NURS8035 Assessment 3 Latest 2021 January

NURS8035 Foundations of Evidence-Based Practice in Nursing

Assessment 3

Instructions: Normal Data Distribution and Two-Variable Correlation Testing

•             For this three-part assessment you will create a histogram or bar graph for a data set, perform assumption and correlation tests, and interpret your graphic and test results in a 2-to-3 page paper.

In this unit we focus on whether two or more groups have important differences on a single variable of interest. For example, for the dependent variable stress score, we may want to know if there is a difference in stress between males and females, or maybe we would like to know if there is a difference in stress levels between people who drink chamomile tea and those who do not, or maybe we would like to determine if a group of expectant parents is less anxious (this is the dependent variable) about the birthing experience after a series of discussions with experienced parents. In each of these examples we have two groups (two groups being compared or the same group being compared before and after), and one dependent variable that is being compared in each group. In this unit you will begin exploring popular statistical techniques (and their assumptions) that are used to compare two or more groups.

The independent t-test, also called unpaired t-test, is typically used in health care to compare two groups of individuals that are entirely unrelated to each other (that is, independent), thus the one group cannot influence the other group. For example, we may wish to compare a drug treatment group to a control group (those not receiving drug treatment) for a specific clinical characteristic (dependent variable) that can be measured at the interval or ratio level (such as cholesterol, depression scale, or memory test).

The dependent t-test, also called paired t-test, compares two groups for a dependent variable measured at the interval or ratio level as well; however, these two groups are in reality just one group. But because they are measured before and after an intervention, we consider them as two groups for analytical purposes. This group is considered dependent because nothing is expected to vary in the nature of the individuals being measured except as a result of the intervention, as the group is composed of the same individuals.

Overview

One of the most important steps along the researcher’s path to data analysis is to become familiar with the character of the raw data collected for the project. Before weaving the strands of data into an analytical story that is related to a study’s goals, researchers typically inspect the completeness and quality of the data with various visualization techniques (graphics), summary tables, and mathematical tests of quality (assumption tests), as discussed in Assessment 2. One of these latter tests is a correlation analysis. With this approach, the researcher performs a very basic series of exploratory tests on variable pairs to identify any potentially interesting (yet unknown) relationships between groups of data (variables). Correlational analyses are often later performed as part of the predetermined data analysis plan to answer a specific research question.

Demonstration of Proficiency

By successfully completing this assessment you will address the following scoring guide criteria, which align to the indicated course competencies.

•             Competency 1: Describe underlying concepts and reasoning related to the collection and evaluation of quantitative data in health care research.

o             Interpret the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress.

                Competency 2: Apply appropriate statistical methods using common software tools in the collection and evaluation of health care data.

o             Create a histogram and scatter plot for variables tested for normal distribution.

o             Perform a normal distribution assumption test for two variables to determine if data is normally distributed.

o             Perform an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables.

                Competency 3: Interpret the results and practical significance of statistical health care data analyses.

o             Interpret the effect size for correlation analysis results.

                Competency 5: Address assignment purpose in a well-organized text, incorporating appropriate evidence and tone in grammatically sound sentences.

o             Articulate meaning relevant to the main topic, scope, and purpose of the prompt.

o             Apply APA formatting to in-text citations and references.

Instructions

For this three-part assessment, complete the following, referring to Yoga Stress (PSS) Study Data Set [XLSX], which you have used previously, as needed.

Software

The following statistical analysis software is required to complete your assessments in this course:

                IBM SPSS Statistics Standard or Premium GradPack, version 22 or higher, for PC or Mac.

You have access to the more robust IBM SPSS Statistics Premium GradPack.

Please refer to the Statistical Software page on Campus for general information on SPSS software, including the most recent version made available to Capella learners.

Part 1: Graphic Representation of the Data from the Yoga Stress (PSS) Study Data Set

6.            Create a histogram or bar graph (according to the measurement level of the data) of the following variables: Age, Education, Pre-intervention Psychological Stress Score (PSS).

o             Refer to the following resources as needed while creating your histogram:

o             SPSS Tutorials. (n.d.). What is a histogram? Retrieved from https://www.spss-tutorials.com/histogram-what-is-it/

o             SPSS Tutorials. (n.d.). Creating histograms in SPSS. Retrieved from https://www.spss-tutorials.com/creating-histograms-in-spss/

o             Creating Histograms in SPSS.

7.            Create a scatter plot of the following pair of variables: Age versus Pre-intervention Psychological Stress Score (PSS).

o             Refer to the following resources, as needed, while creating your scatterplot:

o             Displaying Relationships: Scatterplot.

o             Interpreting Scatterplots.

Part 2: Statistical Tests

8.            Perform a preanalysis assumption test for a normal distribution test to determine if the data you intend to use for the correlation tests passes the assumption of being normally distributed.

o             You will use this test for Age and Pre-intervention Psychological Stress Score (PSS).

9.            Perform the appropriate correlation test to determine the direction and strength or magnitude of the relationship between these two variables from Step 1.

o             Remember, we are not concerned about causation at this point and want to determine only if there is a statistical association.

Part 3: Yoga Stress (PSS) Study Paper

•             Include the histogram and scatter plot graphics you created earlier for Age and Pre-intervention Psychological Stress Score (PSS).

o             Provide an interpretation for these graphics.

•             Report the statistical outcome of the correlation analysis using appropriate scholarly style, including a brief interpretation of the effect size of the correlation.

•             Interpret the practical, real-world meaning (and limitations of the interpretation) of the relationship of these two variables based on the correlation analysis you performed.

•             Include the SPSS “.sav” output file that shows your programming and results from Parts 1 and 2 for this assessment.

•             Provide at least one evidence-based scholarly or peer-reviewed article that supports your interpretation.

Normal Data Distribution and Two-Variable Correlation Testing Scoring Guide

CRITERIA              NON-PERFORMANCE     BASIC    PROFICIENT        DISTINGUISHED

Create a histogram and scatter plot for variables tested for normal distribution. Does not create a histogram for variables tested for normal distribution.   Creates a histogram for variables tested for normal distribution, but the histogram or curve is flawed.             Creates a histogram for variables tested for normal distribution.               Creates a histogram for variables tested for normal distribution and explains what the histogram shows.

Perform a normal distribution assumption test for two variables to determine if data is normally distributed.      Does not perform a normal distribution assumption test for two variables to determine if data is normally distributed.      Performs a normal distribution assumption test for two variables to determine if data is normally distributed, but the test is performed incorrectly or is otherwise flawed.           Performs a normal distribution assumption test for two variables to determine if data is normally distributed.        Performs a normal distribution assumption test for two variables to determine if data is normally distributed and explains how the test was conducted.

Perform an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables.    Does not perform an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables.               Performs an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables, but the test is performed incorrectly or is otherwise flawed. Performs an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables.                Performs an appropriate correlation test to determine the direction and strength or magnitude of the relationship between two variables, and explains how the test was conducted.

Interpret the effect size for correlation analysis results. Does not interpret the effect size for correlation analysis results.                Interprets the effect size for correlation analysis results, but the interpretation is inaccurate or otherwise flawed.                Interprets the effect size for correlation analysis results.               Interprets the effect size for correlation analysis results and explains how the effect size was identified.

Interpret the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress.         Does not interpret the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress.     Interprets the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress, but the interpretation in incomplete, inaccurate, or otherwise flawed.            Interprets the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress.          Interprets the overall clinical meaning and limitations of the relationship of two variables, based on a correlation analysis and literature regarding age and stress. Explains how the clinical meaning and limitations might affect future decisions.

Articulate meaning relevant to the main topic, scope, and purpose of the prompt.           Writing is unrelated to the assignment prompt.       Addresses a specific topic with unclear intent or insufficient depth.         Articulates meaning relevant to the main topic, scope, and purpose of the prompt.  Articulates a focused response to the assignment prompt and demonstrates a thorough understanding of the main topic, scope, and purpose.

Apply APA formatting to in-text citations and references.             Does not apply APA formatting to in-text citations and references.        Applies APA formatting to in-text citations and references incorrectly and/or inconsistently, detracting noticeably from good scholarship.            Applies APA formatting to in-text citations and references.         Exhibits strict and nearly flawless adherence to APA formatting of in-text citations and references.

Resources: Chi-Square Analysis

•             Chi-square is a very flexible data analysis tool to answer relationship or association questions that cannot be answered by the other inferential tests you have learned during this course. Although there is actually a “family” of chi-square tests, we will focus on just two of the most commonly used: the chi-square test for independence (also called chi-square test for association) and the chi-square goodness of fit test. The chi-square test for association is used to explore an association or independence between two variables where data for both variables is measured at the category level (also referred to as nominal data). The chi-square test is flexible enough to allow us to test data measured at the ordinal level, but we will stick with categorical data to keep things simple.

Interestingly, whether one variable is dependent or independent is not important in this test. All that matters is that there are two categorical data-type variables. Remember too that when categorical data has only two levels of possible answers (such as yes or no), then we say this is dichotomous data. If you see these terms, they probably refer to categorical data.

Understanding what chi-square is, as well as how to perform the analysis will be helpful as you complete this assessment.

Chi-square Analysis

•             Heavey, E. (2019). Statistics for nursing: A practical approach (3rd ed.). Burlington, MA: Jones & Bartlett. Available in the courseroom via the VitalSource Bookshelf link.

o             Chapter 8, “Chi-square.”

                Geher, G., & Hall, S. (2014). Straightforward statistics: Understanding the tools of research. New York, NY: Oxford University Press.

o             Read Chapter 13, “Chi-square and Hypothesis Testing With Categorical Variable.”

                DrNic’sMaths and Stats. (n.d.). Analysing data in a two-way table (including chi-squared test) [Video]. Retrieved from https://www.youtube.com/watch?v=jhz0ubW0EWk

                DrNic’sMaths and Stats. (n.d.). Understanding and calculating the chi-squared statistic in two-way tables [Video]. Retrieved from https://www.youtube.com/watch?v=qfxzG6FgVlM

Resources: Displaying Data

•             The following resources provide information on displaying and interpreting data from histograms and scatterplots. This type of visual display is often used when attempting to communicate the type of relationship between two variables.

Displaying Data

•             SPSS Tutorials. (n.d.). What is a histogram? Retrieved from: https://www.spss-tutorials.com/histogram-what-is-it/

•             SPSS Tutorials. (n.d.). Creating histograms in SPSS. Retrieved from: https://www.spss-tutorials.com/creating-histograms-in-spss/

•             Creating Histograms in SPSS.

•             Displaying Relationships: Scatterplots.

•             Interpreting Scatterplots.

Activity: SPSS Practice: Performing T-tests

•             PRINT

•             SPSS PRACTICE: PERFORMING T-TESTS

Context

To best prepare for the upcoming assessment, you must understand how to perform and interpret paired several different t-tests. SPSS makes it relatively easy to perform parametric testing.

To add to your statistical programming skills and prepare you for the next unit, we will be using SPSS to perform two different types of parametric t-tests.

Again, like the Assessment 1 practice activity, you will first complete tasks in SPSS and then answer questions in a formative activity, SPSS Practice: Performing T-tests, to check your work.

Activity Instructions

For this activity, use the data found in the Emotional Well-Being (SF-36) Study [XLSX] data set you created previously. You will have to use two t-tests: independent samples (unpaired) t-test and the dependent samples (paired) t-test.

INDEPENDENT SAMPLES (UNPAIRED) T-TEST

To determine the varying effects of the dietary treatment on males versus females (which are, admittedly, “independent” of each other), perform an independent samples t-test on the Well-Being dependent variable in male participants compared to the same scores in female participants.

Note: You will need to create a new dependent variable related to the treatment-related change in well-being scores (Hint: Change score = Post-Tx Well-Being – Baseline SF-36 Well-Being Score).

DEPENDENT SAMPLES (PAIRED) T-TEST

Next, to compare the effects of the Dietary Treatment on the well-being of males at baseline to well-being scores in the same males after treatment, perform a dependent (paired) samples t-test for the dependent variable Well-Being (Hint: use Baseline versus Post-Tx).

Save your work on the Emotional Well-Being data set. Once you have completed the practice tasks, leave SPSS open for reference and answer the questions in the SPSS Practice: Performing T-tests formative activity to check your work.