Two Sample t test using SPSS with Dr Ami Gates

hey everybody welcome to a how-to video with dr. Amy gates this video is going to focus on how to perform a two-sample independent t-test using SPSS but before we get started with that or looking at what hypothesis we're going to be working with let's take a look at the data set that we're using to perform this test that's always the first and most important thing to do when performing any statistical test is to really understand the data that you're working with so in this particular data set I have information about people in this case for example I know their gender and you might notice here that gender is actually a string male or female but in some data sets that we may have seen in other how-to-videos our gender was labeled as ones and twos that's why it's so important to look at the data set and understand how the data is being represented and in what form not only gender but we know the age of the people we know their height shoe size hair color the number of terms they've been in school if we continue to scroll over we even know kind of fun things like how many Facebook friends they have and over here how they feel about golf and sports now under golf and sports you may notice that there are numbers here so these numbers obviously represent something but we're not sure what it represents yet and so to determine what a variable represents or what its values mean we actually want to click on variable view on the lower left hand corner and take a closer look at that variable so when I click on variable view and I go and find the golf variable at the bottom this golf variable actually represents golf and sports does all sports in general they're listed here and the values that this variable can take on I can determine by clicking this so if I click it it tells me that one the number one represents least favorite are not liking it so anybody with a 1 doesn't like golf they don't like sports they're not sports oriented anybody with a 5 they do like golf they do like sports the very sports oriented so now we know what these numbers mean and let's go back to our data and have a close sir look so we'll click back on data view and here when we look at golf again we see that we've got all kinds of numbers between 1 and 5 and we know again that 5 is our people who really likes golf and sports and ones are people who really don't like them so what if I had a strange thought what if I wondered how does liking sports affect the number of Facebook friends a person might have it's an interesting question and I don't want to look at all the ranges I don't want people who are undecided like threes and fours and twos I want to look at the extreme ones I want ones and fives I want people who either really liked it or really don't and that's what I'm going to compare it's going to be my two groups so let's take a quick look at my hypothesis and what I'm going to be comparing and then we'll run our tests in SPSS and then we'll analyze all the results so first and again this is going to be an example of an independent to sample t-test using SPSS and our first job is to figure out what is our research question in this case my research question is does loving or hating golf or sports create a significant difference between the number of Facebook friends a person has kind of a strange question well my first group is going to be called Group one and I'm going to put all the people in Group one who have ones in our data set these are the people who really don't like golf or sports in my second group I'm going to put all the people who have fives in my data set these are the people who really love or like golf in sports and so my null hypothesis is that there is no significant difference between the mean number of Facebook friends for group 1 and the mean number of Facebook friends for group to remember the null hypothesis always has the equal sign it's the status quo and it assumes that there's no significant difference between your two research groups my alternative or research hypothesis H a is that there is a difference that the mean number of Facebook friends for my first group the people who don't like sports is actually not going to be the same as my mean number of Facebook friends for the people who do like sports have an interesting question and again my grouping variable is whether or not people like sports to the extreme either ones or fives and the variable that I'm using to make my comparison is going to be the number of Facebook friends that each group has that's why here in my hypothesis I'm looking at the mean number of Facebook group friends for each group all right let's see what happens next now in order to perform this test I'm going to go into SPSS I'm going to click on analyze I'm going to click on comparing means because that's what I want to do I'm looking at the means between two groups and in this case I'm going to be using an independent samples t-test why independent samples t-test because I have two separate samples of people that I want to compare I have the people who like sports and the people who don't there's no overlap between those two groups they either have a 1 meaning they really don't like it or a 5 meaning they really do like sports so it's an independent samples t-test as a side note notice that there are other types of T tests you can perform and further how-to videos will show these other T tests and how to know which of these tests you want in this case we have two sample means we're comparing and they're independent and that's why we're choosing this option now once I select it SPSS wants to know okay what's the variable you're using to compare your two groups and what are your groups how are you defining your groups I want to do the grouping variable first I'm going to be defining my groups by people who love or hate golf and sports so I'm going to scroll down to find that option here's my golf and sports option I'm going to put it into my grouping variable area but I can't stop there I have question marks here and that's because I have to click this next button that says how am i defining these groups all right well let's see I won't group one to be everybody who does not like the sports I'll put a 1 in there and I want group 2 to be everybody who loves the sports and now I'm going to continue and I want to now use my test variable which is the number of Facebook friends we've got so let's scroll around to find Facebook friends there it is that's the variable on comparing so I have the lovers of golf and sports and the haters of golf and sports and I want to compare their Facebook friends to see if there's a significant difference that's all I have to tell SPSS if I were to click the options button it would ask me what is my confidence interval the 95 percent tells me that my alpha value is 0.05 and that's what I want my alpha value to be for this test so I'm going to leave it as it is and click continue then I'm going to click OK now SPSS is going to run our test and it's going to give us two large boxes of values the first one just shows us the mean of both groups look at the mean number of Facebook friends in our people who don't like off in sports and now look at the mean number of Facebook friends on people who do like off in sports these numbers look pretty far apart so my guess is going to be that there is going to be a significant difference here and that my t-test is going to allow me to reject the null and go with my research hypothesis this is the result of my t-test it gives me an F test first and it's p-value so I can determine whether or not to assume that my variances are equal once I do that and I know which row I'm using I can determine what my t-test value is and what my p-value of the t-test value is and from there I can make my final conclusion so let's take these two tables and look at them more closely in my PowerPoint slides okay so we were talking about loving and hating golf and our two groups we had our null and alternative hypotheses and we ran SPSS our very first step is to determine which one of these rows the top row or the bottom row are we going to use to determine the final results or conclusions of our test and in order to do that we're going to look in this big pink circle here SPSS runs an F test with a p-value for that F test and it tells us whether or not we can assume that our two groups have equal variance or not that's what this is all about here's the quick roll if our significance or p-value for this F test is less than our alpha value or 0.05 which it is in this case our variance ours are not assumed to be equal if this value is bigger than 0.05 then we can assume them to be equal but in this case our p value is less than 0.05 and so the variances are not assumed to be equal meaning our two groups are not necessarily assumed to be from the same population because we're assuming that the variances are not necessarily equal that means we want to focus our attention to the bottom row of the data so that's what's circled here in this red box okay next given that we're going to use the bottom row and we know that alpha is 0.05 for our test we can see here that our T value is minus two point eight eight three and the p-value for that T test is point zero zero six anytime our p value is less than 0.05 which it is in this case we can reject the null we can conclude that our T test value has fallen into the rejection regions and finally based on this because we can reject the null we can conclude that there is significant evidence to support our research proposal that loving or hating golfer sports actually affect the number of Facebook friends that a person has and so here we can actually see this people who didn't really like sports did have a much lower average number of Facebook friends and so we were able to use a t-test for independent means to show that our research hypothesis was actually significant in this case and that there is a significant difference between the number of Facebook friends people have depending on whether they love or hate golf and sports so this was an example of an independent t-test using SPSS please join me again for future examples thanks

Channel: ProfessorAmiGates

Views: 197,234

Rating: 4.8971429 out of 5

Keywords: hypothesis, testing, t-test, means, sample, SPSS, hypothesis testing, comparing means

Id: J0nPaZ2z1VE

Length: 11min 25sec (685 seconds)

Published: Wed May 23 2012