week5. Multi-regression

Question 1

what is the perfect positive correlation

Answer 1

1.00

Question 2

what is the perfect negative correlation

Answer 2

-1.00

Question 3

if the scatter plot has a curve in the data, is there a correlation

Answer 3

no

Question 4

what is an outlier

Answer 4

an odd one out of the regression line

Question 5

what is the difference between correlation and regression

Answer 5

correlation = association between variables
regression = enables us to predict the value of 1 variable if we know the value of the other variable

Question 6

what is the regression formula

Answer 6

y = bx + c

Question 7

what to the different letters stand for in the regression formula

Answer 7

y = the thing you’re trying to predict (the y axis on a graph), b = the slope angle , x = the predictor (the x axis on a graph), c = the constant

Question 8

what are the different types of regression

Answer 8

bivariate linear, multilinear, logistic, curvilinear

Question 9

what is bivariate linear

Answer 9

a simple linear regression - a linear equation describing the relationship between an explanatory variable and a outcome variable, specifically with the assumption that the explanatory variable influences the outcome variable

Question 10

what are the residuals

Answer 10

is it the distance (drawn parallel to the y axis) from the data points to the line

Question 11

is it best to have big residual or small residuals and why

Answer 11

the smaller the residuals the less errors and this betters the prediction (the the data is scattered less)

Question 12

ia regression symmetrical

Answer 12

no

Question 13

is correlation symmetrical

Answer 13

yes

Question 14

how is regression line scissors and how does it help explain

Answer 14

it looks like them, if the scissors are closed, that means r= either -1.00 or 1.00, if the scissors are open at a 90degree angle, then r=0

Question 15

what is explanatory power

Answer 15

R2 is assessed by the square of the correlation coefficients R2 (sometimes called the coefficients of determination)

Question 16

why do we use multiple regression

Answer 16

it helps predict complex situations involving several variables where experimental control isn’t possible

Question 17

what are some problems with using multi-regression

Answer 17

the sample size must be at least 5 times bigger than the number of variables

Question 18

how is multiple regression an extension of a bivariate regression

Answer 18

you just need to add more x’s to the equations ( y= b1x1 + b2x2 + b3x3… +c)

Question 19

what info does R2 adjusted give

Answer 19

it gives the percentage of variance accounted for by the model

Question 20

how to do present a multiple regression

Answer 20

f= f value, df = df value, p = p value)
mention the R2 too, find this from the ANOVA table

Question 21

what table do you look at to report the results of a multi regression

Answer 21

look at the ANOVA table and coefficients table

Question 22

what do you find on the coefficients table

Answer 22

shows significant predictor, which variables are significant predictors, the regression equation