Transcript:
Hello, this is Dr. Grande. Welcome to my video on creating and interpreting a scatter plot matrix using SPSS, oftentimes in counseling research. We want to examine the relationships between variables and it’s helpful if we have several variables to see those relationships all at the same time, and that’s where a scatter plot matrix is particularly useful so looking at these fictitious data have loaded in the data view. I have an ID variable. And there are 45 participants, an independent variable program that has three levels, individual counseling, group counseling and treatment as usual and three dependent variables, And these are all reported as T scores. T scores have a mean of 50 and a standard deviation of 10 we have severity functioning and motivation and we’ll presume these data are gathered from a substance use treatment facility, where these type of variables would be of interest. And then I have a variable named RV, One random variable one, which is a random variable generated by the RV, normal function and SPSS, so the random values that should follow a normal distribution. So I’m going to show you a couple different ways to generate a scatterplot matrix and the first method. I’m going to go into graphs chart builder, and then you can see down here under gallery have choose from and there’s several different types of charts available. I’m going to move down to scatter dot and you can see here. There are eight different charts to choose from under Scatter Dot, and I want the one named Scatter Plot Matrix. So I’m going to drag this icon into the chart preview, and you can see it just says, scatter matrix down here, and that’s why I’m going to load the variables of interest so first I’m going to load severity functioning and motivation together, just those three, and then I’m going to click OK, so you will notice that all the combinations are represented here so along the x-axi’s, you have severity functioning and motivation and similarly along the y-axis you have severity functioning and motivation now we can interpret the scatter plot matrix directly from this view, But it’s a little easier. If we double click, we can make a few changes to this one is. I want to add a fit line at total and I want to leave it. As linear, which is the default fit line and then under chart size. I want to increase the size of this chart. A little bit you can see. The height is 375 by default here. I’m going to make this fort. Make this 450 like that. A little larger. I’m going to maintain the aspect ratio there, so click apply there and then under options you can see, there’s an option here. Show charts in the diagonal. So what that means is? If you look through the side of the way, if you look here at the scatter plot matrix, there’s no chart in the diagonals because this would be, for example or severity is the Y-axis and it’s also in the x-axi’s, so there’s no. There’s no chart that would be in here by default, similarly functioning functioning and motivation motivation. There’s nothing there, so you have two choices for this for so show charts in the diagonal. You have a histogram or normal curve, so I’m going to let it stay as default, which is histogram and then close this, so you can see the chart is larger and it has the histograms in the diagonal, so these histograms represent the variable in the Y Axis. So this is the histogram for severity Here in the center is the histogram for functioning and the one down here. The bottom right is the histogram for motivation and just to demonstrate this quickly. I’ll go to chart builder. I’m going to reset this. Go down to histogram and drag in a simple histogram and then put severity as the x-axis and click OK. So if you look up here at this histogram the general pattern of it here and move down to the Y just created, you can see they are identical so now taking a look at the relationships between the variables and interpreting what they mean, so if we look here at functioning in the y-axis and severity and the x-axi’s, you can see that the line moves from the top left to the bottom, right so that there is a strong negative correlation between these two variables as severity increases, functioning decreases and similarly as severity increases. Motivation decreases. All those you can see here. The relationship is not quite as strong as as the as was the case with severity and functioning then looking at functioning of motivation. We can see that as functioning increases, motivation increases, and, of course, viewed another way with motivation on the x-axis and function of y-axis as motivation increases functioning increases. So the scatter plot matrix tells us about the relationship, but doesn’t tell us anything about causality. It’s not suggesting that motivation causes a change in functioning but simply, there’s a relationship between the two in this case, a strong, positive relationship between these two variables, So let’s go back to graphs and charts. I’m going to reset this. Move back to scatter Dot and matrix of scatter plot matrix. Now I’m going to load in severity and hold down control, functioning motivation and random variable one. So I’m going to drag them all down into a chart preview. I’m going to click OK, And as you can see by default because we have four variables. We have 16 sells here. This is a little difficult to Rica’s the size, so I’m going to double click and first I’m going to change the chart Size 450 click apply, and then I’m going to add the fit line total under options. I’m going to display the histogram so this is a little easier reading, Alka, this larger and the only change here is I’ve added the random variable, so we wouldn’t expect a random variable to be associated with these other variables to a statistically significant degree, and as you can see by adding the random variable if we look at the random variable here on the x-axis and move up the line is it moves down word a little bit, but there’s really no relationship there with motivation and not much of one with functioning and even less a relationship with severity and, of course. The histogram. Down here in this bottom right, cell is for the random variable. I’m going to move back to the data view and show you another way to do this. You can go to graphs and set a chart builder. Your legacy dialogues and you can select scatter dot from here and you can see. It gives you these choices, and you want to select matrix scatter and click define, so if we want the same scatter plot matrix that I had before just hold down CTRL and select severity function, motivation and random variable 1 move them over and click. OK, and you can see, of course, it’s smaller. It doesn’t have the lines and histograms again. That’s change this by double clicking, and we can add all these features in fairly quickly, so we have the same result just using the legacy dialogues instead of the chart builder, so by interpreting the the scatter plot matrix and looking at the individual scatter plots, we have an idea about the relationships between these variables. Let’s take a look at that under, analyze correlate bivariate. I’m just going to throw all four of these variables be working with in two variables under options. I’m just going to add means and standard deviations and then click OK, and we can see that between severity and functioning. There is a statistically significant negative correlation as we had suspected, and then we have statistically significant negative correlation between severity and motivation and a non statistically significant result for the random variable again, which would make sense. We’d have the reason to believe the random variable would be associated with severity and in a strong way, it’s certainly not a statistically significant way, and if we look down functioning motivation, we have similar results non statistically significant results when compared to the random variable, we look at functioning and severity with the same value here negative point a one two, but if we look at functioning motivation, something, we didn’t see up here, we can see that is also statistically significant, It’s a positive relationship positive correlation, so these correlations confirm what we would have suspected from the scatter plot matrix, strong negative relationships between severity and functioning is very motivation, a strong, positive relationship between function motivation and really no relationship, not a statistically significant relationship anyway, but the random variable and any of the other three severity functioning or motivation. I hope you found this video in creating and interpreting a scatter plot matrix using SPSS to be helpful as always. If you have any questions or concerns feel free to contact me. I’ll be happy to assist you.