Pearson Correlation - SPSS

hi welcome to how to stats comm in this video I'm going to demonstrate something relatively easy which is a Pearson correlation so this is very basic univariate analysis that is performed very very frequently and is very useful statistic and actually forms the basis of a large number of other more sophisticated useful statistics like multiple regression and factor analysis now when you're interested in performing a Pearson correlation typically you should be performing your analysis you have a hypothesis that one variable is associated with another variable and that could be a positive correlation or a negative correlation and a Pearson correlation is one method of estimating the association between two variables that are scored on a interval or ratio level in this case I've got sample data set that I pulled up from the samples in SPSS it's called bank loan dot save and it's got some ink it's got a it's got an age variable with people's ages in years and it's also got an income variable and my hypothesis is that older people will tend to have larger incomes so I'm expecting a positive correlation between age and income as age increases income increases so to calculate a Pearson correlation going to analyze correlate bivariate and you put your age variable into the variables box and household income in thousands income goes into that box now in the bottom here we've got P correlation coefficients and Pearson is the default which is checked and there's also a couple of other ones and I'll do separate videos on those we got tested significance to tail then flag significant correlations and that just means it's going to put an asterisk next to the correlation if it's statistically significant that's not especially useful when you're only doing a court with a single correlation but when you have a correlation table that's become more useful so I click OK we can see the syntax at the top of the output file this is how you before is what you have to type into the syntax if you were to go through the syntax editor and we have the correlation table right here and this is actually the totality of the output we have a correlation matrix so the correlation is actually reported twice but it's actually the same value it's the same estimate of the same effect so agen years correlated with household income in thousand is 0.47 six and our two Asterix here and you can see at the bottom of the table two asterisks correlation is significant at the P a significant at point zero one level in fact it's it's less than point zero one you can see that the significance level is point zero zero zero and you can double click that to see the sample size is very large so it's going to be very large now it's in scientific notation but we can see that at a very very statistically significant like p equal point zero zero forty nine times zero zero point two five level of significance so anyway at lille s than point zero five and def and less than point zero one it's even less than point zero zero one and the final piece of information is the sample size so we get the correlation to three decimal places usually you'd only ever report a correlation at two decimal places you have the significance level P less than point zero zero one and you have the sample size of 850 and it's duplicated that information is duplicated here it's simply saying the correlation between household income in thousands and age is in this section of the matrix and the correlation between age and household income is in this section of the bottom of the lower portion of the of the table and now and the higher portion look and age in years and agent ears this number one here represents a perfect linear correlation between age in years and age in years it's really just a it's an effect of having a matrix that you're having that SPSS is having the output age and years correlation and household income with household income correlation it's basically nonsense information but it's just it's included in the output it doesn't give you any significance level or anything like that so this is how you perform and identify the results associated with that Pearson correlation in SPSS thanks for watching

Channel: how2stats

Views: 1,093,261

Rating: 4.745863 out of 5

Keywords: correlation, Pearson, SPSS

Id: VOI5IlHfZVE

Length: 5min 0sec (300 seconds)

Published: Thu Sep 01 2011