You are on page 1of 3

Project 3

-I think height don’t have anything to do with the numbers of candies per bag because there are no
possible criterion to connect the two variables to a linear relating equation.

-If Height and number of candies per bag purchased by that same person is related then it should be
height is explanatory variable while numbers of candy per bag is response variable because if these two
variables are related then I think it should be the height that is independent because we can choose the
bag of candy to buy but we can’t choose our height.

-Simple linear regression results:

Dependent Variable: var6


Independent Variable: var7 var6 =
63.116326 - 0.0606596 var7
Sample size: 86
R (correlation coefficient) = -0.093031197
R-sq = 0.0086548036
Estimate of error standard deviation: 2.5639763

Parameter estimates:

Parameter Estimate Std. Err. Alternative DF T-Stat P-value


Intercept 63.116326 4.6385057 ≠0 84 13.607039
<0.0001
Slope -0.0606596 0.070834323 ≠0 84 - 0.3942
0.85635887

Analysis of variance table for regression model:

Source SS MS F-stat P-
DF value
Model 1 4.8210275 0.73335051
4.8210275 0.3942
Error 84 552.21386
6.5739745
Total 85
557.03488
-R (correlation coefficient) = -0.093031197, so the critical value that has n>30 is = 0.361 and
since the absolute value R is smaller than the critical value> there is no linear relation. This is
pretty much expected because there shouldn’t be any linear correlation between heights and
number of candies.
-y hat= B1x+B0
B1=r*(Sy/Sx); Sy=2.55995; Sx =3.926s B1= -0.093031197*(2.55995/3.926)= -0.06066

And B0 = mean of Y – B1*mean of X; mean of x=65.367442; mean of Y= 59.151163 so


B0=59.151163-65.367442*-0.06066=63.116352

And the equation should be Y hat= -0.06066*x+63.116352; my height is 61 inch so


the number of candies that I get is -0.06066*61+63.116352=59 candies

I think this regression line shouldn’t be able to predict how much candies someone is
going to get because height and numbers of candy are not relative with
each other.

-R=-0.093031197 so R^2=-0.008655, so 0.865% of the variation in number of candies per bag

is explained by the least squares regression line

-It isn’t appropriate even if there is actually a significant linear relation, since 90 is out of the

scope for our data we can’t predict for Yao Ming because we don’t have anyone in class near to

90 inch height.
I just got the calculator so the instruction told me to go to stat and choose 1(edit) and there I

type in my x and y value, my 4th and every 10th value is different from yours so I used yours

(already ignored the headers) The x-values are the heights: 63.5, 60, 62, 67.2, 66, 62.5, 60, 68,

68

The y-values are the totals: 55, 57, 58, 59, 59, 60, 61, 61, 64

After enter the value I go to stat and right to calc and choose LinReg (a+bx) enter and then press

calculate and it gives me this r (correlation coefficient)= 0.4453851936, now the mean of x and

y I did the same step but instead of linqu I choose the first one (stat something) both stat crunch

and the calculator gave me 64.1333333 I do the same for Y in calculation I choose that as L2

which is the second column of data and both gave me 59.33333, the same goes for standard

deviation; now from here I guess this is when I need to do it by hand and plug numbers into B1

and B0 and I got equation: Y hat= B1*x+ B0

B1=r*sy/sx=0.4453851936*2.5980762/3.2515381=0.3557

B0=y mean- b1* x mean= 36.5186

So the equation is y hat= 0.3557*x+ 36.5186

According to critical value table 9 is 0.666 and since r=0.4454 which is smaller than critical

value 0.666 there is no significant linear relationship between x and y

You might also like