Computing Central Tendency in SPSS (5-5)
Articles,  Blog

Computing Central Tendency in SPSS (5-5)


Let’s learn how to compute the three measures of central tendency: mean, median, and mode using SPSS. To follow along with this example you should download the data set StatsClass.sav to your desktop and open it in SPSS. Alright, so here We are in SPSS, and we are going to calculate the mean, median, and mode for some variables in this data set. Again, this is StatsClass.sav To calculate the mean, median, and mode there are actually several ways that we could go about it, but here’s the one that we’re going to use. Go to the Analyze menu, so Analyze, Descriptive Statistics, Frequencies This opens up our Frequencies dialog box. We’re going to put two variables into the SPSS variables box, and the variables that we are going to use are the standardized IQ score (that’s IQ). Move that one in and attitude towards statistics (that is StatATT). Let’s move that one, If you would like to see the labels for the variables rather than their names, you can right-click on your PC, or control click on your Mac and change to “display variable labels” So again, standardized IQ scores and attitude towards statistics scores. Next we want to click on the statistics button, and this opens up a new dialog box. Here in the upper right of this new dialog box you can see mean, median, and mode. Click on all three of those to select them. Let’s click on continue, and do one more thing before we select Ok. By default, SPSS will show us the frequency table for these distributions. We don’t want that here, so let’s unclick on “display frequency tables”, and now we’re ready to click OK. The output window opens, and here are our scores for standardized IQ score. The mean: 97.335. Median: 97.3 Mode: 97.3. For attitude towards statistics, the mean: 5.64 The median: 6.00 The mode: 1 So the first question: which of these two distributions is normally distributed? IQ scores or attitude toward statistics? The answer is: IQ scores because the mean, median, and mode for IQ scores very, very similar. Whereas, The mean, median, and mode for attitude toward statistics not even close, so this is not normally distributed. This is normally distributed. Great. Let’s do one more thing. Can you make a histogram for both of these distributions? Yes, we can. Let’s do that now. I’m going to minimize the output window. Move that out of the way. I’m going to go to the same dialog box. Analyze>Descriptive Statistics>Frequencies And this time I’m going to choose Charts To get a histogram, click on Histograms. And I want to add a normal curve because ,remember, I said one was normally distributed one was not. Let’s check that out. Let’s superimpose a normal curve on our histogram to find out do the Data match what we think that they should, given whether the distribution is normally distributed or not. Let’s click Continue, and OK. Our new output gets added below our old output. SPSS runs the same analysis as it has just run, so if I scroll up, there’s the first analysis. I ran that. It ran it again. Those numbers are exactly the same. There’s the 97.3. Scroll down a little bit more. I said that IQ scores should be normally distributed. Based on the fact that the mean, median, and mode were almost the same and what do we see? Yeah, that normal curve that we superimposed matches up really closely with the actual data. So yes, that’s normally distributed. It doesn’t look exactly like the normal curve, but it doesn’t have to. Normality is something that we can see both mathematically and through pictures. Now, what did we say about the attitude towards statistics? We said that was not normally distributed, so let’s check it out. Notice how that superimposed bell curve doesn’t match at all the general pattern that we see with attitudes toward statistics, so that is not a normally distributed data set. And we see that visually matches what we saw with the pictures as well. So that is how you calculate central tendency in SPSS.

3 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *