Independent Sample t-test

okay so this is going to be the lesson on the independent sample t-test we have already opened up our data and we can see that we have a data with data set with a five different columns and a bunch of rows in this first column we see it's called species we have a second one called education health money and today called health - if we go down to variable view we can look up and see what these mean we can see that species is referring to some kind of type of alien species and if we click on values we can see that a1 represents some species called the Zetas cons and the two represents some species called the Holly Hobbs if we look at education this is referring to the years of Education these species have had health seems referring to their health before a new drug money's referring to their overall income and health - is some health after their new drug so if we go back to data view one question we may want to ask is do the saticons and Holly Hobbs differ on education do they differ on the amount of money they make do they differ on their health prior to the drug and do they differ on their health after the drug in all cases because these aliens are presumably either a Zetas Khan or a Holly hop a member cannot be one of each condition therefore our variable species is an independent variable with independent levels it is a between-groups variable so we're going to be wanting to use an independent sample t-test to do this we're going to go to analyze compare means and then we're going to go to independent sample t-test you will see in this same box we also have the one sample t-test the paired sample t-test and the one-way anova these are all different types of analyses that you might need to do at some point so we're going to click on independent sample t-test and we're going to get the following window now in this window we have a couple different things here we have a list of our variables our column labels we have something here called test variable and something here called grouping very the grouping variable is essentially the variable whose groups we want to compare in this case we want to compare the variable alien species to see if the groups of Zetas cons and holley ops differ so all we're going to do is we're going to highlight alien species and we're going to click this arrow and it's going to move species over here now you'll notice that there are two question marks remaining here you will not be able to do the analysis until these question marks are gone in order to make them go away we want to define our groups what SPSS wants to know is what are what are the groups we want to compare now we can't click on Group one and type in Zetas cons or Holly hop so we're going to need to put in the numerical symbol the nominal scale that we use to define these groups for instance the number one in our data set was for Zetas cons and we'll put a 1 in here and number two was the other group that was for Holly hump so we're going to put in a 1 in the 2 will click continue now let's say we want to compare these groups to see how they differ when they're years of Education we'll simply click on years of education and move it over to the test variable once we've done that we can click OK and now our output screen will occur now you'll see here in our group statistics this gives us our mean standard deviation standard error of the mean as well as the number of participants in our variable alien species is our independent variable years of education is our DV and we have 20 people in both here are the means and standard deviations the mean certainly look different what we're looking to do next is to determine are they significantly different if we go down to the independent sample test you're going to want to look over years of Education we're always going to say in the top row equal variance is assumed and we're going to go over to the t-test for equality of means what we're looking for are these three columns here this is your T value their T observed these are your degrees of freedom and this is your significance value think of it as your p value if this number is less than point 0 five if it is any number less than point zero five we would say the group significantly differ meaning that the Zetas Khan's and Hawley Hobbes do in fact differ on diverters are measuring in this case years of education and we would reject the null hypothesis which indicated that they did not differ from each other in this case we would reject the null with 38 degrees of freedom at E observed of nine point five and a p-value of less than point zero zero one let's say we wanted to run other independent sample tests on the other variables we can click on independent sample t-test again we get removed years of education and we can actually highlight all three of the remaining variables and we can move them all over at once and we click OK if we did this we would see we again get our groups at two six this time four before the new drugs income and after new drugs and will run all three t-test at the exact same time we can see that four before new drugs the significance value is greater than point zero five meaning that the groups don't differ for income it's less than point zero five they do differ and after new drugs it's smaller but it's still not less than point zero five we would say the groups don't differ on their health either before after new drugs but they do differ on income to determine how they differ we can just go back to our means it looks like the Zetas Khan's make more than do the Holi humps and if we go all the way up here it looks like the Zetas Khan's do in fact have more education than do the Holi humps that concludes the lecture on the independent sample t-test

Channel: bernstmj

Views: 323,796

Rating: 4.8419867 out of 5

Keywords: Statistics, Psychology, Academia, Math

Id: oIpzdTc0reI

Length: 6min 5sec (365 seconds)

Published: Mon Jun 20 2011