If you are running a correlation between two variables with unequal samples sizes, SPSS will simply run the analysis on the data that are available for both variables. So your sample size will equal the smaller N associated with the two variables.
Can you please explain what one should interpret from the value of significance in the output table? Also, should we only look at the Pearson correlation value in the output table to interpret a relationship? I am new at this so please help me.
@@TheSamyben It's the significance level, used to determine whether a significant correlation exists between the variables evaluated. If p-value(sig. as shown in the table) > the significance level (0.01) then the null hypothesis is not rejected and hence there is no significant correlation between age and income.
Not sure if this is the right place to ask, but im going to ask anyway. How do i set up a correlation between many variables up towards a single variable? As in checking the correlation of for example "age" up against "weight", "drinking", "smoking", "number of kids", and so forth? So that i dont get that giant 6x6 diagram but rather a 1x6. Thanks for all the videos :)
I'm a little rusty (I'm reviewing material I learned several years back) but, since you predicted that older people would have more income, shouldn't we have used a 1-tail test here? Input please, if I'm wrong please explain to me why I am.
i want to make correlation between dependent and independent variable,but in my dependent variable it has 4 question in the questionnaire,how can i combine all of this 4 question to get 1 value same goes as independent variable and with that value i use it for correlation?
+anna melissa I think you should combine them by clicking on Transform --> Compute variables, then you add all your variables, name the new variable and click ok. It should look like this "q1 + q2 + q3 + q4".
Hi, if I want to correlate 5-point Likert scale items how would this carried out? For example, trying to correlate a 5-point scale measuring motivation with a 5-point scale about emotions. Would I have to find the total mean or median score of the question items first before correlating? I am so confused, it would be so helpful for an answer please
Technically, the best way to estimate the association between two ordinal scale items (e.g., 5-point) is to use a polychoric correlation. I show how to do that here: ua-cam.com/video/9aJu5MIWCBU/v-deo.html However, as I write in my (free) textbook (www.how2statsbook.com) Chapter 5, I argue (with evidence) that a Pearson correlation can often estimate the association between two 5-point ordinal scaled items. That is, the Pearson correlation and the polychoric correlation will estimate very similar correlations. Once you have 4-point ordinal scale or lower, though, the correlations tend to diverge.
Cian O'Carroll one tailed means when you know what the result is likely to be e.g age and money - the older you are the more money you have whereas two tailed is when you are unsure of what end result.
A correlation coefficient value of .000 means there is no linear association between the two variables, there could be other complex association. And, probably, correlation is a bivariate method, not an univariate method.
No the Pearson Correlation tells you about if a value is a positive or negative correlation based on the proximity of its relation to 0.05 in your case at 95% confidence, the sig is the P value and therefore tells you if something is actually correlated. In this case it is far under 0.05 so you accept the alternative hypothesis of 'there is direct correlation', I think.
No, there is a relationship between them bcuz correlation coefficient is lower than two tailed (0.01) while correlation coefficient is (0.000) understand?
When you are carrying out pearson correlation on factors, do we consider the mean of depends variables of the factors or the sum of dependant variables?
I am doing an assignment that is asking for the coefficient, p value and the significance from a correlation data table based on my own data in spss. What do I need to put in the coefficient column of my assignment? I know the p values and significance from this video and other references already.
So, can we say, age in years and household income have a relation? Bcuz we used two tailed (0.01) and sig no. Is lower than o.o1. Is there a positive relationship between them?
Hi there hope you can help me. I am doing a study with one measure of the independent variable (from a 7point likert scale) and the dependent variable (scores on the dass21) have been taken 2 or 3 times over 2 years. Should i just do a pearsons for each time. then average them out? Or is there another procedure i should do? Would be good to see diferences in time 1, 2 and 3 Or do I have to calculate them all at once for a single correlation? Thanks very much!
@how2stats I was wondering whether there was any way to compare 2 models. Lets say we were interested in finding out if a participant is habituating to our dependent variable. In order to see whether we can generate a model to account for this, each participant will have to have their own model. Is there a way that we can compare the models of individual participants? For Example: - We play a participant a loud tone and record blink amplitude. As the participant is played more sounds, they become used to the sound and as a result, begin to habituate. We can generate a model to see if there is a relationship between the number of startles and their response. At this point, we have 1 model for each participant. Say we have 30 participants. Is there a way to now compare these models to see if they are statistically similar?
Look for Spearman correlation, if the categories are ordinal in nature; ua-cam.com/video/r_WQe2c-ISU/v-deo.html Look for phi correlation if dichotomous in nature (which is really just a Pearson correlation).
hello. i'm currently conducting my studies on assessing the writing proficiency of my students. i already have my rubrics to check it and i was able to rate it already (numerically like 100, 99, etc.) how can i quantify the ratings from 1-5? 5 is excellent. thank you.
Higher scores on X are associated with lower values in Y. A negative correlation example I describe in my textbook is empathy and sales performance. Higher levels of empathy are associated with lower levels of sales performance.
Hi, Can you please tell how I can show that the diameter of a vein at specific level have higher correlation with varix occurence, Than at lower dilated levels. Like when its 10mm the correlation is less. how ever percentage is higher when the diameter is higher. Thanks in advance for help
hi, THANK YOU very much for the psoting, i really like your video, but i was wondering if we could pearsons correlations if the sampel sizes are not equal? big thanks in advacne :)
hi few days back I asked how to form a norm group using spss. I have data of personality test of 200 sample and 35 variable. now I have to make a norm group table , how do I do that?
hi....... i am tarun solanki doing MBA. and i m doing research right now and i am confused that i have used one model in my reseach and there are six component of my model to measure each of them i have made several likert scale statement it is 34 statement now i want to do corelation btwn those 6 component how do i go about it?
i am not sure what i did was 100% correct or not,what i did was that i took the mean value(as the mean is best estimator) of the statement and used multiple correlation.
+hdth07 The asterisks just signify that the Pearson correlation is significant. It's not the one you compare with the alpha value 0.05. The sig. (2-tailed) is what you look at and compare with 0.05. :)
It would depend. For two dichotomously scored variables, sure, you can use Phi (which is really just a Pearson correlation). For categorical data with more than two levels, it's more complicated. It would depend on the nature of the data: ordinal or nominal. I should make a video on this topic.
I'm not sure what you mean by 'norm group table'. I have a video on how to create percentile scores. Maybe that will help you out. Search 'percentile scores' in my channel and it should come up.
![Pearson correlation [Simply explained]](http://i.ytimg.com/vi/k7IctLRiZmo/mqdefault.jpg)
If you are running a correlation between two variables with unequal samples sizes, SPSS will simply run the analysis on the data that are available for both variables. So your sample size will equal the smaller N associated with the two variables.
Great video. What I did not get here is where it explains whether the correlation is negative or positive.
If you get a p= or
you explain all your videos very well. Thank you so much
Thank you very much for all your videos! They truly help with clarifying!
Can you please explain what one should interpret from the value of significance in the output table? Also, should we only look at the Pearson correlation value in the output table to interpret a relationship? I am new at this so please help me.
I need to know as well
hopefully, you will include next time on how to write interpretations.
Thank you for making my assignment a whole lot easier (:
In my analysis i dont get the bottom text that says the level of significant correlation....?
same, help pls
@@TheSamyben It's the significance level, used to determine whether a significant correlation exists between the variables evaluated. If p-value(sig. as shown in the table) > the significance level (0.01) then the null hypothesis is not rejected and hence there is no significant correlation between age and income.
same here, i need help pls
thanks stats dude you're the best
Can this be run on three variables?
Not sure if this is the right place to ask, but im going to ask anyway. How do i set up a correlation between many variables up towards a single variable? As in checking the correlation of for example "age" up against "weight", "drinking", "smoking", "number of kids", and so forth? So that i dont get that giant 6x6 diagram but rather a 1x6. Thanks for all the videos :)
VERY EDUCATIVE. I am fully helped by your tutorial . thanks very much
How did I do correlation if it is a variable that has groups like tendon pain and no pain
I'm a little rusty (I'm reviewing material I learned several years back) but, since you predicted that older people would have more income, shouldn't we have used a 1-tail test here? Input please, if I'm wrong please explain to me why I am.
Haaah, thank you thank you thank you!!! You just saved my life! :D
Am confused in variable picking method we will choose variable names or items
Very concise info, easy to follow demo. Thanks!!
Very informative video, thank you very much!
At 1:58 would it not be correct to tick the "one tailed" Box, because you have an hypothesis? You predict that there will be a positive correlation.
i want to make correlation between dependent and independent variable,but in my dependent variable it has 4 question in the questionnaire,how can i combine all of this 4 question to get 1 value same goes as independent variable and with that value i use it for correlation?
+anna melissa I think you should combine them by clicking on Transform --> Compute variables, then you add all your variables, name the new variable and click ok. It should look like this "q1 + q2 + q3 + q4".
I have the same problem, can u pls give some tips? thanks
Use the Transform --> Compute... utility in SPSS. I'll have to make a video on that.
Did you ever get round to making that video?
Can you please share the link on how to do this?
Hi, if I want to correlate 5-point Likert scale items how would this carried out? For example, trying to correlate a 5-point scale measuring motivation with a 5-point scale about emotions. Would I have to find the total mean or median score of the question items first before correlating? I am so confused, it would be so helpful for an answer please
Technically, the best way to estimate the association between two ordinal scale items (e.g., 5-point) is to use a polychoric correlation. I show how to do that here: ua-cam.com/video/9aJu5MIWCBU/v-deo.html
However, as I write in my (free) textbook (www.how2statsbook.com) Chapter 5, I argue (with evidence) that a Pearson correlation can often estimate the association between two 5-point ordinal scaled items. That is, the Pearson correlation and the polychoric correlation will estimate very similar correlations. Once you have 4-point ordinal scale or lower, though, the correlations tend to diverge.
@@how2stats thank you very much . How about more than 2 5-point ordinal scale item & instead several ?
How do I create the variables though? How do you group all the questions say about income into one variiable before I perform the correlation?
I have input the variable for Pearson correlation and press ok button but don't show any result, please response about this problem
How do I know when to use two-tailed or one-tailed test of significance?
Cian O'Carroll one tailed means when you know what the result is likely to be e.g age and money - the older you are the more money you have whereas two tailed is when you are unsure of what end result.
@@xFinalxArenox Thanks for responding to that other person :) that is what I came here for too, but it wasn't explained in the video.
How do we know the degree of freedom in pearson correlation
thank you for saving my dissertation
Plz save mine as well if you can
A correlation coefficient value of .000 means there is no linear association between the two variables, there could be other complex association. And, probably, correlation is a bivariate method, not an univariate method.
thank you, it is really helpful
you have done a great job.
I didn't get it. Was your hypothesis right or wrong? Do people make more money when they get older?
if p= or
No the Pearson Correlation tells you about if a value is a positive or negative correlation based on the proximity of its relation to 0.05 in your case at 95% confidence, the sig is the P value and therefore tells you if something is actually correlated. In this case it is far under 0.05 so you accept the alternative hypothesis of 'there is direct correlation', I think.
If I have 5 dv and 5 iv how do I calculate the correlation
a correlation coefficient value of .000 indicates no association between two variables.
No, there is a relationship between them bcuz correlation coefficient is lower than two tailed (0.01) while correlation coefficient is (0.000) understand?
Hello, Performing correlation after transforming the data to log to the base 10 is that a correct method?
Sir do i put variables in percentage or in numbers ??
When you are carrying out pearson correlation on factors, do we consider the mean of depends variables of the factors or the sum of dependant variables?
What if there's a negative in front
hey so I did this and my significance is smaller than 0.05 but larger than 0.01, can I still say it is significant?
If you're alpha aka confidence interval is 95%, which is the standard then your finding is significant. So, yes it is significant.
I am doing an assignment that is asking for the coefficient, p value and the significance from a correlation data table based on my own data in spss. What do I need to put in the coefficient column of my assignment? I know the p values and significance from this video and other references already.
So, can we say, age in years and household income have a relation? Bcuz we used two tailed (0.01) and sig no. Is lower than o.o1.
Is there a positive relationship between them?
Yes
Hi there hope you can help me. I am doing a study with one measure of the independent variable (from a 7point likert scale) and the dependent variable (scores on the dass21) have been taken 2 or 3 times over 2 years. Should i just do a pearsons for each time. then average them out? Or is there another procedure i should do? Would be good to see diferences in time 1, 2 and 3 Or do I have to calculate them all at once for a single correlation? Thanks very much!
@how2stats I was wondering whether there was any way to compare 2 models. Lets say we were interested in finding out if a participant is habituating to our dependent variable. In order to see whether we can generate a model to account for this, each participant will have to have their own model. Is there a way that we can compare the models of individual participants?
For Example:
- We play a participant a loud tone and record blink amplitude. As the participant is played more sounds, they become used to the sound and as a result, begin to habituate. We can generate a model to see if there is a relationship between the number of startles and their response. At this point, we have 1 model for each participant. Say we have 30 participants. Is there a way to now compare these models to see if they are statistically similar?
what if you don't have the age but just participant numbers do we include that instead of age/gender?
very well , and easily described
Do we have any correlation between 2 categorical variables? If we don't, tell me please, how do we know the relationship between them, thanks!
Look for Spearman correlation, if the categories are ordinal in nature; ua-cam.com/video/r_WQe2c-ISU/v-deo.html
Look for phi correlation if dichotomous in nature (which is really just a Pearson correlation).
thank you a lot you are the Man!!
Is it possible to analyze correlation of ranked data ( Google survey) questions
Yes, use the Spearman rank correlation: ua-cam.com/video/eE6GkdpSnK0/v-deo.html
hello. i'm currently conducting my studies on assessing the writing proficiency of my students. i already have my rubrics to check it and i was able to rate it already (numerically like 100, 99, etc.) how can i quantify the ratings from 1-5? 5 is excellent. thank you.
Sir please explain about sample size calculation using spss for quantitative study
thanks, excellent presentation and very helpful.
So pearson correlation is not relevant to comparing two ordinal variables?
How would you understand if it was a minus number
Higher scores on X are associated with lower values in Y. A negative correlation example I describe in my textbook is empathy and sales performance. Higher levels of empathy are associated with lower levels of sales performance.
can i use four variable for Pearson correlation coefficient in SPSS ?
That’d be an ANOVA I’m pretty sure
I have a question regarding to ranked data
Hi, Can you please tell how I can show that the diameter of a vein at specific level have higher correlation with varix occurence, Than at lower dilated levels. Like when its 10mm the correlation is less. how ever percentage is higher when the diameter is higher. Thanks in advance for help
hi, THANK YOU very much for the psoting, i really like your video, but i was wondering if we could pearsons correlations if the sampel sizes are not equal? big thanks in advacne :)
Thankyou! You were a great help
hi
few days back I asked how to form a norm group using spss. I have data of personality test of 200 sample and 35 variable. now I have to make a norm group table , how do I do that?
Hello, would the number of participants be 850 since n is equal to 850?
Yes
So it's positive?
I'm not getting the "Correlation is significant at the 0.01 level (2-tailed) at the bottom
Trey Miskoffkitch check the "flag significance correlations" in the bivariate correlations screen
is the pearson correlation the same as the r value?
Yes.
how2stats thank you!
hi....... i am tarun solanki doing MBA. and i m doing research right now and i am confused that i have used one model in my reseach and there are six component of my model to measure each of them i have made several likert scale statement it is 34 statement now i want to do corelation btwn those 6 component how do i go about it?
hi, were you able to solve your problem? i have the same problem too,
i am not sure what i did was 100% correct or not,what i did was that i took the mean value(as the mean is best estimator) of the statement and used multiple correlation.
thank you Solanki Tarun
how do you explain the relationship between the independent and dependent variables?
The explanation was not clear for me.
hlw,pls tell me how can i calculate overall value of 3 column vs overall value of 5 column correlation
Why has it got 2 asterisks when it is not significant i.e >0.05? I don't understand.
+hdth07 The asterisks just signify that the Pearson correlation is significant. It's not the one you compare with the alpha value 0.05. The sig. (2-tailed) is what you look at and compare with 0.05. :)
great video! thanks
What do I do if my variables are inappropriate for this dialog box? Cant perform correlation do you know why?
and what if it shows "correlation is significant at the 0.05 level (2-tailed)" ? what does it mean?
He's really the best
what about if it above .000?
Well explain.. Thank you
Thank you - so helpful!
Please let me know how you proved that high age group have high income.
what make the data is not correlate? i do conduct reliability n validity and it is appropriate figure but it is not correlate.
Wonderful, thank you!
Please give me data for correlation analysis
Super helpful. This video helped me to find my way. I definitely added a like to your many likes how2stats.
Thank you very helpful
Link to the data set please
Hi. Do you use a correlation with categorical variables? Could you please explain more detail?
It would depend. For two dichotomously scored variables, sure, you can use Phi (which is really just a Pearson correlation). For categorical data with more than two levels, it's more complicated. It would depend on the nature of the data: ordinal or nominal. I should make a video on this topic.
I'm not sure what you mean by 'norm group table'. I have a video on how to create percentile scores. Maybe that will help you out. Search 'percentile scores' in my channel and it should come up.
You sound like a young Ross Geller.
Sure, if you subscribe to the one-tailed hypothesis testing point of view. Not everyone does.
Explained well
Dude thank for making this, but please drink some water or something before recording. That lip chomping is kind of hard to listen to
That's why I'm watching with subtitles 8)
many tnx sir. Very helpful,
you did nt tell what is the meaning of .000
is there a correlation???? or No correlation!!!!!!!!.......wht if we get .478 as sig 2 tailed.
My variabes (climate variables) are in asc raster files. So to do this analysis, should i convert these rasters to point data ?
PEARSON'S R CORRELATION
ANALYZE
CORELATE
BIVARIATE
TWO TAILED
easy explanation
Thanks
Haaha that was a joke with students
This is not univariate analysis.
gay