Correlation
📊 What is Correlation?
helps understand whether variables change togther.
Form: understood by scatterplot
direction: ➕ 📈 / ➖📉 / 0 ░
magnitude:
correlation
🦜What does the correlation coefficient (r) measure?
range: r always falls between -1.00 & +1.00
correlation
🔑 What are the properties of the correlation coefficient?
(three qualities)
form
Direction:
- Positive (+): variables move in the same direction
- Negative (-): one increases ⬆️ while the other decreases⬇️
- zero: no linear relationship
Form
- measures linear relationships
magnitude
- strength of the relationship (closer to +/-1 = stronger)
What is the difference between positive, negative, and no correlation?
➕📈Positive correlation:
- As one variable increases, the other also increases.
- scatterplot slopes upward
**ex) **
- Hours studied ⬆️ Exam score⬆️
- salary ⬇️ job satisfaction⬇️
-self-esteem ⬆️ body satisfaction⬆️
➖📉 Negative correlation: ⬆️⬇️
- As one variable increases, the other decreases.
- scatterplot slopes downward
ex.)
-hrs procrastinating ⬆️ Exam score⬇️
- pain⬆️ hapiness⬇️
- self-esteem⬇️ sadness⬆️
- scatterplot looks random
0️⃣ ░ No correlation
- no relationship… two variables not related
- ex.)
- hrs of procrasti - numbers of hairs
- pain - # of snickers on halloween
How do you interpret the magnitude and direction of a correlation?
magnitude (strength):
- small = 0.2
- medium- 0.5
- large > 0.6
- (closer to +/- = stronger relationship)
- (context matters– what counts as “stong” depends on the field)
- looking at graph:
- strong relationship = scatter plots tighter- linear line is visblie
- weak relationship = liear line still visible– scatterplots looser
Direction:
- positive–> both variables move together
- negative –> variables movie in opposite directions
- zero –> no relationship
ex.)
r= 0.7: strong positive correlation
r= -0.30: weak negative correlation
r= 0.00: no correlation
📊 Using a Correlation Table
Why does correlation not mean causation?
Correlation only shows association not cause- & effect
ex.) - Example: Ice cream sales and sunburns are correlated, but hot weather causes both.
What is spurious correlation?
A misleading correlation caused by coincidence or a third variable
ex. nicolas cage films vs. swimming pool drownings