Week 6 - Assignment: Interpret a Correlation Analysis

SAGE Research Methods Video

Correlation and Regression - Pearson

Video Title: Correlation and Regression - Pearson

Originally Published: 2016

Publication Date: Sep. 30, 2016

Publishing Company: SAGE Publications, Inc.

City: Thousand Oaks, United States

ISBN: 9781506359212

DOI: https://dx.doi.org/10.4135/9781506359212

(c) SAGE Publications Inc., 2017

HERSCHEL KNAPP: Welcome to Practical Statistics for Nursing Using SPSS. This video shows how to process the Pearson correlation and regression. You can watch the entire video or use the time slider to navigate directly to any time point. [Correlation and Regression - Pearson, Overview] Correlation and regression analysis

HERSCHEL KNAPP [continued]: computes the nature of the relationship between two continuous variables. The relationship can be characterized using two parameters, direction and strength. The regression ranges between -1 and +1. The regression sign indicates the direction of the correlation. Positive correlations occur when the variables

HERSCHEL KNAPP [continued]: move in the same direction. When x goes up, y goes up. Or when x goes down, y goes down. Negative correlations occur when the variables move in opposite directions. When x goes up, y goes down. Or when x goes down, y goes up.

HERSCHEL KNAPP [continued]: The regression value indicates the strength of the correlation. Values nearer to -1 or +1 are stronger than values nearer to zero. To better conceptualize the data, a scatterplot with a regression line is useful. Each point represents two scores gathered from each individual.

HERSCHEL KNAPP [continued]: For example, this point represents two scores gathered on one of the patients surveyed. The patient had a length of stay of 12 days and a depression score of 58. The regression line can be thought of as the average pathway through the points. The positive slope suggests that lower length of stay is associated with lower depression scores,

HERSCHEL KNAPP [continued]: and higher length of stay is associated with higher depression scores for this group of patients. To better comprehend the notion of regression, consider these three examples. Here we see a strong positive correlation between number of homework hours in quiz scores, where lower homework hours are paired with lower quiz scores

HERSCHEL KNAPP [continued]: and higher homework hours are paired with higher quiz scores. In the second scatterplot, we see a strong negative correlation between alcohol consumption and quiz scores, where higher alcohol consumption is paired with lower quiz scores and lower alcohol consumption is paired with higher quiz scores.

HERSCHEL KNAPP [continued]: Finally, we see a fairly weak correlation between baseball-throwing skills and quiz scores, where baseball-throwing skills have virtually no correlation with quiz scores. [Correlation and Regression - Pearson, Pretest Checklist] Before running a Pearson correlation or regression analysis, there are three pretest criteria

HERSCHEL KNAPP [continued]: that need to be met. First, the data for each of the two groups should be normally distributed. We can check for this by observing a histogram with a normal curve for each group. The second and third criteria, linearity and homoscedasticity, can be verified by observing the scatterplot with the regression line.

HERSCHEL KNAPP [continued]: This example uses the dataset Ch 11 - Example 01 - Correlation and Regression.sav. This dataset contains three variables. Patient ID is a string variable, along with two continuous variables, Length of Stay

HERSCHEL KNAPP [continued]: and Depression scores for each patient. To check for normality, order histograms with normal curves for the two variables that will be involved in the correlation, Length of Stay and Depression. Click on Analyze, Descriptive Statistics, Frequencies. Move Length of Stay and Depression into Variables

HERSCHEL KNAPP [continued]: and click Charts. Select Histogram with Normal Curve. Click Continue, and uncheck Display Frequency Table. Click OK, and it'll process. The symmetrical curve

SAGE (c) SAGE Publications Inc., 2017

SAGE Research Methods Video

Page 2 of 3 Correlation and Regression - Pearson

on the histogram for Length of Stay shows a normal distribution.

HERSCHEL KNAPP [continued]: And the curve on the histogram for Depression also shows a normal distribution. The pretest criteria of normality is satisfied. To finalize the pretest checklist, we'll order a scatterplot with a regression line. This will also give us a more comprehensive understanding of the relationship between Length of Stay and Depression.

HERSCHEL KNAPP [continued]: Click on Graphs, Chart Builder. In the Choose From list, select Scatter/Dot and select the Simple Scatter option. Drag Length of Stay to the x-axis and Depression to the y-axis. Click OK, and it'll process.

HERSCHEL KNAPP [continued]: To order the regression line, double-click on the scatterplot and click on the Add Fit Line at Total icon. In terms of linearity, we see that the points lie in a fairly straight line. There are no unexpected curves or twists in the arrangement of the points. This satisfies the linearity criteria.

HERSCHEL KNAPP [continued]: As for homoscedasticity, we see that the field of points is thicker in the middle and tapers at the ends. This satisfies the homoscedasticity criteria. [Correlation and Regression - Pearson, Test Run] To process a Pearson correlational analysis, click on Analyze, Correlate, Bivariate.

HERSCHEL KNAPP [continued]: Move Length of Stay and Depression to variables. Click OK, and it'll process. [Correlation and Regression - Pearson, Results] The correlation table shows a strong positive correlation of 0.789 between Length of Stay and Depression. We also see that the P value is less than 0.05,

HERSCHEL KNAPP [continued]: suggesting that this is a statistically significant correlation. This concludes this video.

SAGE (c) SAGE Publications Inc., 2017

SAGE Research Methods Video

Page 3 of 3 Correlation and Regression - Pearson