SPSS - Reliability Analysis - Cronbachs Alpha - Combining groups of variables into one | Ton Van Voorden | Skillshare

SPSS - Reliability Analysis - Cronbachs Alpha - Combining groups of variables into one

Ton Van Voorden, MSc.

Play Speed
  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x
2 Lessons (19m)
    • 1. 12

      3:39
    • 2. 12

      15:01

About This Class

Sometimes the answers to specific questions are vague or subjective to interpreter. A question like Are you Happy? Could be answered with exactly the same words, but mean something completely different. In order to be an objective researcher we make use of more questions to make the happiness more objective :)

Transcripts

1. 12: Hello. This is stolen Florida, Your statistics teacher. In this video, I'm going to tell you something about scales and about comas. Often most important thing about this video is how do you go into create a number of questions To find a concept? Sounds a bit difficult, but a question like, How are you doing? If my kids Hey, how are you doing? It looks quite okay. So he would be a human. You would say good, Good. But it would be different if someone would say good for someone who and because of all this differences. And if you want him down as good all that he become exactly the same. You don't know what is what anymore. So you're a picture. It's not my cat. How are you doing? So what we need to do is first want to know what we want to know. For example, Question like, Are you happy? It's quite a vague question. If you ask people that will say yes. I'm happy. Yes, even if they're generally not so happy. So it's a bad question toe. Ask people directly if they're happy, but sometimes for questions. If you do want to do your statistic work. Who really want to know if people are happy or not? In order to do that, you have to create a number of questions. He created six questions people with people generally have to answer with. Hungry slightly every neutral, slightly disagree and disagree. And here for Are you happier? Created the questions people will do. The happy dance are funny. People are not happy in general, they hate people. Happy dances. I generally wake up with its mouth and full of energy. I love to endure little things in life. I love going out and be social. I'm easily annoyed by people on goldfish. Is should have more space in that bowl. For the last question is a bit last two A bit questionable if you agree on them, the banks to happiness. But I'm going to show you in the next video what to do about it and how to check it within SPS. If if they would be perfect, you cannot learn from little mistakes. So let's step one created scale. You can also create find a lot off already tested skills on Internet. I would really recommend you doing that if you're wining your teaches or your own work. Second step, I'm gonna teach you to check the scale. You do this, uh, where the reliability analysis within SPS and you get output, which is called Come back often some more things. Last step if you get your skills. All right. You're good to go to take the mean or the same off the scale. So just if people score like a 543 and three times before if they mean the skill mean becomes four and people get hurt for as a school for happiness, I'm failed to go into the next video. I'm going to show you how to do all the steps within SBS. Thank you for watching this video. 2. 12: Hello. Doesn't storm of order your statistics teacher in this video I'm going to show you how to do a reliability analysis with comas. Offer within f business. I hope you're ready. Wants my previous video. We might get in What I explained to you the idea behind reliability analysis. Let's go forward. So over here you will see SPS s And what you see over here is I created the happiness skill to find how happy people are. Really? If you ask them. How happy are you with our happy are really so I created six questions people will do the happy dance are funny. I generally wake up with the smell and full of energy. I loved him and your little things in life. I love going out and the social and easily annoyed about evil and goldfish is should have more space in that both First of all, if you look good to this questions, maybe you really like what does a cold fish have to do with your happiness? Actually, therefore, edit is very cool. Probably the goldfish does not have anything to do with happiness. So it should be removed at the end next to it. If you look to the question, I'm easily annoyed by people. It's like opposite, Lee asked, than the four questions above it. Because if you completely agree, like I love going out and be social and probably and not easily annoyed by people so and he would expect a low score goldfish, it does not matter at all. Let's put this all within a reliability analysis and let's see if SPS gives you the same answers as we came before. Normally, first step is always to check these things before yourself and already do the steps we're gonna do later, like remove terrible's or like Reese killed in. But let Jack what s being said will say if he would have forgotten it. Click on, analyze, take a scale reliability analysis. We good little menu over. You See that items just put all the six items within, hear a click on statistics. Wouldn't the statistics been though I mark the scale if I've been deleted? I put it on press continue and press OK. What you will see over here is firstly, is the Commerce Office score 0.387 Commerce office score goes for minus 1 to 1 in which only one is good, not with correlation. Minus one is good, only one is good and you want the Commerce Office score and generally to be above 0.7 for good. It created good score, so the 0.7 is important to remember. So let's see. This value over here should be are both who Wait. This is a different one. My excuse Skopje. This value over here should be above 0.7 and it's not hear the case. But let's look further to the item Total statistics over a year you see a girl and in this go you have to look to first of all negative values. Why do you look to negative failures? General, they need to be re skilled and what you see over you is I'm easily annoyed that people has a really big negative value on Therefore it should be re skilled. This very is also negative. Was first start with really big negative one. Then we will look for so first that look at the corrected. I didn't tell the correlation. I'm easily annoyed that people, How does the read skill work like in my happiness scale? If you look here. One is completely disagree on five. It's completely agree. So this is disagreed and do this. I agree. So this means happy. And there's a great means not happy in general except this variable and easily annoyed people. Because if you agree to it, you are not happy. If you don't agree to it, you're happy, so I can create in 54321 The skill works on this way. And actually, this course should become the fashion. Become a one before should becoming 23 Phase 24 and that one should be comma five. How to do it is the easiest way will be with a little mathematical calculation. You pick the highest number of five you do plus one. So you get six. 12345 six If you got a seven point scale which also often used 1234567 eight So you get eight and now it's quite easy. New score will be six minus the number here because six minus five equals 16 minus four equals 26 minus tree six minus one equals five. That's bullets within SBS s transform Compute variable. I say With the right mouse, I can display the variable names sometimes easier. I see the happiness kill annoying and put wedding numeric expression six minus. That annoys skill. This part a copy. I base it over here and I put it for re skill behind. So this means the target fable re skilled will be six minus the old skill and have a best Okay, Now you see, just compute. I cannot see it because it's too low. Only thing you see will be compute eyes. Well, nothing is. He seems like nothing happens. But if you look in your data set, you will see Over here your new eyes annoy. And if I put the variables next to its order I see one becomes five failed for 251 for two . So adverts always jacket Afterwards they create a new variable right mouse instead available. Double click on the name Yeah, I double click again on the name within the variable view I call it just excess And now I will be putting the old variable annoy skill behind the excess. Why do I do this? So that skill stays together and I don't need to delete the old one. If you forget to do this, there's nothing warm. But it makes life easier. Because now, if I want to do that reliability analysis again. I moved my old valuables, and now it just grab display variable names all the variables before excess and I can not by accident include the highs and no it again put him in here. In this statistic, there still the scale of items related press. Okay. Again, I'm gonna check it again. First of all, you see that call most often is no higher. Seven higher as 70.7, which is great next to it. We're going to check the corrected item. Told the correlation Woods will be, um, or positive values except off the goldfish. Should have more space in that bowl. But next of it, yeah, this value. And especially because goldfish should have more space in that boy. It doesn't makes sense. And this value is like lower as minus 0.3. I will not go. I'm not going to re scale it. Of course, if you want, you can first re skillets also. But for now, I'm going to continue to the next step, which will be checked the coma offer if I didn't delete it. And in here you check the commerce Alphonse, and you compare it with old school myself. So 0.867 and what you will see is it's better to keep. The people will do the happy dance are funny within your skill because it keeps the skill high saying for this one's for all except off the goldfish is shoot have more space in their bowl. If you remove this item, you will see the commas offer will go up even more 0.2 point 957 So always check for yourself like shouldn't be included or not. If you have ah, already created skill and you call must offer is already higher as 0.7. Just leave it there, especially if the difference between the two numbers of small like 20.1 points to something like that. But here the difference is almost 0.1 higher and next to it. If I look to this question, the goldfish would have more space in the bowl has nothing to do with your happiness. So therefore, I decides to delete this question and do that comp analysis again so what I will do here in our data, said. I also put this again behind the excess. I press, analyze skill, reliability, analysis I remove to fish press OK and checking the data over here. He will just see that the commerce offer is now really high. Born 957 In this column There's nothing negative. And in this column all the values are lower, which is great, too. So now we're ready to create a skill. Last step. When you're finished with the video, I click on Tense for computer available. I say Over here, Daggett variable equals happiness. Gil equals the meat. You really have to die there. Sorry. You can even put like points three or point for behind. It means like you need at least three questions off four questions in order to make a new skill with the every if everything, it doesn't matter so much mean off, and I really have to double click on all the variables and put dots in between. It's a bit of a work, especially if you have a lot of variables, but just make it like this. I mean, bracket open puts variables in it dot variable dot fable Don't family and close having this best. Okay? And now the new available. Well, we created over here If instead you want to some just put instead of me some James existing I would create happiness some. Okay. And now Oh, so we'll see the happiness some if you create a some verbal really, really, really important that from all variables you have all the data order. But if you miss a day that somewhere you're some for that variable will be lower and Robien era. So what we've done in this video a lot off steps. We went into the analyze scale reliability menu. Put everything inside there. Don't forget. Don't forget to click the statistics skill of item The leader on last but not least, we we checked how this stables work and corrected. I don't told the correlation if they're heavily negative, re skilled, available and comas often item deleted. If this one is higher than the value over here, you have to delete variable, but always check it. Check it like if it's logical to do it. And that's old. Thank you so much for watching this video. Looking forward to seeing you in the next video