Real and Illusory Correlations, Scatterplots, and Causation (Intro Psych Tutorial #12)
HTML-код
- Опубликовано: 17 ноя 2016
- In this video I explain the dangers of illusory correlations and confirmation bias, correlational studies, patterns of variation, r-values, scatterplots and conclusions about causation.
Try guessing the r-value from a scatterplot at www.guessthecorrelation.com
Don’t forget to subscribe to the channel to see future videos! Have questions or topics you’d like to see covered in a future video? Let me know by commenting or sending me an email!
Need more explanation? Check out my full psychology guide: Master Introductory Psychology: amzn.to/2eTqm5s
Enable closed-captioning or find a full transcript of this video here: www.psychexamreview.com/real-a...
My exam is in 2 days and these videos are helping me so
much thank u!
Let me know if you have any questions, best of luck on the exam!
Great video!
Thanks!
I am preparing to start a Psychology study next month. I have the feeling I will be prepared very well after watching your videos. That is a causation, not correlation right:-)?
I'd like to hope my videos play a role, but I think your willingness to prepare might be the causal factor here
@@PsychExamReview That is true. Thank you very much again for making this great quality material accessible for everybody.
Dear Sir,
Thank you so much for your valuable inputs in enhancing our knowledge and understanding of psychology.
Could you please let me know whether causation can exist when correlation is zero? I am literally stuck at this point.
And Sir is there any relevance to the statement that, ' Correlation is neither necessary nor sufficient condition for causation' ?
I would be really grateful to you if you could enlighten me regarding the issue.
Warm regards,
I'm glad to hear that you've found my videos helpful. For the question about causality without correlation, this depends on how we measure correlation. It is possible for two variables to have a causal relationship and yet lead to a Pearson's r value of 0. This is because this calculation only tells us about linear correlation, not non-linear correlation.
One example is to imagine that as variable X increases it causes Y to increase then decrease, then increase then decrease, etc. like a wave. The variables may be causally related, but this could cancel out when calculating a linear correlation to show a correlation of 0.
The variables would have a clear relationship (and the pattern would be clear in a scatterplot), but they wouldn't show a linear correlation, so other non-linear methods of calculating correlation would be needed for describing their relationship. It's also possible for 2 variables to have a causal relationship that is only seen in the context of a 3rd variable (X cannot predict Z, and Y cannot predict Z, but X and Y together can predict Z).
Hope this helps!
@@PsychExamReview I would like to extend my profound gratitude to you for taking my doubts under consideration. This was helpful to a great extent.
Thank you so much Sir!
And Wish you all the best of luck for all your future endeavours.
@@sigmawarrior2727 Thanks, all the best to you too!
@@PsychExamReview
Dear Sir,
Is it true that if two variables are proportional then they will always be correlated?
Love the videos, but man, you gotta lose that highlighter, scraper sound at the beginning.
Yeah, will replace it with something else in future videos
PsychExamReview na it sounds cool
@@danielt7165 same, it feels like he is personally teaching it to me!!!