|Home > RNJ > 2007 > November/December > Research Insights|
This is the final part of a four-part series focusing on sections of experimental or quasi-experimental quantative research studies. Part IV will focus on the results and discussion sections of an article. This series has two purposes: (1) to present sections of a research study and (2) to offer guidelines to critique and evaluate a study for use in clinical practice.
The results section is organized by the study questions or hypotheses. The researcher reports the results of the data analysis by stating the name or symbol for the statistical test, the values obtained by data analysis, and the level of significance. The typical level of significance for nursing research is equal or below p of .05. This means the researcher is willing to accept that 5% of the results of data analysis are by chance alone, not the intervention. If the researcher is willing to accept that 1% of the data analysis results are by chance alone, the level of significance is set at or below p of .01. If the level of significance by data analysis is found to be higher that a p value of .05 or .01, there is no significance difference noted between or among intervention outcomes, plus no significant difference between outcome data and baseline data (Burns & Grove, 2001; Polit & Beck, 2004). The usual way of expressing results is as follows: “The scanned volumes underestimated catheterization volumes by an average of 80.6 111.3 ml (t = 3.9, p = .001). The two measures were positively correlated with a Pearson product-moment correlation coefficient of 0.87 (p < .001)” (Borrie et al., 2001, p. 189). The t indicates that the researchers used the t test for data analysis, the researchers used this test to indicate if differences occurred between scanned urine volumes and catheterized urine volumes. The p is the level of significance found with data analysis. In the case of p = .001, the researchers have results that are significantly different. In the second data analysis, the correlation was reported to be .87 using the Pearson product-moment correlation coefficient that is a statistic that analyzes relationships of variables. The correlation value can vary from –1.00 to + 1.00. The reported .87 is a strong, positive correlation. The p of < .001 indicated a significant relationship between scanned urine volumes and catheterized volumes of urine.
Results of data analysis are often presented in the text and on tables. Ideally, the tables are well organized with the table number, title of the data set, names of the values being described, the statistic used, and the results of the analysis. Table 1 is an example of a well-presented table.
Table 1 is clearly titled Patient Characteristics and Conditions; N means the total number of subjects, n means the number of subjects noted with this characteristic or condition, and % is the percentage of the total number of subjects for each characteristic or diagnosis. Descriptive data can also be displayed using pie charts, histograms, or bar graphs. In summary, the results section should be well organized, clearly presented, and concisely presented to allow the reader to quickly and accurately see the results of data analysis.
The first portion of the discussion section briefly restates the results of data analysis. The data analysis is then compared to prior like studies that were presented in the literature review. The results of the current study may or may not be similar with the findings of prior studies. The next part of the discussion shows how the results of the current study support or does not support the conceptual framework or theoretical basis for the study. This section also includes the strengths and limitations of the current study (Burns & Grove, 2001).
The final portion of the discussion section presents ways that the findings of the current study can be used in practice. Borrie and colleagues (2001), for example, concluded that the bladder scans under predicted the actual volume of urine found as compared with a catheterization procedure. The shape of the bladder can contribute to the variability in accuracy of the scan. They recommended that repeated scans of the bladder be done until there is little variation in the results. When that point is reached, the scan can be considered reliable, and catheterization procedures can be omitted if urinary retention is not noted by the scan. The last portion of the discussion section also presents ideas or study questions for further research.
In summary, the discussion portion of the research article briefly presents the findings of the data analysis and compares the results of the current study with prior studies, plus details of how the current study supports or does not support the conceptual framework. The discussion portion also details how the results of the current study can be used in practice. Finally, the researcher expresses thoughts about further studies that would be based on the findings of this current study.
About the Author
Barbara Brillhart, RN PhD CRRN, is an assistant professor at Arizona State University, College of Nursing. Address correspondence to her at Box 872602, Tempe, AZ 85287-2602 or firstname.lastname@example.org.
Borrie, M. J., Campbell, K, Arcese, Z. A., Bray, J., Hart, P., Labate, T., et al.(2001). Urinary retention in patients in a geriatric rehabilitation unit: Prevalence, risk factors, and validity of bladder scan evaluations. Rehabilitation Nursing, 26(5), 187–191.
Burns, N., & Grove, S. K. (2001). The practice of nursing research (4th ed.). Philadelphia: Saunders.
Polit, D. F., & Beck, C. T. (2004). Nursing research: Principles and methods (7th ed.). Philadelphia: Lippincott Williams & Wilkins.