L17 - Regression 1

Question 1

what is a model?

Answer 1

model is a mathematical, physical or otherwise logical representation of a system that is used to convey real world events or phenomenon.

Question 2

what is the main difference between correlation and regression?

Answer 2

correlation just looks at how two variables are related, Regression looks at dependent and independent variables (X,Y) and explains the relationship between them.
regression attempts to make a causal inferance

Question 3

how do you form a Hypothesise a regression model between two variables?

Answer 3

determine which variable may be dependent on another (which one is X and Y).
then hypothesize the nature of the relationship, positive or negative?

Question 4

what is the equation for the linear regression model?

Answer 4

Y = B0 + B1X + E
where:
B0 = y intercept
B1 = lope of the line
X = indipendente variable
E = Error

Question 5

what are the assumptions for a simple linear regression (4)

Answer 5

1. Data is interval or ratio
2. the relationship between x and Y is linear
3. we can identify one variable
4. the x variables are measured without error

Question 6

what does a residual/error plot do?

Answer 6

it plots variables residual scores on a graph to see if a trend is visible. if a trend is visible then the data variables understudy may be linked through another variable not considered.

Question 7

in order for the regression model to be accepted there are four requirements that mus be satisfied

Answer 7

1. the errors must be normally distributed
2. the errors have a mean of zero
3. the standard deviation of error is constant for all values of X (homoscedasity)
4. The set of errors associated with different values of Y are all independent.

Question 8

what is error in a regression model?

Answer 8

error is the difference between the observed and predicted value of Y. for each value X

Question 9

draw out the sum of squares diagram including Total, Unexplained (error) and total

Answer 9

refer to slides

Question 10

why do we find the line of best fit? what does it tell us?

Answer 10

the line of best fit is a form of model that is used to predict values of Y for given X that may be outside the dataset.

Question 11

what is the Coefficient of determination?

Answer 11

R^2 tells us how much of the variation in Y can be explained by X.

          Explained variation  R^2 =\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_
             Total variation

Question 12

what does R^2 = 0.67 mean?

Answer 12

it means that 67% of the variation in Y can be explained by X

Question 13

what is the error term within a residual plot

Answer 13

it is the amount of error within a regression model. depending on the distribution of the error there will be different

Question 13

what is the error term within a residual plot

Answer 13

it is the amount of error within a regression model. depending on the distribution of the error there will be different