Pearson r Correlation in SPSS - Part 3
HTML-код
- Опубликовано: 23 авг 2024
- How to calculate the correlation coefficient in SPSS is covered in this video. The correlation is also tested for significance and a scatterplot is constructed.
For more SPSS help: tinyurl.com/nyw...
Video Transcript: I had indicated earlier that I would show you what the linear relationship looks like. So let's go ahead and spend a few moments doing that now. And what I'll do is I'll go ahead and graph the relationship between hours media and college GPA for our 20 people. So to do that we want to go to Graphs and then go to Legacy Dialogs and then down here towards the bottom select Scatter/Dot. And then notice Simple Scatter has the box around it. We want to go ahead and go with this default option, so go ahead and click Define. And then here I'll put college GPA in my Y-axis box and I'll put hours media in my X-axis box. Now if you wanted to you could put hours media in the Y-axis and college GPA in the X-axis, in other words flip these variables and that would be fine as well. Next we'll click OK. And then here we get a graph which is output and notice I have all these circles here. There are actually 20 of these circles here, where each circle represents the values for a given individual in my data set. And the way to read this is, if I double-click this for a moment, this circle here notice how on my X-axis variable, the horizontal axis, which we also can think of as X, this is labeled hours media. So here is my hours media variable. And over here, this circle, notice it looks like it's about 30 hours of media. This individual, this is one person's score on hours of media, how much media they watched. And if you scroll over here to the Y-axis, labeled college GPA, we can see that their GPA was somewhere around more or less 2.3, maybe 2.35, but somewhere in there. So this point here is high on hours of media if you notice that they're all the way to the right, but they're low on college GPA as they're way down here. These other two points are high on hours media, but they're low on college GPA. Next, notice these points over here. This point is low on hours media, they're somewhere around 7 or so, but they're high on GPA; if you notice they're close to 4. This person is fairly low on hours media and they're fairly high on GPA, and so on. And then these people in the middle, they're moderate on hours of media, and they're moderate on their GPA, as far as a GPA was measured, which is about from 2 to 4 here. Notice this characterizes a negative relationship: high on one variable, low on the other, or low on that variable, high on the other. That's how we define a negative relationship. And there's one last thing we can do here. If we right-click this plot, and a shortcut menu opens here, and if we go ahead and select Add Fit Line at Total. Click on that. And then here we're going to select Linear, and then Close. And you'll see here this line which is used to represent, or fit, the data points in general. And notice the straight line is applied. This is exactly what we're doing here with Pearson correlation. We're using a correlation that estimates the linear or straight line relationship. And these points they do do a pretty good job of fitting a straight line here. As opposed to a situation where we don't have a linear relationship and if you think of something like a parabola if we have dots like this and they came around like a horseshoe or a 'U' upside down, that would be a nonlinear relationship, and we would not want to use Pearson's r for that. But in many situations our points do fall in a pattern that mimics a straight line relationship, as opposed to some nonlinear relationship, like a parabola. This concludes the presentation on correlation in SPSS. Thanks for watching.
SPSS videos statisticsinstructor/Quantitative Specialists:
RUclips channel: / statisticsinstructor
Udemy inferential course: www.udemy.com/...
Udemy descriptives course: www.udemy.com/...
Udemy ANOVA course: www.udemy.com/...
Udemy MANOVA course: www.udemy.com/...
What a life saver for my final project in SPSS! Thank you!
23jroyer
Glad to hear it helped, 23jroyer!
Superrrrrrbbbbbbb these three video of 12 minutes duration are juicer than my seven months of studying and trying to understand the correlations. May God bless you.
Thank you for the explanation. i was clueless about pearson, and now u have just saved the day.
What a huge sigh of relief-i just didnt this concept during our lectures.Now I have mastered a few tricks which i will use in an exam
Thanks for letting me know, Keith. Very glad to hear the video helped!
SO helpful, thank you for helping me get through my final project. This made everything easy to understand!!
Thank you. This is an excellent resource. I am using Pearson's r for my dissertation data analysis.
Thank you so much for explaining it very easy and step by step
Was verrryyy informative for me
Thank you so much
what a life saver! Thank you so much!
This saved my dissertation.Thank you!!
Thank you so much. It is very useful.
Very informative,Thanxs for the good explanation
Congratulations for your videos!! The step-by-step explanations make clear not only the way to conduct analyses, but also the assumptions and previous ideas neccesary to understand each technique. I would also like to ask for the way to run a normality test as an assumption needed to satisfy in several techniques...Should I run just one test for the whole grouping factor or a separate test for each level of such factor?
Thanks from Spain!!
Juan Antonio Arce Altamirano
Juan,
Thanks for the feedback - I'm glad you found them helpful!
I would run an overall test, but I would also run tests for the separate factors. This way, if you have unequal variances you can pinpoint exactly where they are and run alternative tests if necessary on these separately. If your sample sizes are equal (or very close) for each cell though you should be generally okay, as long as the variances are not substantially unequal.
Ron @ QS
OK. Thanks for your explanation. Now I 'll assess my investigations on a safer basis.
I was trying to read statistics textbooks and they were way too technical. I don't have much time to sink all the information in. But thanks to your videos, I was able to understand all those unfamiliar concepts. :)
Cristina Amarille
Thrilled to hear it, Christina! Thanks for letting us know!
Quantitative Specialists
My apologies for the misspell, Cristina.
Hi, Thanks for your nice video. I am wondering, is it possible to do correlation for unequal sample size?
this is very helpful to understand and It helped me a lot
Very beautifuly explained,
hi, how will i interpret the result in pearson r correlation if i used likert scale on my data?
thank you. very clear explanation and demonstration !
very informative and easy to follow.
Thank you so much!!!
Clear explanation. Thanks
Thank you this video, makes it easier to understand!
But what i needed to know is what i should use to do the correlation for a likert scale? Use mean score or sum?
Glad it helped!
As long as you are calculating the mean or the sum of the exact same number of items, the correlation will be the exact same either way, so it really doesn't matter. What usually determines which way it is done is the original instructions by the author of the scale. But the correlation is the same either way...
Thank you so much! U're doing a great job :)
Thank you so much
good job
many thanks for your vid
thank u so much! this helped me a lot
Glad to hear it helped, Kimniie. Thanks for letting me know!
Thank u sir , thanks so so so much.
thank you so much
very useful
Thanks for sharing ....thumb up
Mushtaq Ahmad
Thanks, Mushtaq. Glad to hear you enjoyed it!