Research methods- correlations

Question 1

What does a scatterplot describe

Answer 1

is a visual way to describe the assossiation between 2 variables. Every dot on the scatterplot is a ppt

Question 2

how can we describe the relationshp between two variables

Answer 2

direction
strength
shape

Question 3

describe a positive asssossiation

Answer 3

Larger values on variable 1 are associated with larger values on variable 2​

Smaller values on variable 1 are associated with smaller values on variable 2​

Question 4

describe a negative assossiation

Answer 4

Larger values on variable 1 are associated with smaller values on variable 2​

Smaller values on variable 1 are associated with larger values on variable 2​

Question 5

describe the strenght of a correlation

Answer 5

If the strength is zero- no association.​

Dots tightly on the line – one variable predicts the next variable

Question 6

do correlations have to be linear

Answer 6

no - they can have a u shape or inverted u shape

Question 7

What are correlation co-efficient

Answer 7

e descriptive statistics that describe an association between variables​

the two covered in this lecture was Pearsons Product-Moment correlation

spearmans Rank Correlation

Question 8

What is pearsons correlation

Answer 8

Describes linear associations between variables​

Used for interval/ratio level data that is normally distributed​

Question 9

what is a spearmans correlation

Answer 9

Describes linear associations between variables​

Used for ordinal level data or interval/ratio data that is not normally distributed (skewed)

Question 10

in Pearsons correlation what are the values to describe the strength and direction

Answer 10

-1 to 1

Question 11

what is the formula for pearsons correlation

Answer 11

r= Cov( x, y)
————-
SDx +SDy

cov ( x, y) = sum ( xi - mean of x) x (yi - mean of y )

Question 12

What does Spearmans rank order correlation use

Answer 12

Uses ranks instead of raw data values​

Interpretation identical to Pearson correlation (-1 = perfect negative correlation, 0 = no correlation, +1 = perfect positive correlation)​

Question 13

what are the first 3 steps for calculating spearmans correlation

Answer 13

1) rank the data seperately from lowest to highest.
2) if there are any ties complete the median rank
3) compute the pearson’s correlation for ranked variables.

cov ( x, y) / sd1 +sd 2

cov (x,y) - E(xi - mean of x ) x ( yi - mean of y)/ SD1 +SD2

Question 14

how do we determine which hypothesis to use for a correlation test

Answer 14

Test significant (p < .05)  Reject the null hypothesis  Decision for H1
Test not significant (p > .05)  Cannot reject the null hypothesis  Decision for H0

Question 15

how to report a correlation test in apa-7

Answer 15

Spearman’s correlation analysis showed that there was a significant positive correlation between fear of missing out and the number of drinking establishments attended, rs (33) =.62, p<.001.

name the correltion
what is significant
positive or negative correlation
variables
inferential statistic gotten from R studio

Question 16

in the example : rs (33)=.62, p< .001

what does the values mean

Answer 16

rs= spearmans rho
33- degrees of freedom ( n-2)
.62- co-efficient
p<.001 = p-value

Question 17

What is degrees of freedom

Answer 17

number tells you something about the sample size and the specific test that was conducted

’Degrees of freedom tell you how many values in your data can vary independently after applying the rules of the statistical test.

Degrees of freedom are usually a bit smaller than the number of observations

Question 18

What is a one-tailed correlation test

Answer 18

A one-tailed correlation test checks whether the correlation between two variables is significantly positive OR negative, but not both.
You must choose the direction (r > 0 or r < 0) before performing the test.

Question 19

When should you use a one-tailed correlation test?

Answer 19

Use it when you have a specific, theory-driven directional hypothesis (e.g., you expect a positive or negative relationship).
Example: Expecting higher attendance to be positively related to higher exam scores.

Question 20

What are the hypotheses for a one-tailed test predicting a positive correlation?

Answer 20

H₀ (Null):
𝑟
≤
0
r≤0 → no correlation or a negative correlation

H₁ (Alternative):
𝑟
>
0
r>0 → positive correlation in the population

Question 21

How do you decide whether to reject the null hypothesis in a one-tailed correlation test?

Answer 21

f p < .05, reject H₀ → evidence supports a positive correlation.

If p > .05, fail to reject H₀ → no evidence of a positive correlation.

Question 22

How does a one-tailed test differ from a two-tailed test?

Answer 22

one-tailed: tests for a correlation in one direction (r > 0 or r < 0).

Two-tailed: tests for a correlation in either direction (r ≠ 0).
A one-tailed test only looks at one side of the sampling distribution.

Question 23

What happens to the rejection region in a one-tailed test?

Answer 23

The rejection region is placed on only one tail of the sampling distribution — the direction specified by your hypothesis (e.g., the right tail for r > 0).

Question 24

How can you convert a two-tailed p-value to a one-tailed p-value?

Answer 24

Divide the two-tailed p-value by 2 — but only if the observed correlation is in the predicted direction (e.g., positive for a right-tailed test).
If the correlation is opposite, the one-tailed p-value is not valid.

Question 25

What are the risks of using a one-tailed test incorrectly?

Answer 25

Switching to a one-tailed test after seeing a non-significant result is not acceptable and considered cheating One-tailed tests must be pre-registered, theoretically justified, and used only when direction matters.

Question 26

In scientific articles how are correlations reported

Answer 26

in correlation tables the top half of the table is empty as the correlation has already been shown

Question 27

Why is correlations different from causation

Answer 27

because the relationships is measured so the correlation could be surpious. Meaning there is a presence of confounding variables.