Skip to main content
SAGE
Search form
  • 00:00

    [MUSIC PLAYING][An Introduction to Correlation & Regression]

  • 00:11

    HERSCHEL KNAPP: Hi, I'm Dr. Herschel Knapp.This video provides an overview of correlation and regression.[Herschel Knapp, PhD, MSSW.Adjunct Lecturer, Department of Study.University of Southern California] A correlationand regression assess the relationshipbetween two continuous variables gathered from each memberof a single group.So we could ask the question, is therea significant relationship between the number

  • 00:31

    HERSCHEL KNAPP [continued]: of homework hours per week and the score on a Friday quiz?[Example]So for our example, we'll gather weekly homework hours and quizscores from one group of students.And our data will look like this.For each student, we'll have a homework score and a quiz

  • 00:53

    HERSCHEL KNAPP [continued]: score.Now, regression and correlation determine the relationshipbetween two continuous variables,and the relationship ranges in terms of their regressionscore, which can go from minus 1 to plus 1.The regression sign, either plus or minus,indicates the direction of the correlation.

  • 01:15

    HERSCHEL KNAPP [continued]: So for a positive correlation, we'llsee an r ranging from 0 to plus 1.And what this is indicating is that the variablesmove in the same direction.When one goes up, the other goes up, and vice versa.That's a positive correlation.Alternatively, we could have a negative correlation.This is where the correlation r, the regression score,

  • 01:39

    HERSCHEL KNAPP [continued]: ranges between minus 1 and 0.In this case, it's telling us that the variablesmove in different directions.When one goes up, the other goes down, or vice versa.The regression value indicates the strengthof the correlation.Correlations at minus or plus 1 are stronger than those nearer

  • 01:60

    HERSCHEL KNAPP [continued]: to 0.[Conducting Correlation & Regression]This is a scatterplot representing the coordinatesof quiz scores and homework scoresall together on one graph.In order to understand this, let's lookat each of the elements that are on this graph, the points

  • 02:22

    HERSCHEL KNAPP [continued]: and the line.We'll start with this point here.This point represents Johnny's score,which is 5 hours of homework and 80 on the quiz.The remaining points represent all of the other students'pairs of scores.The regression line represents the average pathwaythrough the points.And as we see here, we see a positive upward slope

  • 02:44

    HERSCHEL KNAPP [continued]: of the line.Now, if we were to look at this graphically again,we could see it with the regression score, 0.883.The 0.883 shows us that we have a strong positive correlationbetween the quiz score and the homework hours.In other words, when we see low homework hours,

  • 03:05

    HERSCHEL KNAPP [continued]: we're going to see a low quiz score.When we see high homework hours, we'regoing to see higher quiz scores.Another way we could have measured thisis what if we gathered the quiz score and weekly alcoholconsumption?In this case, the r comes out to minus 0.835,a negative correlation-- in fact, a very strong

  • 03:27

    HERSCHEL KNAPP [continued]: negative correlation.In this case, the variables are moving in different directionsfrom each other.Where we see high alcohol consumption,we see a low quiz score, and vice versa.And finally, suppose we measured quiz scoreand how far a student could throw a baseball?

  • 03:47

    HERSCHEL KNAPP [continued]: Now, at face value, we would expectthese variables wouldn't have muchof a correlation with each other.And as we see here, the r value of 0.047 is very close to 0.And with that, we see a virtual flatteningof the regression line.Also notice that as the regression score drops closerto 0, we see more of a scattering of the points,

  • 04:10

    HERSCHEL KNAPP [continued]: whereas the other graphs, we see that the points arehugging the line much closer.[Conclusion]For further details on conducting correlationand regression, please download the videosat sagepub.com/knapp.For further reading, please refer

  • 04:32

    HERSCHEL KNAPP [continued]: to Introductory Statistics Using SPSS, or Practical Statisticsfor Nursing Using SPSS, both by Herschel Knapp.[Further Reading. www.sagepub.com/knapp.Knapp (2013).Introductory statistics using SPSS.Knapp (2016).Practical statistics for nursing using SPSS.]

Video Info

Publisher: SAGE Publications Ltd

Publication Year: 2017

Video Type:Tutorial

Methods: Correlation, Regression analysis, Regression coefficient

Keywords: alcohol consumption; baseball; homework; mathematical concepts

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:

Keywords:

Abstract

Professor Herschel Knapp discusses regression and correlation in statistical calculations. Using hypothetical examples and visual aids, he demonstrates how correlation and regression reveal the relationship between continuous variables.

Looks like you do not have access to this content.

An Introduction to Correlation & Regression

Professor Herschel Knapp discusses regression and correlation in statistical calculations. Using hypothetical examples and visual aids, he demonstrates how correlation and regression reveal the relationship between continuous variables.

Copy and paste the following HTML into your website