OLS and Regression

Question 1

why do you divide by variance of x in a BLP

Answer 1

Because you want to understand how Y varies for every change in X

Question 2

What does the BLP produce

Answer 2

The expected change in Y for every 1 unit change in X

Question 3

OLS estimator or regression to estimate the BLP

Answer 3

We try and estimate it

Question 4

What can this estimator do?

Answer 4

Using the closed form solution can use the formula for B

Question 5

Error term

Answer 5

Non systematic noise

Question 6

Formula for B

Answer 6

Sum of (Yi-Expectation of Y) (xi - Expectation of X) / sum of (Xi - Expectation of X)2

Question 7

Intercept

Answer 7

B0

Question 8

B1

Answer 8

Effect of change in one unit of the explanatory variable

Question 9

standard deviation of normal variables

Answer 9

approximately two thirds of a normal distribution is within one standard deviation of the mean

Question 10

difference between standard deviation and standard error

Answer 10

the SD quantifies the variation within a set of measurements, whilst SE quantifies the variation in the means from multiple sets of measurements

Question 11

SD

Answer 11

Quantifies how much the data are spread out- quantifies how much much measurements are spread around their means

Question 12

what is the standard normal distribution

Answer 12

has a mean of 1 and SD of 1

Question 13

z-score

Answer 13

how many standard deviations an observation a number is from the mean

Question 14

Z-score table

Answer 14

Lets us know how much of an area that corresponds to a Z score, which corresponds to an X value

Question 15

normal dis density

Answer 15

bell curve -describes tendency of values to cluster around the mean - most of the data values located near the mean, it is symmetric around its mean

Question 16

B0 interpretation

Answer 16

The average value of the Y when the X (B1) = 0

Question 17

Cont x Cont B1

Answer 17

For every one additional X, there is a B1 change in Y

Question 18

Bin x Cont

Answer 18

For every additional X, there is a B1 change in the probability that an individual/unit is a Y

Question 19

Cont x Bin

Answer 19

The difference in Y between those that do and do not have/do X is B1

Question 20

Bin x Bin

Answer 20

The difference in the probability of being Y between those that do or do not X is B1

Question 21

what happens if regression doesnt mean much?

Answer 21

use sd

Question 22

how to calculate sd in this case

Answer 22

x : sd(x xB1) / SDy

Question 23

how to interpet sd in a regression if used

Answer 23

a one standard deviation increase in X is associated with a … standard deviation change in Y

Question 24

what do we assume in regression

Answer 24

that values are normally distributed

Question 25

B0 is...

Answer 25

gives us the expected value for y when X is 0

Question 26

limitations to BLP

Answer 26

Often, you want to control for stuff but doesnt let you

Question 27

Multivariate

Answer 27

Y = alpha + B1x1 + B2X2 + e

Question 28

B1

Answer 28

partialling out interpretation

Question 29

what happens when you 'hold constant'

Answer 29

are imagining you are only comparing between countries w similar variable held constant

Question 30

what does linear mean in linear regressions

Answer 30

linearity in our betas