Nursing - Research Methods for Professional Practice - Essay Example

Nursing (Research Methods for Professional Practice) Nursing (Research Methods for Professional Practice) Introduction The two articles provided for this paper were mainly studying the extent to which research results are used by nurses in practice as relates to self-compassion. Research is a critical and informative tool in the nursing practice, especially with the professional and policy developments that have been witnessed in the past two decades (Frasure 2009, p. 10). Nurses are facing growing pressure to be more accountable for the decisions they make and actions they take, which calls for evidence-based practice that can best be provided by research (Thompson & Estabrooks 2009, p. 15). For nurses and midwives, an understanding of research methods is important because all decisions arrived upon regarding evaluations, diagnosis, treatment and care are informed by research (Kumar 2009, p. 173). As shown by its usage in the two articles provided for the purpose of this paper, sampling is a key element of research methods and is chosen because it presents advantages and challenges equally. The paper shall compare and contrast the use of sampling by discussing three issues associated with it as an element, which are defining the population, generalising data and the practicality of random samples in a study. Defining the population In order for a target population to be defined, the research element must be specific on the attributes that identify its members (Belkhodja 2010, p.379). Such attributes include but are not limited to education, income, age, marital status and area of work or residence. The first article by Heffernan et al (2010) had a sample of 350 registered nurses (RNs) and nurse managers from three counties in New York, out of which 143 RNs participated via an online link provided by the health system. The second article by Horsburgh and Ross (2013) had a sample of 47 newly qualified staff nurses who had practiced for just one year. One similarity in how sampling is used in the two articles is that it is random. Here, the significance of random sampling in both researches was to get participants that are as representative of the nursing population as is practically possible. Random techniques have the potential to limit biases (Wallin & Ewald 2008, p. 512). This is also in view of the fact that sampling makes it possible to study a portion of the target population affordably and further, studying an entire population increases the probability of errors. However, sampling is also used differently in the two researches, where it is particularly limited in Horsburgh and Ross (2013). Their study only defined the population as newly qualified staff nurses who had practiced for approximately a year. This, unlike Heffernan et al (2010), overlooks many other variables such as age and, more importantly, income and experience. The main disadvantage for this move is that the results obtained from a purely nurse population may not be as valid as those from nurse managers, who probably have many years of experience and earn enough to afford comfortable lives. For the purpose of the research subject the two articles were addressing (self-compassion), experience and income, and hence comfort of life, are critical variables in determining how an individual is compassionate to the self. This is because nurses must first be compassionate to themselves before they can disseminate it to patients. For instance, all participants in the article by Horsburgh and Ross (2013) claimed that they felt they had not received adequate guidance in the working environment for the year they had practiced. If analysed critically, that sample had worked for a relatively short time and may not even be aware of the questions they need to ask in a clinical environment unlike academic setting. This explains how the validity of findings may be compromised by a poorly defined population in random sampling, especially one that only emphasises on a single attribute of the larger population (Estabrooks & Cummings 2010, p. 290). Biases in this study rest dominantly in the fact that the participants of the second research were limited to a particular job group. An alternative would be to put a ceiling on the eligibility basing on experience and earnings, or test different job groups separately, albeit the cost. Generalising Data It is common practice to consider whether data can be generalised meaningfully from a sample considered to be too small even if it was recruited randomly (Parikh, Parikh & Thomas 2009, p. 217). Previous studies have declared random samples with not more than 50 participants to be small, and Tuckett (2009, p. 49) confirmed this by showing how the margins of error from such a small sample can be enormous. The essence of a random sample is that the results obtained from it can be generalised to represent an appropriately defined population (DeVellis 2008, p. 74). However, smaller samples have proven to be rarely worth the effort, trouble and costs, whether or not they can theoretically be generalised (Halperin, W 2010, p. 89). When researchers attempt to transfer findings from such studies to, say, a similar population in a different geographical region, even though within the same population, margin errors increase significantly (Jackson & Waters 2008, p. 369). This can be explained by the lack of considering specific attributes in the different areas, for example, demographics (Niederhauser & Kohr 2010, p. 82). To explain the tendency of higher error margins in smaller samples, Tuckett (2009, p. 60) uses an example from general voting. Suppose candidate A is leading B by 15% in a poll represented by a sample of 500 with an error margin of 5%. Then, to use the error margin, it is simply added and subtracted from that sample’s results and, in this example, 15% plus or less five implies A’s lead was between 20% and 10%. However, because the results were obtained from a sample and not an entire population, there are high chances of inaccuracies, specifically from smaller samples as tabulated below and also agreed by Ploeg (2009, p. 211). Size of Sample Error Margin (Percentage) Range for Estimate (Percentage) 10 31 19-81 20 22 28-72 40 15 35-65 50 13 37-63 Table 1: Sample size vs margin of error (Tuckett 2009, p. 60). Heffernan et al (2010) used a sample of 350 participants. On the other hand, Horsburgh and Ross (2013) only had a sample of 47 participants, a fact that is further compounded by their relative lack of experience in the professional setting. The difference here is that the first sample can be termed more “solid” in the sense that it included an array of job groups in the hierarchy of the nursing profession while the second one had no variety. Hence, the results obtained from the first article can be generalised to represent a significant portion of the population of nurses. In contrast, the second one was too small for any practical results representative of the nursing fraternity. For the sake of clarity, results of a study done on a certain category of nurses cannot be generalised to the entire nurse population. Practicality of Random Samples The third issue evident in the use of random sampling in the two studies is that random samples can actually turn out to be impractical for most researches, whether they are quantitative or qualitative analyses. For practicality purposes, researchers must itemize each member of the potential population they intend to draw the sample from then contact them to implore their participation (Cozby 2009, p. 114). However, the study that only had 47 participants yet intended to collect views believed to represent majority of nurses is characterised by a greater than practical compromise, which locks out validity of findings as earlier shown by Tuckett (2009, p. 60). In fact, the key purpose the sample was chosen seems to have been to address a very particular population of nurses: newly registered staff nurses. For objective purposes, since compassion is an inextricable characteristic of nursing, the participants in any nursing research should be more diverse in order to provide evidence based results (Bernstein & Kramer 2009, p. 132). That aspect presents another difference in the way in which random sampling is used in the two studies. From the perspective of any researcher, validity indicates how sound and accurate their research is and applies particularly to both methods and design of the research (Squires 2011, p. 21). Similarly, reliability should be an indication of the consistency of the findings (Creswell 2010, p. 69). Although the study established there is need for undergraduate programs to prepare students for the professional environment, its disadvantage is presented by the lack of subject variability, compromising the validity of the findings. The only advantage realised from such a small sample size is that it is less costly to conduct a research on (Streiner & Norman 2009, p. 201). The results were published, but that does not guarantee validity as does the entirety of the picture covered and how credible the sample is (Nayak 2011, p. 410). On the contrary, the first study had a considerable sample size that can meet reliability tests because it offers results that are inherently repeatable. Repeatability is one of the strongest measures of reliability, because it highlights consistency (Haines, Kuruvilla & Borchert 2008, p. 725). It was successful because the researchers also had their focus on the larger picture in their methodology. The two studies can confidently be said to have different levels of validity, reliability, consistency and trustworthy. Using 350 RNs and nurse managers from three counties, Heffernan et al (2010) can be considered to be more reliable and valid. They were more random, and the broader selection of participants is a strong indicator of consistency and trustworthiness. Although not every survey result can be believed in, the lack of biasness in the first study shows that it was reliably conducted, unlike the second one that relied only on a specific nurse population. The second study was not exactly designed to comprehensively address the research question, although it did adhere to research protocols in terms of randomly selecting participants. Although the title of study in the second research informs the reader what the research is about, it is also its key misgiving because of the participants it directly targets. There is a need for the study, but Horsburgh and Ross (2013) greatly limit their scope of participants, unlike Heffernan et al (2010). Again, results from the two studies were described adequately, although the second one fails to adequately satisfy all research requirements mainly by the underlying biasness in terms of participants. This is because random selection of participants requires that no exceptions are imposed (Frasure 2009, p. 12). Viewed from another perspective, imposed exceptions do not contribute positively towards reliability and validity of evidence obtained from the results of a research, mainly because it locks out the most concerned sources of information and limits generalization (Tuckett 2009, p. 52). Conclusion It can be concluded that the research designs in the two studies matched the research question and sampling as an element of research method was appropriate. Further, both can be said to have been successful, albeit in their own unique categories. The aspect of coercion can be eliminated, but personal views played the key roles. However, the random samples may not have been genuinely random because of the limitations presented by the second study. The research question had not been answered appropriately previously, but these two studies indicate that there is a degree of usage of research findings in practice. In terms of quantitative studies, the target population was only significantly identified by the first study. The first study found that nurses actually have a significant degree of self-compassion and are able to care for patients with equal compassion. This finding has a great degree of validity and reliability. This conclusion is drawn from the credible sample size used and the diversity of the participants involved. The second study, although limited in the diversity of its sample group, was able to point towards a miscellaneous kind of support rather than systematic for newly qualified staff nurses. Nurses were unable to apply research results mainly because of the institutionalised negativity from long-serving colleagues. However, the validity and reliability of this result is greatly compromised by the sample size. References Belkhodja, O 2010, ‘The extent and organizational determinants of research utilization in Canadian health services organizations’, Science Communication, vol. 28, no. 2, pp. 377-417. Bernstein, D & Kramer, G 2009, Introduction to clinical psychology, Pearson, New Jersey. Brains, W & Manheim, R 2011, Empirical political analysis, Longman, Massachusetts. Cozby, P 2009, Methods in behavioral research, McGraw, Massachusetts. Creswell, J 2010, Research design: qualitative, quantitative and mixed method approaches, Sage, California. DeVellis, R 2008, Scale development: Theory and applications, Sage, London. Estabrooks, C & Cummings, G 2010, ‘A comparison of research utilization among nurses working in Canadian civilian and United States Army healthcare settings’, Research in Nursing and Health, vol. 30, no. 3. pp. 282-296. Frasure, J 2009, ‘Analysis of instruments measuring nurses attitudes towards research utilization: a systematic review’, Journal of Advanced Nursing, vol. 61, no. 1, 5-18. Haines, A, Kuruvilla, S & Borchert, M 2008, ‘Bridging the implementation gap between knowledge and action for health’, Bulletin of the World Health Organization, vol. 82, pp. 724-732. Halperin, W 2010, Field investigations of occupational disease and injury, Oxford University Press, New York. Heffernan, M, Quinn, MT, McNulty, SR, & Fitzpatrick, JJ 2010, ‘Self compassion and emotional intelligence in nurses’, International Journal of Nursing Practice, vol. 16, pp. 366-373. Horsburgh, D & Ross, J 2013, ‘Care and compassion: the experiences of newly qualified staff nurses, Journal of Clinical Nursing’, vol. 22, pp. 1124-1132. Jackson, N & Waters, E 2008, ‘Criteria for the systematic review of health promotion and public health, vol interventions’, Health Promotion International. 20, no. 4, pp. 367-374. Kumar, B 2009, ‘Why learn research methodology’, Indian Journal of Ophthalmology, vol. 57, no. 3, pp. 173-174. Mack, N & Macqueen, K 2010, Qualitative research methods: a data collector’s field guide, Family Health International, North Carolina. Nayak, B 2011, ‘Conflict of interest in biomedical publications’, Indian Journal of Ophthalmology’, vol. 65, no. 1, pp. 409-410. Niederhauser, V & Kohr, L 2010, ‘Research endeavors among pediatric nurse practitioners study’, Journal of Pediatric Health Care, vol. 19, no. 2, pp. 80-89. Oppong, H 2013, ‘The problem of sampling in qualitative research’, Asian Journal of Management Sciences and Education, vol. 2, no. 2, pp. 202-210. Parikh, R, Parikh, S & Thomas, R 2009, ‘Likelihood ratios: clinical application in day-to-day practice’, Indian Journal of Ophthalmology, vol. 57, no. 4, pp. 217-221. Ploeg, J 2009, ‘Factors influencing best-practice guideline implementation: lessons learned from administrators, nursing staff and project leaders’, Worldviews on Evidence-Based Nursing, vol. 4, no. 4, pp. 210-219. Squires, J 2011, ‘To what extent do nurses use research in clinical practice? A systematic review’, Implementation Science, vol. 6, no. 1, pp. 21-22. Streiner, D & Norman, G 2009, Health measurement scales: a practical guide to their development and use, Oxford University Press, Oxford. Thompson, D & Estabrooks, C 2009, ‘Interventions aimed at increasing research use in nursing: a systematic review’, Implementation Science, vol. 2, no. 2, pp. 15. Tuckett, A 2009, ‘Qualitative research sampling: the very real complexities’, Nurse Researcher, vol. 12, no. 1, pp. 47-61. Wallin, L & Ewald, U 2008, ‘Sustainability in changing clinical practice promotes evidence based nursing care’, Journal of Advanced Nursing, vol. 41, no. 5, pp. 509-518. Read More