Assignment Prompt & Instructions
– 2 years ago
**files (5)**
– ## WK4DISCDATA.docx
– ## NURS_8211_WK4_ComparisonofMeansPart1.pptx
– TheImpactofDiabetes.pdf
– PalliativeCommunication.pdf
– CriticalCareNurse.pdf
**## WK4DISCDATA.docx**
Self-Study: Comparison of Means, Part 1
Throughout the course, there will be a self-study Discussion pertaining to an important concept or topic covered within the assigned week. These Discussions are designed to give you the opportunity to collaborate with your peers and faculty, test your knowledge, ask questions, practice research analysis, and assist your peers.
You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
Resources
Be sure to review the Learning Resources before completing this activity.
Click the weekly resources link to access the resources.
[**WEEKLY RESOURCES**](https://waldenu.instructure.com/courses/157739/modules/items/6373691)
[Required Resources](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Dang, D., Dearholt, S. L., Bissett, K., Ascenzi, J., & Whalen, M. (2021).
_Johns Hopkins evidence-based practice for nurses and healthcare professionals: Model & guidelines_ (4th ed.). Sigma Theta Tau International Honor Society of Nursing.
· Chapter 6, “Evidence of Appraisal: Research” (pp. 155–158)
· Appendix E, “Research Evidence Appraisal Tool” (pp. 298–302)
· Appendix F, “Nonresearch Evidence Appraisal Tool” (pp. 307–314)
· Appendix G, “Individual Evidence Table” (pp. 315–318)
· Salkind, N., & Frey, B. (2019).
_Statistics for people who (think they) hate statistics_ (7th ed.). SAGE Publications.
· Chapter 12, “t(ea) for Two: Tests Between the Means of Different Groups” (pp. 236–239, 243)
· Chapter 13, “t(ea) for Two [Again]: Tests Between the Means of Related Groups” (pp. 253–256, 259)
[Required Media](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Niedz, B. (2024).
_Comparison of means, part 1_ [Video]. Walden University Canvas. http://waldenu.instructure.com
[PowerPoint Presentation](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
[Required Resources for Topic: \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Cook, H. E., Garris, L. A., Gulum, A. H., & Steber, C. J. (2024).
[Impact of SMART goals on diabetes management in a pharmacist-led telehealth clinicLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/08971900221125021).
_Journal of Pharmacy Practice, 37_(1), 54–59.
[Required Resources for Topic: Paired \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· DeFusco, C., Lewis, A., & Cohn, T. (2023).
[Improving critical care nurses perceived self-efficacy in providing palliative care: A quasi-experimental studyLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/10499091221094313).
_American Journal of Hospice and Palliative Medicine, 40_(2), 117–121.
· Markiewicz, A., Hickman, R. L., McAndrew, N. S., & Reimer, A. (2023).
[Enhancing palliative communication in the intensive care unit through simulation: A quality improvement projectLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://www.sciencedirect.com/science/article/pii/S1876139923000075).
_Clinical Simulation in Nursing, 77_, 1–5.
**To prepare:**
· Read and view the Learning Resources in Doherty & Skalsky and in Dang et al. (2021) in Required Readings.
· View the video on comparison of means.
**Use** this Discussion to collaborate with your peers and faculty as an open office hours/ Q&A forum.
**Post** answers to the following:
· Interpret independent sample
_t_-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples
_t_ test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired
_t_ test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
·
For this Self-Study Discussion, you may post throughout Week 4. You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
**Post** answers to any or all of the following:
· Interpret independent sample t-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples t-test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired t-test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
Our interactive discussion addresses the following learning objectives:
· Interpret comparison of mean (
_t_-test) data in a published study
· Differentiate between dependent observations and independent observations
[Previous](https://waldenu.instructure.com/courses/157739/modules/items/6373696)
[Next](https://waldenu.instructure.com/courses/157739/modules/items/6373710)
**## NURS_8211_WK4_ComparisonofMeansPart1.pptx**
NURS 8211
Research for an evidence-based practice
Week 4:Comparison of Means Part 1
Week 4: Comparison of Means Part 1
Objectives
Interpret a comparison of means (t test) data in a published study.
Differentiate between dependent and independent observations.
2
Appraising Evidence
Tips on reading a research study
Qualitative vs. Quantitative
Finding the null hypothesis: what are the dependent and the independent variables implicit in the study?
How were they measured, in the same people or in different in the groups?
Dang et al. Critical Appraisal tool (Appendix E pp.298-302
Theory generating (qualitative) vs. theory testing research (quantitative)
Qualitative deals with words; quantitative deals with numbers and statistics
3
The impact of smart goals on diabetes Management in a pharmacist-led telehealth clinic
Dependent variable: A1c level
Independent variable: SMART Goals
Observations are from different people, although they are “matched”
Null Hypothesis: There will be no difference in A1c levels when the intervention group with SMART goals is compared to the control group (no SMART goals setting)
Read the abstract, you can immediately get a feel for the study right from the the first page.
Identify the variables: what is the dependent what is the independent.
Form a null and an alternate hypothesis.
Clearly, they expected to see a statistically significant outcome: that A1c levels in the intervention group would be lower over 3 months as compared to the control group over a similar 3-month period.
Discuss clinical vs. statistical significance.
4
Cook et al. (2024)
Take a look at the table in Cook et al. (2024) on p. 56 in the article. Table 1 includes descriptives statistics which are presented in each group (the intervention group and the control group).
This was at a VA outpatient clinic, and there were 50 in the intervention group and 50 in the control group. The researchers set specific inclusion and exclusion criteria. For example, they specified that patients should be approximately the same age + or – 5 years. So the mean age in the SMART goal group was 64 years with a (9.2) standard deviation. And the mean age in the control group was 65, just a year older on the average with a SD of 8.6. Although there was statistical testing on age, you can see that this met their preset criteria for differences between the groups on no more than + or – 5 years difference.
You can take at other variables that they collected, including Mean A1c, in each group 9.1 and 9.0, a bit higher than should be and only slightly different, but as there are descriptive statistics, you really can’t see statistical significance, yes or no from these descriptive statistics. You need an inferential test and a p value.
5
Cook et al. (2024)
Here is the change in A1c from baseline to the 3 month mark.
In the text of the article, on p. 55 they disclose that the comparison used a Student’s t test (another way to describe an independent samples t test.)
In this figure, on p. 56, you can see a nice bar graph showing that both groups had a reduction in A1c: the SMART goal group, decreased 1.2 points, and the control group also decreased by a bit less (to .85). The p value that resulted from the independent samples t test was p = .287 as you can see in the note at the bottom of the graph.
Because the p value was so much larger that .05, we cannot reject the null hypothesis. Statistically we cannot say with any statistical confidence that these two numbers are due to the intervention. The difference is more likely, just due to chance.
Although this is true, the reduction in A1c is indeed clinically significant and this is validated in the study on page 58. Take a look in the discussion.
Now what might explain this result and what might be some suggestions for the future.
The first thing that comes to mind is that this might be a type II error…….Preventing a type 1 error (the risk of finding significance when it is really there) is really protected by the p value, set at .05 to reduce this risk.
But a type II error occurs when you don’t find significance but it is really there. And a very small difference, like this one, is definitely a greater risk. There is a technique called power analysis that researchers often use a priori or in advance. To detect a small effect size (that what they had here) with a 95% power these researchers would have need 131 cases in each group (262 total). As they only had a total of 100 cases, you can see one pertinent reason as to why they did not find statistical significance.
The Mann Whitney U test is the nonparametric equivalent of the independent samples t test. When the data do not meet the requirements (the assumptions) of the parametric test, they often substitute a nonparametric test instead. IN general, parametric tests are stronger than nonparametric tests. In published studies, sometimes the authors will describe the reasons for a choice of a statistical test. Sometimes they do not.
6
Improving critical care nurses perceived self-efficacy in providing palliative care: a quasi-experimental study
Dependent variable: nurses’ self efficacy in palliative care
Independent variable: an educational process using online videos
Observations are from the same nurses pre to post
Null Hypothesis: There will be no difference in self efficacy as measured in nurses before and after an educational process
Let’s use the same method we used before to take a look at this study. Defusco et al. (2023)
The nurses completed a 12 item survey that measured nurses’ self efficacy on palliative care for both psychosocial support and symptom management before and after viewing 5 videos and 2 documents from the VitalTalk website.
A convenience sample of self-identified nurses was used.
Power analysis was conducted effect size of .45 (moderate difference…..remember, I used .20 for the Cook study).
Sample of 41 participants . Ended up with 40.
Because the observations (the results of the survey pre to post) are from the same people, they are said to be DEPENDENT. (don’t confuse this language with the dependent and independent variables). It just indicates that because the observations are from the same people, you can use a paired t test or its non-parametric equivalent, the Wilcoxon Signed Rank test.
The nonparametric choice was a good one in this study. That’s because the paired t test has some requirements (these are called assumptions) and one of them is that the data are normally distributed. With a small sample, that can be challenging. The authors explain this on p.119 in the data analysis plan.
Null hypothesis is: There will be no difference in self efficacy as measured in nurses before and after an educational process.
Let’s look at the results.
7
Defusco et al. (2023)
Descriptives: Medians were used because data were not normally distributed
Wilcoxon Signed Rank test used to measure nurses’ self efficacy on overall self efficacy and two subscales pre and post:
Overall: 38 to 43.5 z = -4.868 p < .001
Psychosocial subscale: 18.0 to 21.0 z = -4.867, p < .001
Symptom management: 21 to 22.5 z = -3.861, p <.001
8
Key Points
Using a deliberate approach to reading a research study has benefit.
The Dang et al. (2022) critical appraisal tool asks some pertinent questions about published studies that help you to summarize carefully and thoroughly.
Although Cook et al. (2024) did not achieve statistically significant results, the outcome was clinically significant and useful.
DeFusco et al. (2023) did achieve both statistical and clinical significance in their study of nurses’ self efficacy in palliative care.
9
## image1.jpg
## image2.jpg
## image6.jpg
## image5.jpg
## image8.png
## image9.png
## image3.jpg
## image4.jpg
**TheImpactofDiabetes.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/theimpactofdiabetes-pdf)
**PalliativeCommunication.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/palliativecommunication-pdf)
**CriticalCareNurse.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/criticalcarenurse-pdf)
**CriticalCareNurse.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/criticalcarenurse-pdf)
**## WK4DISCDATA.docx**
Self-Study: Comparison of Means, Part 1
Throughout the course, there will be a self-study Discussion pertaining to an important concept or topic covered within the assigned week. These Discussions are designed to give you the opportunity to collaborate with your peers and faculty, test your knowledge, ask questions, practice research analysis, and assist your peers.
You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
Resources
Be sure to review the Learning Resources before completing this activity.
Click the weekly resources link to access the resources.
[**WEEKLY RESOURCES**](https://waldenu.instructure.com/courses/157739/modules/items/6373691)
[Required Resources](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Dang, D., Dearholt, S. L., Bissett, K., Ascenzi, J., & Whalen, M. (2021).
_Johns Hopkins evidence-based practice for nurses and healthcare professionals: Model & guidelines_ (4th ed.). Sigma Theta Tau International Honor Society of Nursing.
· Chapter 6, “Evidence of Appraisal: Research” (pp. 155–158)
· Appendix E, “Research Evidence Appraisal Tool” (pp. 298–302)
· Appendix F, “Nonresearch Evidence Appraisal Tool” (pp. 307–314)
· Appendix G, “Individual Evidence Table” (pp. 315–318)
· Salkind, N., & Frey, B. (2019).
_Statistics for people who (think they) hate statistics_ (7th ed.). SAGE Publications.
· Chapter 12, “t(ea) for Two: Tests Between the Means of Different Groups” (pp. 236–239, 243)
· Chapter 13, “t(ea) for Two [Again]: Tests Between the Means of Related Groups” (pp. 253–256, 259)
[Required Media](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Niedz, B. (2024).
_Comparison of means, part 1_ [Video]. Walden University Canvas. http://waldenu.instructure.com
[PowerPoint Presentation](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
[Required Resources for Topic: \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Cook, H. E., Garris, L. A., Gulum, A. H., & Steber, C. J. (2024).
[Impact of SMART goals on diabetes management in a pharmacist-led telehealth clinicLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/08971900221125021).
_Journal of Pharmacy Practice, 37_(1), 54–59.
[Required Resources for Topic: Paired \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· DeFusco, C., Lewis, A., & Cohn, T. (2023).
[Improving critical care nurses perceived self-efficacy in providing palliative care: A quasi-experimental studyLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/10499091221094313).
_American Journal of Hospice and Palliative Medicine, 40_(2), 117–121.
· Markiewicz, A., Hickman, R. L., McAndrew, N. S., & Reimer, A. (2023).
[Enhancing palliative communication in the intensive care unit through simulation: A quality improvement projectLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://www.sciencedirect.com/science/article/pii/S1876139923000075).
_Clinical Simulation in Nursing, 77_, 1–5.
**To prepare:**
· Read and view the Learning Resources in Doherty & Skalsky and in Dang et al. (2021) in Required Readings.
· View the video on comparison of means.
**Use** this Discussion to collaborate with your peers and faculty as an open office hours/ Q&A forum.
**Post** answers to the following:
· Interpret independent sample
_t_-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples
_t_ test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired
_t_ test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
·
For this Self-Study Discussion, you may post throughout Week 4. You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
**Post** answers to any or all of the following:
· Interpret independent sample t-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples t-test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired t-test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
Our interactive discussion addresses the following learning objectives:
· Interpret comparison of mean (
_t_-test) data in a published study
· Differentiate between dependent observations and independent observations
[Previous](https://waldenu.instructure.com/courses/157739/modules/items/6373696)
[Next](https://waldenu.instructure.com/courses/157739/modules/items/6373710)
**## NURS_8211_WK4_ComparisonofMeansPart1.pptx**
NURS 8211
Research for an evidence-based practice
Week 4:Comparison of Means Part 1
Week 4: Comparison of Means Part 1
Objectives
Interpret a comparison of means (t test) data in a published study.
Differentiate between dependent and independent observations.
2
Appraising Evidence
Tips on reading a research study
Qualitative vs. Quantitative
Finding the null hypothesis: what are the dependent and the independent variables implicit in the study?
How were they measured, in the same people or in different in the groups?
Dang et al. Critical Appraisal tool (Appendix E pp.298-302
Theory generating (qualitative) vs. theory testing research (quantitative)
Qualitative deals with words; quantitative deals with numbers and statistics
3
The impact of smart goals on diabetes Management in a pharmacist-led telehealth clinic
Dependent variable: A1c level
Independent variable: SMART Goals
Observations are from different people, although they are “matched”
Null Hypothesis: There will be no difference in A1c levels when the intervention group with SMART goals is compared to the control group (no SMART goals setting)
Read the abstract, you can immediately get a feel for the study right from the the first page.
Identify the variables: what is the dependent what is the independent.
Form a null and an alternate hypothesis.
Clearly, they expected to see a statistically significant outcome: that A1c levels in the intervention group would be lower over 3 months as compared to the control group over a similar 3-month period.
Discuss clinical vs. statistical significance.
4
Cook et al. (2024)
Take a look at the table in Cook et al. (2024) on p. 56 in the article. Table 1 includes descriptives statistics which are presented in each group (the intervention group and the control group).
This was at a VA outpatient clinic, and there were 50 in the intervention group and 50 in the control group. The researchers set specific inclusion and exclusion criteria. For example, they specified that patients should be approximately the same age + or – 5 years. So the mean age in the SMART goal group was 64 years with a (9.2) standard deviation. And the mean age in the control group was 65, just a year older on the average with a SD of 8.6. Although there was statistical testing on age, you can see that this met their preset criteria for differences between the groups on no more than + or – 5 years difference.
You can take at other variables that they collected, including Mean A1c, in each group 9.1 and 9.0, a bit higher than should be and only slightly different, but as there are descriptive statistics, you really can’t see statistical significance, yes or no from these descriptive statistics. You need an inferential test and a p value.
5
Cook et al. (2024)
Here is the change in A1c from baseline to the 3 month mark.
In the text of the article, on p. 55 they disclose that the comparison used a Student’s t test (another way to describe an independent samples t test.)
In this figure, on p. 56, you can see a nice bar graph showing that both groups had a reduction in A1c: the SMART goal group, decreased 1.2 points, and the control group also decreased by a bit less (to .85). The p value that resulted from the independent samples t test was p = .287 as you can see in the note at the bottom of the graph.
Because the p value was so much larger that .05, we cannot reject the null hypothesis. Statistically we cannot say with any statistical confidence that these two numbers are due to the intervention. The difference is more likely, just due to chance.
Although this is true, the reduction in A1c is indeed clinically significant and this is validated in the study on page 58. Take a look in the discussion.
Now what might explain this result and what might be some suggestions for the future.
The first thing that comes to mind is that this might be a type II error…….Preventing a type 1 error (the risk of finding significance when it is really there) is really protected by the p value, set at .05 to reduce this risk.
But a type II error occurs when you don’t find significance but it is really there. And a very small difference, like this one, is definitely a greater risk. There is a technique called power analysis that researchers often use a priori or in advance. To detect a small effect size (that what they had here) with a 95% power these researchers would have need 131 cases in each group (262 total). As they only had a total of 100 cases, you can see one pertinent reason as to why they did not find statistical significance.
The Mann Whitney U test is the nonparametric equivalent of the independent samples t test. When the data do not meet the requirements (the assumptions) of the parametric test, they often substitute a nonparametric test instead. IN general, parametric tests are stronger than nonparametric tests. In published studies, sometimes the authors will describe the reasons for a choice of a statistical test. Sometimes they do not.
6
Improving critical care nurses perceived self-efficacy in providing palliative care: a quasi-experimental study
Dependent variable: nurses’ self efficacy in palliative care
Independent variable: an educational process using online videos
Observations are from the same nurses pre to post
Null Hypothesis: There will be no difference in self efficacy as measured in nurses before and after an educational process
Let’s use the same method we used before to take a look at this study. Defusco et al. (2023)
The nurses completed a 12 item survey that measured nurses’ self efficacy on palliative care for both psychosocial support and symptom management before and after viewing 5 videos and 2 documents from the VitalTalk website.
A convenience sample of self-identified nurses was used.
Power analysis was conducted effect size of .45 (moderate difference…..remember, I used .20 for the Cook study).
Sample of 41 participants . Ended up with 40.
Because the observations (the results of the survey pre to post) are from the same people, they are said to be DEPENDENT. (don’t confuse this language with the dependent and independent variables). It just indicates that because the observations are from the same people, you can use a paired t test or its non-parametric equivalent, the Wilcoxon Signed Rank test.
The nonparametric choice was a good one in this study. That’s because the paired t test has some requirements (these are called assumptions) and one of them is that the data are normally distributed. With a small sample, that can be challenging. The authors explain this on p.119 in the data analysis plan.
Null hypothesis is: There will be no difference in self efficacy as measured in nurses before and after an educational process.
Let’s look at the results.
7
Defusco et al. (2023)
Descriptives: Medians were used because data were not normally distributed
Wilcoxon Signed Rank test used to measure nurses’ self efficacy on overall self efficacy and two subscales pre and post:
Overall: 38 to 43.5 z = -4.868 p < .001
Psychosocial subscale: 18.0 to 21.0 z = -4.867, p < .001
Symptom management: 21 to 22.5 z = -3.861, p <.001
8
Key Points
Using a deliberate approach to reading a research study has benefit.
The Dang et al. (2022) critical appraisal tool asks some pertinent questions about published studies that help you to summarize carefully and thoroughly.
Although Cook et al. (2024) did not achieve statistically significant results, the outcome was clinically significant and useful.
DeFusco et al. (2023) did achieve both statistical and clinical significance in their study of nurses’ self efficacy in palliative care.
9
## image1.jpg
## image2.jpg
## image6.jpg
## image5.jpg
## image8.png
## image9.png
## image3.jpg
## image4.jpg
**TheImpactofDiabetes.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/theimpactofdiabetes-pdf)
**PalliativeCommunication.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/palliativecommunication-pdf)
**CriticalCareNurse.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/criticalcarenurse-pdf)
**## WK4DISCDATA.docx**
Self-Study: Comparison of Means, Part 1
Throughout the course, there will be a self-study Discussion pertaining to an important concept or topic covered within the assigned week. These Discussions are designed to give you the opportunity to collaborate with your peers and faculty, test your knowledge, ask questions, practice research analysis, and assist your peers.
You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
Resources
Be sure to review the Learning Resources before completing this activity.
Click the weekly resources link to access the resources.
[**WEEKLY RESOURCES**](https://waldenu.instructure.com/courses/157739/modules/items/6373691)
[Required Resources](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Dang, D., Dearholt, S. L., Bissett, K., Ascenzi, J., & Whalen, M. (2021).
_Johns Hopkins evidence-based practice for nurses and healthcare professionals: Model & guidelines_ (4th ed.). Sigma Theta Tau International Honor Society of Nursing.
· Chapter 6, “Evidence of Appraisal: Research” (pp. 155–158)
· Appendix E, “Research Evidence Appraisal Tool” (pp. 298–302)
· Appendix F, “Nonresearch Evidence Appraisal Tool” (pp. 307–314)
· Appendix G, “Individual Evidence Table” (pp. 315–318)
· Salkind, N., & Frey, B. (2019).
_Statistics for people who (think they) hate statistics_ (7th ed.). SAGE Publications.
· Chapter 12, “t(ea) for Two: Tests Between the Means of Different Groups” (pp. 236–239, 243)
· Chapter 13, “t(ea) for Two [Again]: Tests Between the Means of Related Groups” (pp. 253–256, 259)
[Required Media](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Niedz, B. (2024).
_Comparison of means, part 1_ [Video]. Walden University Canvas. http://waldenu.instructure.com
[PowerPoint Presentation](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
[Required Resources for Topic: \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· Cook, H. E., Garris, L. A., Gulum, A. H., & Steber, C. J. (2024).
[Impact of SMART goals on diabetes management in a pharmacist-led telehealth clinicLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/08971900221125021).
_Journal of Pharmacy Practice, 37_(1), 54–59.
[Required Resources for Topic: Paired \
_t_ Test](https://waldenu.instructure.com/courses/157739/pages/week-4-learning-resources?module_item_id=6373691)
· DeFusco, C., Lewis, A., & Cohn, T. (2023).
[Improving critical care nurses perceived self-efficacy in providing palliative care: A quasi-experimental studyLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://journals.sagepub.com/doi/10.1177/10499091221094313).
_American Journal of Hospice and Palliative Medicine, 40_(2), 117–121.
· Markiewicz, A., Hickman, R. L., McAndrew, N. S., & Reimer, A. (2023).
[Enhancing palliative communication in the intensive care unit through simulation: A quality improvement projectLinks to an external site.](https://go.openathens.net/redirector/waldenu.edu?url=https://www.sciencedirect.com/science/article/pii/S1876139923000075).
_Clinical Simulation in Nursing, 77_, 1–5.
**To prepare:**
· Read and view the Learning Resources in Doherty & Skalsky and in Dang et al. (2021) in Required Readings.
· View the video on comparison of means.
**Use** this Discussion to collaborate with your peers and faculty as an open office hours/ Q&A forum.
**Post** answers to the following:
· Interpret independent sample
_t_-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples
_t_ test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired
_t_ test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
·
For this Self-Study Discussion, you may post throughout Week 4. You are not required to post to this forum; however, you are encouraged to post, review the posts of others, as well as answer questions and/or provide clarity and collaboration with your peers. Discussions will be graded as either Complete or Incomplete.
**Post** answers to any or all of the following:
· Interpret independent sample t-test results and explain the relevance to the research question (Cook et al., 2024).
· Compare and contrast Cook et al. (independent samples t-test) to DeFusco et al. (2023) or Markiewicz et al. (2023) (paired t-test) results.
· How did the subjects differ in the two studies?
· Differentiate between clinical value and statistical significance.
Our interactive discussion addresses the following learning objectives:
· Interpret comparison of mean (
_t_-test) data in a published study
· Differentiate between dependent observations and independent observations
[Previous](https://waldenu.instructure.com/courses/157739/modules/items/6373696)
[Next](https://waldenu.instructure.com/courses/157739/modules/items/6373710)
**## NURS_8211_WK4_ComparisonofMeansPart1.pptx**
NURS 8211
Research for an evidence-based practice
Week 4:Comparison of Means Part 1
Week 4: Comparison of Means Part 1
Objectives
Interpret a comparison of means (t test) data in a published study.
Differentiate between dependent and independent observations.
2
Appraising Evidence
Tips on reading a research study
Qualitative vs. Quantitative
Finding the null hypothesis: what are the dependent and the independent variables implicit in the study?
How were they measured, in the same people or in different in the groups?
Dang et al. Critical Appraisal tool (Appendix E pp.298-302
Theory generating (qualitative) vs. theory testing research (quantitative)
Qualitative deals with words; quantitative deals with numbers and statistics
3
The impact of smart goals on diabetes Management in a pharmacist-led telehealth clinic
Dependent variable: A1c level
Independent variable: SMART Goals
Observations are from different people, although they are “matched”
Null Hypothesis: There will be no difference in A1c levels when the intervention group with SMART goals is compared to the control group (no SMART goals setting)
Read the abstract, you can immediately get a feel for the study right from the the first page.
Identify the variables: what is the dependent what is the independent.
Form a null and an alternate hypothesis.
Clearly, they expected to see a statistically significant outcome: that A1c levels in the intervention group would be lower over 3 months as compared to the control group over a similar 3-month period.
Discuss clinical vs. statistical significance.
4
Cook et al. (2024)
Take a look at the table in Cook et al. (2024) on p. 56 in the article. Table 1 includes descriptives statistics which are presented in each group (the intervention group and the control group).
This was at a VA outpatient clinic, and there were 50 in the intervention group and 50 in the control group. The researchers set specific inclusion and exclusion criteria. For example, they specified that patients should be approximately the same age + or – 5 years. So the mean age in the SMART goal group was 64 years with a (9.2) standard deviation. And the mean age in the control group was 65, just a year older on the average with a SD of 8.6. Although there was statistical testing on age, you can see that this met their preset criteria for differences between the groups on no more than + or – 5 years difference.
You can take at other variables that they collected, including Mean A1c, in each group 9.1 and 9.0, a bit higher than should be and only slightly different, but as there are descriptive statistics, you really can’t see statistical significance, yes or no from these descriptive statistics. You need an inferential test and a p value.
5
Cook et al. (2024)
Here is the change in A1c from baseline to the 3 month mark.
In the text of the article, on p. 55 they disclose that the comparison used a Student’s t test (another way to describe an independent samples t test.)
In this figure, on p. 56, you can see a nice bar graph showing that both groups had a reduction in A1c: the SMART goal group, decreased 1.2 points, and the control group also decreased by a bit less (to .85). The p value that resulted from the independent samples t test was p = .287 as you can see in the note at the bottom of the graph.
Because the p value was so much larger that .05, we cannot reject the null hypothesis. Statistically we cannot say with any statistical confidence that these two numbers are due to the intervention. The difference is more likely, just due to chance.
Although this is true, the reduction in A1c is indeed clinically significant and this is validated in the study on page 58. Take a look in the discussion.
Now what might explain this result and what might be some suggestions for the future.
The first thing that comes to mind is that this might be a type II error…….Preventing a type 1 error (the risk of finding significance when it is really there) is really protected by the p value, set at .05 to reduce this risk.
But a type II error occurs when you don’t find significance but it is really there. And a very small difference, like this one, is definitely a greater risk. There is a technique called power analysis that researchers often use a priori or in advance. To detect a small effect size (that what they had here) with a 95% power these researchers would have need 131 cases in each group (262 total). As they only had a total of 100 cases, you can see one pertinent reason as to why they did not find statistical significance.
The Mann Whitney U test is the nonparametric equivalent of the independent samples t test. When the data do not meet the requirements (the assumptions) of the parametric test, they often substitute a nonparametric test instead. IN general, parametric tests are stronger than nonparametric tests. In published studies, sometimes the authors will describe the reasons for a choice of a statistical test. Sometimes they do not.
6
Improving critical care nurses perceived self-efficacy in providing palliative care: a quasi-experimental study
Dependent variable: nurses’ self efficacy in palliative care
Independent variable: an educational process using online videos
Observations are from the same nurses pre to post
Null Hypothesis: There will be no difference in self efficacy as measured in nurses before and after an educational process
Let’s use the same method we used before to take a look at this study. Defusco et al. (2023)
The nurses completed a 12 item survey that measured nurses’ self efficacy on palliative care for both psychosocial support and symptom management before and after viewing 5 videos and 2 documents from the VitalTalk website.
A convenience sample of self-identified nurses was used.
Power analysis was conducted effect size of .45 (moderate difference…..remember, I used .20 for the Cook study).
Sample of 41 participants . Ended up with 40.
Because the observations (the results of the survey pre to post) are from the same people, they are said to be DEPENDENT. (don’t confuse this language with the dependent and independent variables). It just indicates that because the observations are from the same people, you can use a paired t test or its non-parametric equivalent, the Wilcoxon Signed Rank test.
The nonparametric choice was a good one in this study. That’s because the paired t test has some requirements (these are called assumptions) and one of them is that the data are normally distributed. With a small sample, that can be challenging. The authors explain this on p.119 in the data analysis plan.
Null hypothesis is: There will be no difference in self efficacy as measured in nurses before and after an educational process.
Let’s look at the results.
7
Defusco et al. (2023)
Descriptives: Medians were used because data were not normally distributed
Wilcoxon Signed Rank test used to measure nurses’ self efficacy on overall self efficacy and two subscales pre and post:
Overall: 38 to 43.5 z = -4.868 p < .001
Psychosocial subscale: 18.0 to 21.0 z = -4.867, p < .001
Symptom management: 21 to 22.5 z = -3.861, p <.001
8
Key Points
Using a deliberate approach to reading a research study has benefit.
The Dang et al. (2022) critical appraisal tool asks some pertinent questions about published studies that help you to summarize carefully and thoroughly.
Although Cook et al. (2024) did not achieve statistically significant results, the outcome was clinically significant and useful.
DeFusco et al. (2023) did achieve both statistical and clinical significance in their study of nurses’ self efficacy in palliative care.
9
## image1.jpg
## image2.jpg
## image6.jpg
## image5.jpg
## image8.png
## image9.png
## image3.jpg
## image4.jpg
**TheImpactofDiabetes.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/theimpactofdiabetes-pdf)
**PalliativeCommunication.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/palliativecommunication-pdf)
**CriticalCareNurse.pdf**
This file is too large to display. [View in new window](https://www.sweetstudy.com/files/criticalcarenurse-pdf)