomenology, ground theory, ethnography and case study are also explored. The chapter concludes with a discussion about trustworthiness of the analysis and conclusions. In general, this chapter allows the novice consumer of research to understand how analysis is completed in the qualitative research domain. LEARNING OUTCOMES After reading this chapter, the student should be able to do the following: Describe the processes of qualitative data analysis. Outline the steps common to qualitative data analysis. Describe how data are reduced to meaningful units (themes). Describe the process of identifying themes and categories and the relationships between them. Assess the validity of a data analysis from a study.
• CHAPTER 15 Qualitative Data Analysis TEACHING STRATEGIES As a teacher of nursing students, you can facilitate your students’ understanding of qualitative analysis. It is a very complex process; thus it is expected that students may express some frustration with the process. It is diffi cult for students to appreciate the process if they are not immersed in the data collection, transcription, and analysis. The teaching strategies focus on allowing student to be comfortable with a basic understanding of the process and an awareness of the more common methods and terminology. 1. Critical Thinking Challenges The following lists of questions are intended to facilitate review of chapter content and to foster the critical thinking skills needed to fulfi ll the role of research consumer: RECALL AND UNDERSTANDING What is the basic text analyzed in qualitative research? What is the common form of data? When a researcher uses interviews, what steps must be taken to collect the data? What are the basic steps of analysis? How is trustworthiness established? ANALYSIS AND SYNTHESIS How does qualitative analysis differ from quantitative analysis? What are some issues with qualitative analysis? How does the researcher ensure that the data analysis is trustworthy? 2. Reading and Understanding Some Qualitative Analysis Examples Have the students work in groups. They must choose two qualitative studies from the appendix in the textbook or from the following list: Fournier, B., Mill, J., Kipp, W., & Walusimbi, M. (2006). Discovering voice: A participatory action research study with nurses in Uganda. International Journal of Qualitative Methods, 6(2). Retrieved from http://www.ualberta. ca/∼iiqm/ O’Donnell, M. E. (2000). The long gray tunnel: The day-to-day experience of spouse caregivers of people with Alzheimer’s disease. Scholarship & Inquiry in Nursing Practice, 14, 47–71. Sword, W. (2003). Prenatal care use among women of low income: A matter of “Taking care of self.” Qualitative Health Research, 13(3), 319–332. Varcoe, C. (2001). Abuse obscured: An ethnographic account of emergency nursing in relation to violence against women. Canadian Journal of Nursing Research, 32(4), 95–115. As they individually read the articles, ask them to focus on the following: What is the specifi c analytic method used? Did they understand how the researchers found their results? Are the fi ndings useful for the practice of nursing? Do the author(s) use examples of text to establish trustworthiness with the reader? Have the students form groups of three or four to share their fi ndings. Ask each group to summarize their discussion to share with the class. 3. Providing Experiential Learning Activities A. Interviewing Have the students interview each other with a s pecifi c series of questions. Cut the interview after 10 minutes as they will be working with the transcript. Each dyad will need an audio-r ecorder so you may wish to assign this as homework. Students can borrow audio-recorders from the media centre or library. A sample interview guide is included, but you may develop an alternative research question and guide. RESEARCH QUESTIONS: What makes nursing students anxious about practicum/clinical placements?
CHAPTER 15 Qualitative Data Analysis 83
Can you tell me about your practicum/clinical placements? Did you ever feel anxious before or during your placement? Probing questions such as: Can you tell me more about that? What does anxiety feel like for you? Is your anxiety different in some placements? Do you fi nd there is a difference as you gain more experience? B. Transcribing Once the students have their interview, ask them to listen to the tape and transcribe the interview. They can take note of themes as they transcribe. C. Analysis Using their transcripts from the previous exercise, ask them to code the data and come up with some themes. D. Sharing of Findings Have the students share some of their themes with the rest of the class. Are there common themes? What are the fi ndings? Do the fi ndings feel true to the students? Do they generally represent their experiences? Could these fi ndings be used by clinical instructors to help reduce student a nxiety? E. Data Display As a class, attempt to come up with a visual display of the themes. 4. Confirmation of Analysis Have the students compare their analysis to that of other researchers. You can use the transcript from Chapter 15 in the text, from your own research, or Web sites such as: http://hsc.uwe.ac.uk/dataanalysis/qualText- DataEx.asp A suggested interpretation of the transcript is provided on the Web site. Ask students how close they came to the same interpretation as their peers or the Web site. What happens when interpretations differ? What steps can researchers take to come up with common interpretations? How can the fi ndings be verifi ed?
Quantitative Data Analysis CHAPTER OVERVIEW Chapter 16 discusses both descriptive and inferential statistics. Descriptive statistical procedures are used to summarize data in terms of their meaning, use, and limitations. These procedures are presented in sufficient detail to enable the consumer of research to evaluate their appropriate use. Initially, nominal, ordinal, interval, and ratio levels of measurement are defined in relation to the nature of the object or event being measured and the use of each particular scale. Examples of scale measurements are provided to clarify and bring relevance to the discussion. In this chapter, complex topics such as frequency distributions, measures of central tendency, and measures of variability are presented in a form that is palatable to undergraduate and master’s students. Tables and figures facilitate the comprehension of this material. Regarding the critique of the statistical aspects of a research study, it is pointed out that the nursing student does not need an extensive background in statistics to critically evaluate whether the procedures for summarizing data are sensible in light of the study’s purpose. For example, the student can evaluate the “match” between a specific measurement scale and the statistic used to summarize data, the clarity and relevance of results, the relevance of the summary statistic to the study purpose and hypothesis, and the appropriateness of the sample size. Inferential statistics are used to draw conclusions about populations from sample data. Basic concepts and terminology about inferential statistics are presented to give students the appropriate background information to understand their purpose and application. For example, there is a brief introduction to probability theory and a section on type I and type II errors in statistical inference. Parametric and nonparametric tests of statistical significance are differentiated; parametric tests are used more often in the nursing research literature. Tests of difference, tests of relationships, and advanced statistical tests are discussed as to their appropriateness to various research designs and
hypotheses. Because this is a complex, confusing subject, a previously described research study example is employed to provide depth of knowledge and clarity to the use of statistical tests of significance for hypothesis testing. In addition, when critiquing criteria are identified, it is pointed out that, although the critique can be difficult, aspects of the data analysis can be addressed without the benefit of a statistics course. Such indicators as the hypothesis statement, level of measurement, sample size, presented results, clarity, and distinctions made between practical and statistical significance can be used by the research consumer to judge the adequacy of the analysis. LEARNING OUTCOMES After reading this chapter, the student should be able to do the following: Define descriptive statistics. State the purposes of descriptive statistics. Identify the levels of measurement in a research study. Describe a frequency distribution. List the measures of central tendency and their use. List measures of variability and their use. Identify the purpose of inferential statistics. Distinguish between a parameter and a statistic. Explain the concept of probability as it applies to the analysis of sample data. Distinguish between type I and type II errors and their effect on a study’s outcome. Distinguish between parametric and nonparametric tests. List the commonly used statistical tests and their purposes. Critically analyze the types of statistics used in published research studies.
TEACHING STRATEGIES Chapter 16 of the textbook presents the procedures used to describe and summarize data, i ncluding measures of central tendency, measures of variability, and some correlation techniques. The following teaching strategies have been d eveloped to enhance students’ understanding of the meaning and use of these procedures. The presentation of inferential data-analysis techniques is conceptual in nature. Therefore, knowledge of the mathematical operations involved in statistical analysis is not essential for achieving the chapter’s learning objectives. In keeping with this approach, the following teaching strategies focus on helping students to develop an understanding of the purpose of simple data-analysis techniques and an ability to evaluate the appropriateness of researchers’ use of these techniques in published studies. Students usually associate data analysis with mathematical operations and, as a result, approach this topic feeling anxious. It is suggested that you reinforce the need for a conceptual rather than a mathematical understanding of inferential statistics so as to implement the research consumer role. 1. Critical Thinking Challenges The following questions are designed to facilitate students’ integration of this chapter’s content and provide an impetus for thinking critically about inferential data analysis. RECALL AND UNDERSTANDING • What are descriptive statistics? What are the four levels of measurement used to quantify variables? What does a frequency distribution tell us? What do measures of central tendency tell us? What is a null hypothesis? How would you defi ne type I and type II errors? What is the difference between a parametric and a nonparametric statistic? If a researcher wants to test the relationship be- tween two variables (or the difference between two groups), which statistical tests would be appropriate? If the level of signifi cance for hypothesis testing in a study is set at p = .05 and the results of data analysis are found to be signifi cant at p = .07, what can the researcher conclude? ANALYSIS AND SYNTHESIS Why is it desirable to have at least intervallevel data? Why is the mean the preferred measure of central tendency? Why is the range an unstable measure of variability? What are measures of variability intended to describe? When determining the level of signifi cance for hypothesis testing in a study, what factors should a researcher take into account? A researcher is interested in fi nding out whether drinking ice water has any adverse effects on patients who have had myocardial infarctions. Which type of error is more important to control for in this study? Why? If the purpose of a study is to describe the characteristics of a population, and no predictions are made, what type of statistical tests would be appropriate? Why? 2. Allaying Students’ Anxiety Despite the conceptual nature of the presentation of statistical techniques in the textbook, students may still believe that knowledge of the mathematical operations involved in data analysis is necessary to understand the related content. We have found it helpful to allay students’ anxiety about this by saying at the beginning of class that the material to be covered involves concepts and principles—and that it is not necessary for them to know how to perform the statistical tests that will be discussed. It can also be reassuring to students if you share the fact that most researchers today do not perform statistical tests by hand, but instead use computers (and statisticians) to analyze their data. 3. Using Examples of Descriptive Statistics It is useful to provide numerous examples to facilitate students’ learning of the appropriate use of descriptive statistical techniques. Table 16-1 lists several variables, levels of measurement specifi c to each variable, and appropriate descriptive statistical technique(s) that can be used to summarize data collected for each variable. You can present several of these examples, involving students in the process of arriving at the appropriate conclusions. Subsequently, you can name other variables and ask students to identify their levels of measurement and appropriate descriptive statistics. 4. Providing Experiential Learning Activities a. Organizing Data: The Frequency Distribution This learning activity is designed to give the student a better understanding of the frequency distribution. Student objectives for the learning activity: Develop a frequency distribution. Identify the shape of the developed frequency distribution. Directions for implementation. Students can use the surveys they completed earlier (see C hapter 13 of the Instructor’s Manual) to implement this
a ctivity. Data from 20 individuals were collected in relation to their age. These data form the basis for developing a frequency distribution for age. Students plot the ages on either of the forms shown in Figure 16-1 (p. 352) in the textbook (polygon or histogram). They also can identify the age range, mean, median, and mode to give them a fuller understanding of the data collected. Students should work in groups of four to combine data, thereby increasing the possibility of constructing a graph that is somewhat smooth in appearance. Once the graph is completed, its shape can be described in terms of symmetry, modality, and kurtosis. Figures 16-2 (p. 353), 16-3 (p. 354), and 16-4 (p. 355) in the textbook can help the student evaluate the shape of the distribution. The data, graph, and description of the shape of the distribution can be submitted to you for review and commentary. Grading is optional. b. Determining and Interpreting Descriptive Statistical Techniques Through this learning activity, students have an opportunity to determine the appropriate use of descriptive statistical techniques, in this case measures of central tendency, to summarize data. Student objectives for the learning activity: Identify the levels of measurement of selected variables. Determine the appropriate descriptive statistic to summarize data of selected variables. Directions for implementation. Again, students can use the surveys they completed as a basis for identifying levels of measurement and determining appropriate descriptive statistical techniques. Data from 20 subjects were collected in relation to the following items: Whether or not they have an annual health examination Their opinions about using additives How they estimate their pain tolerance (low, moderate, or high) Their opinions regarding their current health status (poor, fair, good, or excellent) Their exact age For each item, students identify the variable, level of measurement, and an appropriate measure of central tendency. Although it is not stated as an objective, you may wish to ask students to examine the data grossly and identify possible fi ndings. An example from Table 16-2 below can be used to initiate this learning activity, after which students independently meet identifi ed objectives. Guidelines for discussion. In addition to the presentation of each item in relation to the objectives of this activity, students should defend the descriptive statistic they selected for summarization of each variable. c. Calculating Measures of Central Tendency Using Hand Calculation and a Computer Once data are summarized using a frequency distribution (see Strategy 4a above), this learning activity gives students an opportunity to elicit more statistical information to describe the data collected. Student objectives for the learning activity: Defi ne measures of central tendency. Calculate the mean, median, and mode, using hand calculation and a computer. Directions for implementation. Hand calculation. Use the data from one of the groups of four students (n = 80) and determine the mean, median, and mode. Put all data on the blackboard and ask students how each measure is determined. Then do the actual calculations with the students. Computer application. Use the same data to determine the mean and standard deviation with SPSS (Statistical Package for the Social S ciences) for Windows software. (SPSS is probably the most popular software package in use for conducting statistical analysis. Thus it would be wise for you, as an instructor of research, to learn the program and use it to help students learn how to interpret descriptive and inferential data analyses.) Given the objectives of this unit, it makes the most sense for the instructor to provide students with an SPSS printout of the data and then help them interpret the data on the printout. You also may consider running a cross-tabulation between two of the above variables—for example, annual heath examination and opinions regarding current health status—and helping students read the results of this analysis. This program can also be used to conduct inferential statistical tests. Guidelines for discussion. Involve students in a discussion of how these measures of central t endency provide further information (more than just a frequency distribution) about the data. E xtend this discussion by identifying what information measures of variability provide. Finally, discuss the advantages and disadvantages of hand c alculation and computer application with respect to descriptive statistics. If you have run cross-tabulations, work with students to interpret the results of data analysis. 5. Using Examples of Relationships Between Hypotheses, Levels of Measurement, and Inferential Statistics One way to facilitate students’ learning of the appropriate use of the statistical tests presented in this chapter is to provide numerous examples of hypotheses that have variables at different levels of measurement and for which various inferential statistical tests are indicated. After you actively present some examples, involve students in identifying the statistical test of choice for others. Encourage students to state their rationale for an identifi ed test, specifying the level of measurement for each variable. Table 16-1 (p. 348) in the textbook provides suggestions. 6. Providing Experiential Learning Activities for Inferential Statistics a. Testing a Hypothesis The following exercise, based on a strategy described by Johnson (1984), is a means of reviewing previously learned research concepts and principles, as well as integrating this content with new concepts of inferential statistics. Specifi cally, the Pearson product moment coeffi cient of correlation (r) is used to test a hypothesis. Student objectives for the learning activity: Identify the levels of measurement of variables in a hypothesis. Determine the method of sampling to be used. Identify the appropriate statistic for hypothesis testing. Identify the level of signifi cance for hypothesis testing. State the probability of committing a type I error. Relate the fi ndings to acceptance or rejection of hypothesis. Directions for implementation. You will need to devote portions of two class periods to complete this activity. During the fi rst class, present students with a hypothesis that predicts a relationship between two continuous variables, which can be measured at the interval or ratio level and for which students can provide data. For example, there is a positive correlation between height and weight. Ask students to identify a level of measurement of each variable, the statistic to be used, and the level of signifi cance for hypothesis testing. In addition, ask students to state the probability of committing a type I error. Coach them if needed, but try not to provide the answers. This portion of the exercise should be a review of previously covered content. Depending on the size of the class, you can use one of two approaches to sampling. If the class is large enough (over 20), random s ampling can be used to determine which students can provide data to test the hypothesis (at least 10). If the class is small, a convenience sample of all students will provide the data. Involve students in identifying the appropriate sampling technique to be used, stating the rationale for their choice, and discussing the implications of the method that is chosen. In either case, each student will have a subject number. Ask those students who are “subjects” to write their number, height, and weight on an index card; then collect the cards. Use the method most convenient for you to analyze the data. If you are good with numbers, you may prefer to do the necessary calculations by hand. Some desktop and pocket calculators can do a simple correlation. It is probably not worthwhile to use computer software to analyze data from such a small sample. During the second class, present students with the results of the data analysis. This can include a review of descriptive statistics (mean, range, and standard deviation), as well as the presentation of the correlation coeffi cient and level of signifi cance. Showing the data graphically—that is, a visual display of the distribution of their heights and weights as well as a scattergram—can facilitate their understanding of what the correlation coeffi cient actually means. Transparencies can be used for this purpose. Guidelines for discussion. After the data are presented, involve students in a discussion of whether the hypothesis is supported or not supported and the possible reasons for each outcome. Relate their response to principles of design and sampling theory. Allow 30 to 45 minutes for each portion of the exercise (Class 1: presentation of the hypothesis through data collection; Class 2: presentation of results and discussion), depending on the depth of discussion required to achieve the learning objectives with your particular class. b. Critical Appraisal of Inferential Statistics The purpose of this exercise is to help students distinguish the uses of the various inferential statistics. In addition, it provides practice in critiquing the use of inferential statistics in published research. Student objectives for the learning activity: Distinguish among different types of inferential statistics and their use. Identify the relationship among the problem, the hypothesis(es), the design, the level of measurement of variables, and the statistical tests used in nursing studies. Critique the presentation of statistical results. Directions for implementation. Students should study Chapter 16 of the textbook and review any notes they may have taken in class. In addition, assign a research study that uses statistics with which students are familiar, for the class to read at home. (Suggested studies are listed below.) Together with students, identify the problem statement, the hypothesis(es), the variables and their level of measurement, the sample size, and the design. Working in small groups (three to fi ve students per group), students should be directed to do the following: Identify the statistical test(s) employed in the study. Determine whether the researcher’s choice of test(s) was appropriate by examining the other aspects of the study as specifi ed previously. Critique the presentation of the results according to the critiquing criteria presented in the textbook (p. 380). A representative can present the group’s conclusions to the rest of the class. During subsequent discussion, encourage students to share their thoughts and opinions with one another and to attempt to reach consensus on the just-mentioned aspects of the assigned study. Suggested studies include the following: Clarke, P. (2008). Validation of two postpartum depression screening scales with First Nations and Métis women. Canadian Journal of Nursing Research, 40(1), 112–125. (Correlations; Logistic Regression) Doran, D., Pickard, J., Harris, J., Coyte, P. C., MacRae, A. R., Laschinger, H., et al. (2007). The relationship between managed competition in home care nursing services and nurse outcomes. Canadian Journal of Nursing Research, 39(3), 150–165. (t tests; Hierarchical Linear Modelling) Ratner, P. A., Johnson, J. L., Richardson, C. G., et al. (2004). Effi cacy of a smoking-cessation intervention for elective-surgical patients. Research in Nursing & Health, 27, 148–161. (t tests; chi-square) Instructor Manual for Nursing Research in Canada: Methods, Critical Appraisal, and Utilization Geri LoBiondo-Wood, Judith Haber, Cherylyn Cameron, Mina Singh 9781926648545, 9781771720984, 9780323447652, 9780323057431
Download
Close
Close
