Research_Critique

RESEARCH CRITIQUE OVERALL SUMMARY This article will critique the study by Cook, Green and Topp —‘exploring the impact of physician verbal abuse on perioperative nurses’. Overall speaking, this is a well conducted and interesting study. The strength of this study is the use of Roy Adaptation Model as the framework. The data collection as well as data analysis closely followed the guidelines given by the model. The thoughts in the discussion session provide the readers a clear understanding of the study. Limitation lays at the recruitment step. Only about half of the sent out questionnaires were received with responses and in total 39% of the sample were used for data analysis. This might result in biased sampling. Title A title is a study’s window to the world. Therefore it should be interesting and complete in order to convey the important elements of a study (Holmi & Liewellyn, 1986). The title of this report is brief (10 words), concise and clear. From the title, the reader can get all the key information, focus of the study (impact), phenomenon of interest (physician verbal abuse) and the target population (perioperative nurses). Abstract An abstract is helpful in determining whether the article is of relevant for a selected topic. Crosby (1990) stated that an abstract is a clear and concise summary of a study. It ranges from 100 to 250 words and usually includes the study purpose, design, setting, sample size, results and conclusion (Burns & Grove, 1995). The abstract of this report is very brief, but it is informative. It well summarizes the study by providing the objective, basis of the study design, description of the study sample, key result, and conclusion. Research Questions This is a nonexperimental descriptive study which means a clear and concise problem statement is the key. Obviously, the author has a good skill in doing so. With a brief background clearly stating the need and importance of such a research, the objective of this study is emphasized again. This enables the readers to easily follow the list of research questions that have been identified by the author, and therefore, grabs the interests of the readers to find out how the study could help to answer the questions. Literature Review The author first gave the definitions of the key research variables by citing existing researches as well as the descriptions of how these variables were measured. This may help the readers to understand the framework (Roy Adaptation Model) of this study. Therefore, it could serve the purpose better if this section was organized together with the theoretical framework section. The definitions of the variables are clear and accurate. In literature review, the selected studies should be relevant and can help to gain broad background knowledge or understanding of the information that is available related to the research problem of interest (Burns & Grove, 1995). In the literature review, 10 relevant studies were reviewed. First the author reviewed the negative impacts of verbal abuse to the health care environment. Then she talked about the causes of physicians’ verbal abuse towards nurses and the coping behaviours of nurses by reviewing 4 studies. And lastly she pointed out the lack of study in the incidence and impact of physician verbal abuse among perioperative nurses as well as some findings from existing researches about the impacts. The manner of review is very comprehensive. However, most of the studies reviewed are classical reviews. The lack of current review made the reader doubted whether the findings are up-to-date. Theoretical Framework The choice of Roy’s Adaptation Model (RAM) as the conceptual model for the study is excellent. The ability in explaining a person as an adaptive system responding to internal and external stimuli best fit the need of this study. The author gave explanations to each variable in the model with relevant example. A schematic diagram visualizing the framework was also provided. This helps readers who are not familiar with RAM to understand this model better therefore understand the study better. Research Design This study used a nonexperimental descriptive design. Since the variables of interest in this study are independent, not subjected to manipulation, a nonexperimental design can be used (Polit & Beck, 2004). The objective of this study is to explore incidence and impact of verbal abuse, without causal or correlation, thus, a descriptive design is sufficient. A nonexperimental descriptive design requires the researcher to be skillful at sampling technique, questionnaire design and data analysis. These aspects will be discussed separately in the respective sessions. Research designs must balance the needs for internal validity and external validity to produce useful results. Internal validity requires the researchers to rule out other factors or threats as rival explanations of the relationship between the variables (Beanland et al, 2000). Variables were well defined in this study. The Verbal Abuse Scale had been verified for validity, clarity and completeness. And the test-retest reliability was checked on a sample of 21 nurses with 1 week interval. This sample size is sufficient and the interval is reasonable. The author reported that the Cronbach alpha ranged from .67 to .95 for different subscales. More clarifications are necessary here as a test with .95 Cronbach alpha is considered a good test while a Cronbach alpha will be just “barely acceptable”. The author needs to identify which are the subscales with low Cronbach alpha and makes less influence on them. Similarly for test-retest reliability, the author reported the correlation between .45 and .79. The reliability of the data is questionable. External validity deals with problems of generalizability of the findings to additional populations and to other environmental conditions (Beanland et al, 2000). This study lacks generalization because of the small sample and the use of a convenience sampling method. Sampling Procedure A convenience sample of 200 was used for the study. The author clearly described the characteristics of this sample as well as the inclusion criteria for recruitments. However, there is higher risk of bias compared to other sampling techniques because the sample is self-selecting and representativeness is questionable (Beanland, 2000). According to the sampling strategy, there is a strong bias and the results might only be representative for the population of AORN of Toledo. This bias has been discussed by the author in the limitation session. As an alternative, Stratified random sampling could be used to produce more precision. Sample can be randomly chosen in categories of hospital including all OR nurses. Data Collection Strategies The accuracy of data source was important and explicit because it might lead to exact replication of a study (Nieswiadomy, 1998). In this study the researchers used mailing as the way for data collection. Although they had made a preliminary announcement at AORN chapter meeting one week before the mailing to facilitate more response, more effort should be put in before sending out the questionnaires. For example, they can use attractive gift or coupon for returning questionnaire or advertisement on AORN magazines. They had made a good point to include a cover letter containing instruction and a stamped, self addressed envelope in each questionnaire which could be costly. However use of mailing still leads to a very low response rate (about 39%) that cause nonrespondent bias which hardly represents the whole population. Nonrespondent bias refers to the differences between participants and those who declined to participate (Polit,Beck & Hungler, 2001).Seeking the help from hospitals by distributing the questionnaires to departments and making it compulsory may give a higher response rate and be less costly. Ethical Consideration Approval for conducting this study was obtained from the Medical College of Ohio’s institutional review board (IRB). When humans are used as study participants, care must be exercised in ensuring the rights of those humans are protected (Polit & Beck, 2004). Therefore confidentiality is very important in research study. The IRB’s approval of consent by returning completed questionnaires instead of signed informed consent allows the subjects to give honest answers. After the researches had collected questionnaires, they stored it in a locked file drawer to protect confidentiality. Data Analysis The methods the author used for data analysis are appropriate. Concepts from RAM were used for qualitative categorization of write-in responses. This made the write-in responses analyzable using the framework. SPSS was used for quantitative analysis. Mean and standard deviation were used to describe the data. The descriptive statistics were reported in table form. The reader can easily capture the information. The author described the results from each subscale in detail using the RAM framework. The reader found it rationale and clearly. However, the reader also found confusions at the beginning of the results session. Firstly, within the returned and completed responses, there were other exclusion criteria. These criteria were not described in any part of the article. Secondly, on the second paragraph of results session, according to the percentages of second mailing reported, about 75% of the returned and completed responses were excluded. This is a significant percentage that needs explanation. Third, the gender composition of the perioperative nurse population is necessary to prove that the sampling is not biased towards female group. Interpretation, Discussion and Clinical Application In the discussion, the researcher aims to examine, interpret, and qualify ther results, as well as draw inferences from them (Stockhausen & Conrick, 2002). The author used the interpretations of results to answer the research questions. And for each question answered, she compared the results to the findings of other existing researches. All the reasoning is sound and agreeable except the conclusion on the similarity of the first mailing group and second mailing group. The author was careful enough to carry out the independent-sample t test on the two sample means for both situations (equal variances assumed and not assumed). Although only the frequency subscale showed significant difference between the two groups, both the stress and similarity of coping subscales showed differences that are very close to significant. Second mailing group included subjects who did not complete the questionnaire and subjects who did not return the questionnaire in the first mailing. There could be many reasons that those subjects failed to complete the questionnaire and return it. One possibility is that subjects from this group are less responsible or more negative. This possibility could lead to the conclusion that they are not coping well in their jobs, therefore, they receive more verbal abuses and are less effective in adaptation. To ensure the representative of the study sample to the population of perioperative nurses from AORN of Toledo, more subjects from second mailing group need to be recruited for balance. The author addressed a few limitations of the study. The discussions of the limitations are comprehensive. To overcome the limitations as well as to improve on further study, the author provided some good recommendations for future researches. The conclusions and implications are brief but thoughtful. The author concluded that the effectiveness of adaptation in the study sample is due to the facts that a large percentage of the subjects has been through formal training in communication. One implication from this conclusion is that education of communication and conflict resolution skill plus anger handling will be helpful for nurses in address verbal abuse. However, more importantly, policies and corrective actions need to be implemented for preventing verbal abuse. The conclusions and implications are important as it contributes to nursing knowledge and provides a direction for improvement. Presentation and Stylistic This article was well written with sufficient amount of details. The author has a good skill of choosing language that is brief but informative. Overall speaking, this article is well organized and very clear for readers to follow the thoughts of the author. References All the citations are accurate and reflected at the end references. The author has cited some current studies. However, a majority of the references are about 8 to 10 years old before this article. REFERENCES Beanland, C., Schneider, Z., LoBiondo-Wood, G., & Haber, J. (2000). Nursing Research: methods, critical appraisal and utilisation. Sydney: Mosby. Burns, N., & Grove, S.K. (1995). Understanding Nursing Research. USA: W.B. Saunders Company. Holmi, K., & Liewellyn, J.G. (1986). Nursing research for nursing practice. Philadelphia: W.B. Saunders Company. Nieswiadomy, R. M. (1998). Foundations of nursing research (3rd ed.). Stamford Connecticut: Appleton & Lange. Polit, D.F. & Beck, C.T. (2004). Nursing research: Principles & methods (7th ed.). Philadelphia: Lippincott Willaims & Wilkins. Polit, D. F., Beck, C.T. & Hungler, B.P. (2001). Essential of nursing research: methods, appraisal and utilization (5th ed.). Lippincott: USA Stockhausn, L. & Conrick, M. (2002). Making sense of research: A guide for critiquing a paper. Contemporary Nursing, 14(1), 38-45.