Introduction
How might you predict future events in your practice? Why is it important, as a future DNP-prepared nurse, to consider such future events?
For a DNP-prepared nurse, future predictions might lead to better patient outcomes and care. Therefore, analyzing factors to predict or evaluate can assist in transforming nursing practice or healthcare delivery. The “statistical procedure most commonly used for prediction is regression analysis” (Gray & Grove, 2020).
Reference: Gray, J. R., & Grove, S. K. (2020).
Burns and Grove’s the practice of nursing research: Appraisal, synthesis, and generation of evidence (9th ed.). Elsevier
What’s Happening This Week?
In Week 8, Regression, you will examine the application of linear regression. You will analyze the strengths and weaknesses of the findings in a research study determined with linear regression, as well as explore alternatives to strengthen the aims of the study. You will also begin work on the Quality Improvement Project. While this Assignment is not due until Week 10, you are encouraged to begin this Assignment this week.
Learning Objectives
Students will:
· Explain how the prediction concept and regression could be a useful analytic tool in nursing
· Differentiate between logistic and multiple regression
· Analyze multiple and logistic output and a published study
· Salkind, N., & Frey, B. (2019). Statistics for people who (think they) hate statistics (7th ed.). SAGE Publications.
· Chapter 17, “Using Linear Regression: Predicting the Future” (pp. 313–319, 329–333)
·
Document:
Critical Assessment, Appendix E: Research Evidence Appraisal Tool (Word document)
Download Critical Assessment, Appendix E: Research Evidence Appraisal Tool (Word document)
·
Document:
Critical Assessment, Appendix G: Individual Evidence Summary Tool (Word document)
Download Critical Assessment, Appendix G: Individual Evidence Summary Tool (Word document)
Required Resources for Topic: Regression
·
Alanazi, F. K., Lapkin, S., Molloy, L., & Sim, J. (2023).
The impact of safety culture, quality of care, missed care and nurse staffing on patient falls: A multisource association studyLinks to an external site.
.
Journal of Clinical Nursing, 32(19/20), 7260–7272.
·
Moradkhani, B., Mollazadeh, S., Niloofar, P., Bashiri, A., & Oghazian, M. B. (2021).
Association between medication adherence and health-related quality of life in patients with chronic obstructive pulmonary diseaseLinks to an external site.
.
Journal of Pharmaceutical Health Care and Sciences, 7(1).
Required Resources for Topic: Quality Improvement Project Assignment Options
·
Frazier, M. C. (2023).
Examining the lived experience of secondary traumatic stress in emergency room nurses: A mixed methods studyLinks to an external site.
(Publication No. 30816304) [Doctoral dissertation, Saint Louis University]. ProQuest Dissertations and Theses Global.
·
Karamichos, S. (2023).
The professional identity of the nurse practitioner: A mixed methods studyLinks to an external site.
(Publication No. 30634134) [Doctoral dissertation, Oklahoma City University]. ProQuest Dissertations and Theses Global.
·
Medina, M. (2024).
Implementing motivational interviewing to improve medication adherence: A staff education projectLinks to an external site.
(Publication No. Order No. 30992260) [Doctoral dissertation, Walden University]. ProQuest Dissertations and Theses Global.
·
Rampersad, M. (2023).
Oncology nurses, compassion fatigue and general health: A mixed-methods studyLinks to an external site.
. Walden University.
·
Document: Simons, A. (2023).
Executive summary: Quality improvement initiative reducing orthopedic surgical site infections through nasal decolonization
Download Executive summary: Quality improvement initiative reducing orthopedic surgical site infections through nasal decolonization
. Walden University.
RESPOND TO THIS DISCUSSION POST
Tania
A correlation matrix and multiple regression analysis are both statistical tools used to examine relationships between variables, but they serve different purposes and can complement each other in research.
A correlation matrix displays the strength and direction of linear relationships between pairs of variables using correlation coefficients (typically Pearson’s r). It provides a quick overview of which variables may be significantly related and therefore good candidates for inclusion in a regression model. For instance, if investigating patient falls, a correlation matrix might show significant relationships between variables such as nurse staffing levels, safety culture, and missed care—data points found to be influential in the study by Alanazi et al. (2023). A strong correlation between these predictors and fall rates would justify further exploration using multiple regression to understand their unique and combined effects on the outcome.
Multiple regression, on the other hand, quantifies how several independent variables predict or explain variance in a single continuous dependent variable. It provides beta coefficients, significance levels, and measures like R² to determine the relative contribution of each predictor. The correlation matrix can be used as a preliminary tool to detect multicollinearity—situations in which predictors are highly correlated with each other, potentially distorting the regression model’s reliability (Laureate Education, 2021).
Together, these tools offer a powerful approach: the correlation matrix for screening and diagnostic assessment, and multiple regression for hypothesis testing and predictive modeling. For example, Moradkhani et al. (2021) used chi-square tests and regression to analyze the impact of medication adherence on health-related quality of life. If they had continuous predictors, a correlation matrix might have helped select appropriate variables for a multiple regression model.
Regarding the second part of the question, the difference between logistic and multiple regression lies primarily in the type of outcome variable they analyze:
·
Multiple regression (also called linear regression) is used when the dependent variable is continuous (e.g., number of falls, quality of life scores). It assumes a linear relationship between independent and dependent variables (Creswell & Creswell, 2018).
·
Logistic regression is used when the dependent variable is binary or categorical (e.g., whether a patient was readmitted: yes/no). It models the probability of an event occurring, using odds ratios to interpret the influence of predictors (Polit & Beck, 2021).
Both regressions can accommodate multiple predictors, but the interpretation and application differ due to the outcome’s nature. Logistic regression is particularly useful in clinical settings where binary outcomes, such as medication adherence or the presence of complications, are common (Moradkhani et al., 2021).
In summary, a correlation matrix is a foundational step in identifying potential predictors for multiple regression. While multiple regression explains variance in continuous outcomes, logistic regression predicts categorical outcomes. Used together, these methods enable comprehensive analysis and evidence-based decision-making in nursing research and practice.
References
Alanazi, F. K., Lapkin, S., Molloy, L., & Sim, J. (2023). The impact of safety culture, quality of care, missed care and nurse staffing on patient falls: A multisource association study.
Journal of Clinical Nursing, 32(19/20), 7260–7272.
Links to an external site.
Creswell, J. W., & Creswell, J. D. (2018).
Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). SAGE Publications.
Laureate Education. (2021).
Statistics and data analysis in nursing research [Video file]. Baltimore, MD: Walden University.
Moradkhani, B., Mollazadeh, S., Niloofar, P., Bashiri, A., & Oghazian, M. B. (2021). Association between medication adherence and health-related quality of life in patients with chronic obstructive pulmonary disease.
Journal of Pharmaceutical Health Care and Sciences, 7(1).
Links to an external site.
Polit, D. F., & Beck, C. T. (2021).
Nursing research: Generating and assessing evidence for nursing practice (11th ed.). Wolters Kluwer.
·
Reply to post from Tania AlfonsoReply
·
Mark as UnreadMark as Unread
