Correlation analysis using Excel

How to run a correlation analysis using Excel and write up the findings for a report

[@Leon-om1lg]

This video saved my life. I thought for sure that I won't complete my thesis. But because of this, I finished it! If it weren't for this vid, I wouldn't be able to have my thesis defense. Thank you so much!!!!!!!!

[@ChristianHigherEd]

I'm very glad to hear this. Best wishes.

[@thabethmatoda913]

Me too i was actually crying 🤣 😂🤣 😂

[@thabethmatoda913]

I didnt know what to do...I just started crying because had no hope.Had only 3 days left before i submit my thesis.

[@ijustfollowthesun]

Thanks so much. You're still helping people 10 years later. THANK YOU!!!!!

[@aldehoyos6803]

Finally!!! Someone who can explain correlation without going all over the place!!!! thank you!

[@PAYLAYP]

This video has just saved my ass more than you know in the dying days of writing my dissertation... thank you so much

[@ChristianHigherEd]

Glad to hear it. Hang in there and congrats.

[@angelikakachnowicz6759]

I don't know how to thank you enough ! such an easy and straight forward guidance ! God Bless.

[@zubaidahyusoff8611]

this is a very easy tutorial to understand for someone who does not have any SPSS background at all. THANK YOU!

[@sadiakhan6227]

Really appreciated. I just finished my 100 mks college project with the help of your tutorial. Great work. Thanks :)

[@ChristianHigherEd]

Sadia Khan Glad to hear it. Congrats!

[@fsolanosmg]

Great thorough breakdown! Very beneficial as I approach writing my dissertation!

[@jessicapreston9808]

This was incredibly helpful! Thank you! I have a horrible DS teacher at the moment..I wish he explained as clearly as you :)

[@johng5295]

Thanks in a million. Awesome. Very well explained. Where have you been all these years.A+

[@MAli-nw7ix]

Very helpful video, Chris, thanks for the effort. Tremendously helped me in my postgrad dissertation. Cheers!

[@mahmoudassaad8212]

Thank you for Sharing This Chris. Very clear

[@lifespotwithmaryanne]

Wow this is helpful.I really needed this

[@georgeoguejiofor9684]

Some of my statistical doubts were cleared . It's impressive!

[@dr.denycedaniels9620]

Thanks!! I understand the concept, with your assistance.

[@uduakokpanefe3944]

Thank you Thank you Than you. Just what I need.

[@DetHanob]

This is so useful and clearly explained. Thanks immensely.

[@edyta3665]

Excellent tutorial. going to look at your other videos. thx

[@RhodaYap]

Super helpful and relevant; and easy to follow. Thank you and I've subscribed! :)

[@sopavanh]

Very helpful VDO! thank you man!

[@michaeldever9055]

love your videos. Basically just saved me for exams,

[@ChristianHigherEd]

Glad to hear it. Best of luck.

[@babyslug]

this was extremely helpful! Thank you so much! Definitely subscribing.

[@RafaelMilaniMedeiros]

Sharp and clear! Thank you!

[@Andreasofspades]

Amazing, thank you! Using this for my quantitative analysis for my dissertation!

[@techexpedient]

good video

[@jinhailongwang]

awsome video, very descriptive. Well done Chris

[@yingzou5484]

thank you so much!! I am doing my dissertation. This is help me a LOT!!!!!

[@MandozaII]

Thank you very much! This helped a lot!

[@MyChidananda]

How can we use Pearson correlation in this case? Likert scale is an ordinal variable, we need to Spearman test.Correct if I am wrong

[@zaccblack103]

i agree with you..

[@simplysanyal1986]

thank you very much.this is why youtube rocks.

[@melissahenehan1685]

Very helpful thank you for sharing!

[@elizabeththomas3582]

This was soo helpful ! Thank you so so much!

[@ravinadageri9540]

its very sensible lesson

[@anfalberier3501]

@chris Olson if my survey questions have sub point, which data should I use for correlation analysis. For example questions 1 has a, b and c. Question 2 has a, b, c and d sub-questions. How could create the correlation analysis. Should I average each question then run the analysis

[@b.b9774]

Thank you! Very helpful!

[@samuchayabehera842]

really sir it is very usefull to me....

[@adeilajaggernauth8295]

Very clear and informative video. Can you tell me how to address the issue of missing data when running correlation. Do i code it or leave it blank when running the test?

[@ChristianHigherEd]

There are a couple ways to address missing values when running a correlation in Excel: delete the cases missing data, delete the variables with missing values, or impute the values for the missing data. This website has some good suggestions and explanations: www.real-statistics.com/descriptive-statistics/missing-data/ I hope this helps.

[@adeilajaggernauth8295]

Thank you very much!

[@fernandojackson7207]

Also, if two independent variables are highly correlated with each other, does it make sense to get rid of one of them ? I assume if the two are highly correlated they do not both contribute on different aspects of the dependent variable, i.e., either variable explains the same underlying aspects as the other?

[@ChristianHigherEd]

Fernando Jackson Sounds right to me.

[@barristars5328]

GOSH!! you saved me there

[@VKjkd]

Very helpful. Question - If you have percentages (overall satisfaction) by group or brands can you correlate them or ate there major issues with using aggregated data in this way?

[@ChristianHigherEd]

VKjkd I'm not sure I completely understand your question, but you may want to consider a Chi Square test kzbin.info/www/bejne/i4HEqIGChrSpY5o

[@MohamedGaber111]

Thanks. Could I use it for event correlation from event log? Is there a method to find out the frequent of an event from events log?

[@asw072]

So sir.. In case of negative correlation coefficient.. Would the effect still be large?

[@saadrehman7769]

very well explained. thank you sir.

[@kanchanasewwandi7579]

I use a likert type questionnaire (5 point scale) in order to measure six variables: the questionnaire I have three different sub scales and each one has likert items which measure one variable. I want to find the correlation between the variables. How to calculate it in Excel?

[@cavejohnson3898]

So good! Thank you!

[@jeanettezarb2888]

Does this work only work on numerical data? Is it possible to do it with sentences from a survey?

[@teacuppy6610]

but this way it only shows the linear correlation through pearson coefficient doesn't it ? if you for example checked your scatter graph for this data (my situation with thesis) and you could see that lets say polynomic trendline is more accurate than linear one, is it still okay to use this way for calculating correlation ?

[@raghadali7998]

Very helpful, Thank you !

[@abdiladif625]

@chris olson if i wanna to find out four variables to there correlation how to do , i tried but does not work

[@ChristianHigherEd]

abdilatif mohamed jama You may want to run a partial correlation analysis on all four variables or a multiple regression. Here are links to some tutorials that may help with that. Good luck. Partial correlation: www.listendata.com/2013/02/excel-formula-partial-correlation-matrix.html Multiple regression: www.excel-easy.com/examples/regression.html

[@eyelsi6109]

Hi, Chris! I want to perform Concurrent Validity and Test-Retest Method. May I use the same set of scores I used in Concurrent for the Test-Retest Method? Thank you so much in advance!

[@ChristianHigherEd]

Hi Ey, I'm sorry, but I don't understand your question.

[@pathoyoutubification]

Awesome job...

[@MohamedGaber111]

Could we do analysis between a several elements and build a heat map?

[@tarikusnumidica8229]

thank you chris its perfect

[@nkeepfer]

You say that you cant calculate the significance of the correlation, but within the "regression" option on the toolpak, you can see the p-value listed. I don't understand what you mean by this statement, please explain

[@ChristianHigherEd]

Yes, the Excel regression analysis does produce a p-value, but the correlation test does not. That is one of the limitations of that analysis when using Excel.

[@charlottegoode975]

Thank you so much

[@fernandojackson7207]

Why can't we do a significance test in Excel? Is it not included in the Analysis Pak? Wouldn't we just do an Anova to test for significance and then do a pairwise test if we reject the initial that they are all zero?

[@ChristianHigherEd]

Fernando Jackson I am not sure why Excel doesn't provide a significance test for the correlation analysis, but I found another video that describes a work around. Specifically, it suggests running a regression analysis. When you only have two variables in the regression, the Multiple R output is the same as the Pearson r and you can use the significance findings from the regression analysis. Here is a link to that video: kzbin.info/www/bejne/rHfGqXiur7GZkKs I hope it helps

[@fernandojackson7207]

Excellent, thanks for both replies.

[@cjs3298]

Thank you sir

[@Drgeius2009]

thank you !

[@devangidesai84]

Hello Chris ...I have a survey data...can I do the correlation analysis between Income and amount willing to spend on any product / service ..its for my hypothesis

[@ChristianHigherEd]

Yes, if both variables are continuous. In other words both variables should be in individual "dollar" increments, not categorical groups such as $10,000 to $15,000, $15,001 to $20,000, etc.

[@devangidesai84]

Chris Olson Well thank you for your feedback. What if we the data in categorical groups ? Will the result be incorrect ?. Please let me know if the some other test that can be performed ...thank you

[@mybabyeatsbroccoli]

Thank you!

[@mariamhusseina.6890]

How to simulate the cross correlation in Matlab between two signals

[@MrsTealFish]

awesome

[@Chiu_Hindusthani]

thank you very much sir!

[@rodbrenbautista218]

how to get the pearson coefficient ? is that the average of all the coefficient ? thank u in advance

[@ChristianHigherEd]

The Pearson coefficient is found in the correlation data analysis output. Its about midway through the video. Hope this helps.

[@anthonydilshanthamel6727]

Thanks a lot

[@croaker260]

WHat do you do if your two sets of data are NOT side by side?

[@ChristianHigherEd]

Steve Cole There are probably more sophisticated ways to do this in Excel, but I would just paste the columns to be tested next to each other and run the analysis. I hope this helps.

[@croaker260]

That's why I ended up doing anyway. It seemed like the simplest solution, but in the future another assignment they may not be quite that easy. So I was hoping for a another solution. Thank you so much for responding. I'm in the healthcare budgeting class and you this video explained in very short time what my instructors lectures could not thank you

[@adellaazer778]

Can anyone recommend essay writing company?

[@myrtiegrinder8687]

*essay-writers.all-about.in* Finance paper was a total disaster for me before I found this website. I didn't know where to start, what to write, but they showed me everything and now I am sure that with their help I can cope with any possible assignment.

[@wiseowlresearch675]

Wise Owl Research

[@harisankaruk]

what is the point of measuring the mean and sd in 5 point scale ? mode makes more sense

[@ChristianHigherEd]

Perhaps, but in my experience the mean is usually used when looking for a correlation.

[@sampathjayasekara294]

Hello. How to conduct correlation and regression test for independent variables. Pls help

[@ChristianHigherEd]

If I understand your question correctly, when you find a strong correlation between two variables in your regression model this is called multicolinarity. You should consider removing one of the variables from the model and run the regression again. I hope this helps.

[@ashishmore7242]

THNKUU SO MUCH

[@zaccblack103]

you said pearson correlation? aren't we supposed to use spearman correlation?

[@ClayREZify]

Subscribed! :)

[@cressnaa4183]

are this satisfaction data here measured in likert scale of 1-5

[@ChristianHigherEd]

Yes

[@cressnaa4183]

hello there thanks for the response. Is it possible to calculate corelation if the data are in frequency like for example price satisfaction 1 28 36 2 34 40 3 26 22 4 18 27 5 10 16 where 1= strongly disagree .... 5= strongly agree

[@ChristianHigherEd]

No. If I understand your question correctly it looks like you have summary data and not raw data. Summary = X number of 1s, X number of twos, etc. for two subjects. You need raw data to run a correlation, I.e., one record for each response from each subject, as in the video. Hope this helps.

[@cressnaa4183]

yes its a summary. great help sir thanks a lot. the headache is finally over

[@BharatBhushan-sr4uy]

sir tel me how to analysis data on genetic diversity in rice using d squre statistics

[@ChristianHigherEd]

Sorry, but I don't understand your question.

[@BharatBhushan-sr4uy]

sir i want to know ,how we analysis data using mahalanobis d square statistics in excel

[@ChristianHigherEd]

Sorry, I don't know how to do that.