SPSS  Comparing Two Means  Do girls score higher than boys?
Ton Van Voorden, MSc.
2 Lessons (18m)


1. 12
7:24 
2. 12
10:54

About This Class
Wooooww! Already getting to comparing two means. Good job!
One movie about the dependent Ttest and one for the independent ttest.
The dependent one you use when the scores are dependent of each other, Like Me receiving First a 8 and then for a second test a 9. Compare that to independent scores of a random boy scoring an 8 and a random girl a 9.
Class Projects See All
Ensure that you understand the difference between Dependent (Paired) and Independent completely. If you have any questions let me know!
(Read More)
Transcripts
1. 12: Hello. This is storm of Order, your statistics teacher and this video. I'm going to explain to you something about a deep end and sample t test. More specifically, what we're gonna test is if the efforts of a great off a second test will be higher than the efforts off the day off. The first test on everyone made the first test and the second test. So let's see valuable duties within SPS over here. Rico, If you're like doing a zero, it's one hyper teases the efforts off the first equals efforts off the second, they will get a deep end and there will be the depend detest whatever I said as possessed there simple teeth. So now we go to be a guest. And over here I have my column off first grade and over. You have my column off second grade. I'm on the comm parodies. So person one. Go for his first examine. What aid? And for a second attend. So he scored to hire person to 1st 6 seconds. Eight person four first, seven second one goes like all the way around. Let's see if the first grade will be higher as the second otherwise said wouldn't down like a hypothesis test The evidence off the second grade will be higher than the efforts off the first grade and the idea will be smaller equal. Then the first grade going to SB. That's we click on analyzed. We want to compare means and here we go to this back sample T dest You get like this little menu pop up over here. You can put in a great first to put this terrible one second. That's favorable to Don't put anything into pair, too. If you want, you can click on options and set different confidence. Interval present it now we will not do it. Click on OK And over here you will see the output. So what you see over here is the mean. I have a copy the output into the other program. It's not working perfectly. Excuse for it. I make two copies. No. Perfect. So what you see over here is the mean off. The first grade will be 6.15 and of the second grade, 6.4636 Furthermore, it also has the on standard deviations. The first group is a bit smaller than the second. But our hypothesis waas new seconds is higher as new one. And in terms off x bar, this is too. So that's important. Secondly, you see some paired samples. Correlations. It's like saying What is correlation between second and the first group? I never use it. Those are not be significant. More important, over here you will see the end results and you will see the I mean, the mean over here is the mean difference. It was the eggs difference. So what is it? Actually, it's like the same as efforts to minus mm Everts won, minus Evert to standard deviation will also be the standard deviation off the difference. Stan. Deviation off the difference in that era off the mean will be de bar of years. He had confidence interval off the difference. These two numbers. More important, he would get to the end results T equals minus de together with 219 degrees of freedom equals minus 1.831 This is the way you have to write it down in the if you white a description like your teases or at the end off some subjects and what you see over here's a 62 tails with equals 0.68 from both sides. There's like a P value off half of it, 0.0 feet for and together there is 0.68 So what's the people in our test? Our test is not two tails out. Test is wanted if it's bigger than a new one. And, uh, devalue is negative and that's because D is eggs one minus X part two. So you can rewind this high petit GIs as you want it smaller as you too. It's of course, also to so the rejection region. This case will be on the left side because you want it smaller. As Mewtwo on the left side is the way SPS s Berg's its stakes of the one on the left side T for you is negative will be over here. So you get like this part is the P values 0.0 c four, of course, given an often off the open 05 be smaller offer So the T value here fall into the projection region, so the crew conclusion will be rejected 80 So pay attention to this sick level if you have , Ah, one tail test and you have to divide it by two. Thank you so much for watching this video.
2. 12: Hello. This is Stone of Order, your statistics teacher and this video. I'm going to show you that independent two sample t test in SPS s. Moreover, what everyone wants to meet nice people in their lives. But when with days in the week are you meeting? There's nice people, Is it especially in the weekend was, especially during the week when you're at your job. That's what we're gonna explore in this question. So what we've got over here is this hyper Jesus Evert's on Efforts on weekdays are different as the efforts off nice people you meet during the weekend, so the is there will be the opposite ever. It's nice people you meet on a weekday Monday to Friday at the same during the weekend. We go into the same SB says data set on over here. You see the day off the week monitor fighting or weekends, and over here in this column, you will see how many nice persons you've met during that specific day. It's different person. Some people were asked on a month later or Friday, or the people were asked on the weekend. When she overhears this little button. It's a 81 If you click it, you will see that day off the big chains and once and zeros so it s possess the days of the week are not explained in I'm not explains a numbers off like in inverts. But in numbers like, if you're clicking on favorable view, you can click on the values off the day. And if you click on this little button, you can see there are means we get one means monitor Friday. If you get a data set where this is not done before, always do it because it makes your output easier. Toe read Going back to the day that click on Analyze we want to compare means. Um, but in member remember the last video. Compared to columns, everyone has a 1st 1st grade. Everyone is a second grade. He everyone. It just asked him one day. So you have, like a group of people were asked on the on the moment of Friday and a group with which is asked on the weekend days. So there's no direct connection between them. Therefore, we have a compare means not a bad sample teachers but independent simple t exist one second to look over here like dependent is like this and independent There's a group invariable and within a grouping variable there like difference. Um, numbers going back to the SPS s example. Click on analyze Compare means in the independent simple T fist over here you put in the test variable. What do you want to test which, ever, Just where we met a nice person like click on this little button. And there's the grouping available with variable defines and with groups that amount of nice people met belong. Uh, this would be the day, you see, like two little question marks over here you click on define groups and he have to say that Group one is they all group to is one because we defined a zero and once as we days on weekend days like monitor Friday and weekend days Let's continue over here. If you click options you can put like in which kind of confidence interval You want just 95% like always click on OK and I get my output. I copied it out put already for you into SPS s or into my little one note program. Let's interpret this output. You see the day off the week met the nice person. On the weekend, we get 73 observations and Jonah, we get 147 more during the week. Furthermore, the mean in the weekends 3.85 on the mean for monitor fighter will be 2.90. So X bar weekend equals 3.85 with standard deviation off 2.4 98 And the egg spot Monday till Friday equals 2.9 c with a standard deviation off 1.84 Question is, of course, there's a difference between those two. But is it because of coincidence or is a statistically significant difference? Remember what we need to do from the former video we got like this hypothesis In this case it will be independent. So first we have to do it f test to check whether the variants are equal. Variants are not equal, like explained you earlier in an earlier video. But most important for now, you don't really need to go back it just decide what you have to go to the equal variant supports or not equal Vince apart. SB assess directly gives you all the relevant output and gives you a little F test over here. So what does this mean? It zeal AIDS. One variances different. It's the offense is equal. Only thing you really need to know is that f value equals 15.509 is a live in test. You don't really do anything with value with what? What you really do is like the sick is the P value B equals It's smaller than you have to write down its smaller than because it zero points it was a Brazil, then you white always own smaller 0.1 of course, 0.1 smaller as the often 0.5 So you rejected 80 hypo teases. And because of that, um, you get into the equal variance not assumes method like in s Digest is upbeat. Looks a bit rumbly, But what's especially important if you pick the under column like the above you just remove it. You don't look to all these numbers. So, like s possess will give you methods going back to here as business will give you methods sue to equal van Suppose a medicine for not equal so here. We met too. He will be met a free. Now we use metal three because we rejected zero Unequal variances oppose equal Vince. No Assumed so Remove the above line. Otherwise, this very had be higher. At 0.5 they would every move to be low land getting into the results. Actual results like from this one. What you will see over here is it's a T test with 110 point for 80 The gates of freedom. It's a bit strange. It's like the first time in degrees of freedom is between one and 10 and 111. Like if you would use a table, you could never find this kind of failure within a table. But within a program like Excel, you can find it. Um little dot equals. So 110 dealt for 18 equals 2.796 This value and the most important is the significance. It's detailed. Our test is also two tails, so we can definitely use it. So B equals 0.6 0.6 smaller than often smaller A 0.5 So we reject 80 and it one is to So the ever It's our moderato. Fire is not the same as that was on the weekend within the population. What you see over here, you see, like the mean difference. What is the way already done the test and just explain the other things within SPS s which could use you don't really need them Normally is the mean difference was the mean difference. It is the difference between these two numbers 30.85 minus 2.93 Those have a standards. Our off the difference equals as X by one minus export to and the conference involved. The difference will be in between. And what you see over here is in between 0.267 and 1.567 They're both above zero. So because of that, we can also conclude Reject it. Zero. Thank you for watching this video.