fuck

Question 1

What is correlation and regression

Answer 1

Looks for relationships

looks at similaritires between samples instead of divergences

see whether one sample varies alongside the variations of the other sample (covariance - how 2 variables shift together)

Question 2

What is covariance

Answer 2

strong covariance = bigger similarity in movement

sum between the products of the individual deviations between two variables

measurement of how each variable deviates together

does not take into consideration error - identify proportion between the covariance and the individual deviations of the variables

Question 3

How to control for error

Answer 3

Coefficient called Pearson’s r

Question 4

What is correlation

Answer 4

statistical technique used ot measure and desicribe a relationship between 2 variables

2 variables are observed as they exist naturally; no attempt to control or manipulate variables

Ex. Height and Intelligence, socioeconomic status and length of marriage

Question 5

What can corelation be used for

Answer 5

Prediction - based on trends, and not based on causality

Validity - measurement and testing; Are scales/tools able to measure the right concepts for a study

reliability - are scales/tools used for measuring able to measure consistently

Theory verification - are theories able to correctly proide explantopns

Question 6

What are the characteristics for correlation

Answer 6

1. Shape - Straight line since measures linear relationship only
2. Direction - either positive (X increases Y also increases) or negative (X increases Y decreases)
3. Strength - measures the degree to which points fit the straight line; if all points fall exactly on line, a perfect relationship exists. The more the scatterplot resembles line, the stronger the correlation.

1.00 = perfect relationship while 0.00 = no relationship

4. Significance - does the observed correlation between X and Y really exist in the population and is not due to chance or error

Use Table F, Pearson r for alpha at .05 or .01 with df = 1-90

only when result is significant we interpret but if not significant do not interpret

Question 7

What is the values for strength in correlation

Answer 7

.00 - .29 - weak
.30 - .69 - moderate
.70 - 1.00 strong

Question 8

What are the strengths of correlation

Answer 8

Describes the relationhip between only 2 variables

naturabl observation - no interference or manipulation

accurately reflects the natural events being examined

Question 9

What are the weaknesses of correlation

Answer 9

Third variable may interfere with the two variables and can be
responsible for the observed relation

Does not determine cause or effect- Non-directional relationship = none of the variables can claim
precedence

Does not produce a clear and unambiguous explanation for the
relationship

Question 10

What is Pearson’s R

Pearson Product-Moment Correlation Coefficent

Answer 10

The most common correlation

Measures the degree of straight-line relationship bewteen two variables at a time

Question 11

What are the assumtpions and requirments of Pearson’s R

Answer 11

A straight line or linear relationship

Both X and Y are variables that must be measured at the interval level

sample members must be drawn from a random sample

Both X and Y variables must be normally distributed

Sample size must be at least 30 to disregard normality violations

Question 12

How do we test for Correlation

Answer 12

1. State the hypothesis
   Ho: There is no correlation between
   X and Y
   Ha: There is a correlation between X
   and Y
2. Set the Level of Significance at .05
3. Compute
4. Interpret - use table F

df = N-2 where N is the number of paired scores

Robt > Rcrit

Question 13

How do we interpret the score that we computed using Pearson’s R

Answer 13

.00 - .29 = Weak
.30 - .69 = Moderate
.70 - 1.00 = Strong

Question 14

What are the final notes on correlation

Answer 14

Correlation score should not be interpreted as a proportion

Looks at the strength and direction of correlation value

Does not imply causation -existence of correlation does not imply the existence of a causal link bewteen two variables

Describes relationship between two variables and does not explain why they are related

Question 15

What are the other final notes on correlation (Factors that can affect correlation)

Answer 15

Possible that correlation is due to a common third variable (causing the 2 variables)

Correlation is affected by the restriction of the range - If only a restricted, more homogeneous and selective subest of entire range is included, expect correlation to be weaker

Must ensure that entire range (or the widest range possible) of X and Y values are sampled

Getting only a subset of the entire range of X and Y value will weaken the correlation

Question 16

How to determine the accurcy of our prediction

Answer 16

By using r squared which is the coefficient of determination

Question 17

What does the coefficient of determination do

Answer 17

Squaring r measures the proportion of percentage variability in the DV determined by the IV

Portion obtained will tell the protion of DV that is predicted by the IV

-0.50 means that one variable is partially associated, but the variability portion is only r squared = 0.25/25% of the total variability

There are other (extraneous) variables that are affecting the relationship