Test Pilot 2.x EDPSY 507 MIDTERM RLH EDPSY 507 Midterm 'Please choose or enter the best answer. Best wishes on this examination. WHITE Helvetica,Arial 4 BLACK 7 1 ` ‡The major equation in this course is MODEL = DATA + ______ {PRE
F statistic
(1-PRE)
ERROR
PRE F statistic (1-PRE) ERROR 0 0 0 1 3Sorry, this is the proportional reduction in error. QSorry, the F statistic is often calculated, but is not this part of the equation. GSorry, this is the proportion of error left after comparing two models. Correct 01 points awarded for selecting ERROR
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ j âWhen two different models are compared using the model comparison approach, the model with the fewer estimated parameters is called the ______ model. w
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Correct Sorry, it is a compact model. ÿÿÿÿÿÿÿÿÿÿÿÿ compact t ×When comparing two models using the model comparison approach, the model that estimates the most parameters is called the _________ model. w
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ÿÿÿÿÿÿÿÿÿÿÿÿ augmented ~ ÅThe smallest number of parameters that can be estimated in a single model used in the model comparison approach is ____. pzero
one
two
three
zero one two three 1 0 0 0 Correct ASorry, no parameters need be estimated to create and use a model. ASorry, no parameters need be estimated to create and use a model. ASorry, no parameters need be estimated to create and use a model. /1 points awarded for selecting zero
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ ˆ ÏGiven the following data [a],[b],[c],[d],and [e], and a model with the single parameter of [f], what is the sum of squared errors? w
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Congratulations ?ð @4 ?ð @3 a?ð @4 ?ð @3 b?ð @4 ?ð @3 c?ð @4 ?ð @3 d?ð @4 ?ð @3 e@ @. ?ð @$ f ÿÿÿÿÿÿÿÿÿÿÿÿ?ð '(a-f)^2+(b-f)^2+(c-f)^2+(d-f)^2+(e-f)^2 ;([a]-[f])^2+([b]-[f])^2+([c]-[f])^2+([d]-[f])^2+([e]-[f])^2 ’ If the following were the dependent variable variable values in a simple regression analysis, what is the compact model's value. The dependent variable values are: [a], [b], [c], [d], [e], [f]. w
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Congratulations >Sorry, the mean of these variables would be the compact model. ?ð @4 ?ð @3 a?ð @4 ?ð @3 b?ð @4 ?ð @3 c?ð @4 ?ð @3 d?ð @4 ?ð @3 e?ð @4 ?ð @3 f ÿÿÿÿÿÿÿÿÿÿÿÿ?ð (a+b+c+d+e+f)/6 ([a]+[b]+[c]+[d]+[e]+[f])/6 œ ¼Looking at the simple regression output below, what is the PRE value of comparing this model to the mean model? w
@ ð
simple.gif 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Correct. aSorry, the R-squared value provides the answer. However, PRE is a proportion, and not a percent. ÿÿÿÿÿÿÿÿÿÿÿÿ .7483 0.7483 0.75 .75 ¦ ¨Looking at the output below, what is the value for the constant in the simple linear model. w
@ ð
simple.gif 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Correct fThe constant and the intercept as reported are the same thing. Thus, the constant is equal to 0.48551 ÿÿÿÿÿÿÿÿÿÿÿÿ 0.48551 .48552 .486 .48 ° ÆLooking at the simple regression output below, is the simple model displayed significantly different from the mean model? çYes
No
@ ð
simple.gif Yes No 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Correct 4Sorry, the p value for the t or F statistic is .0001 .1 points awarded for selecting Yes
ÿÿÿÿÿÿÿÿÿÿÿÿ º ìLooking at the simple regression output below, what is the p value for the F statistic comparing this model with a compact model which uses the mean of Achiev? w
@ ð
simple.gif 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ÿÿÿÿÿÿÿÿÿÿÿÿ .0001 1.0E-4 0.0001 Ä ÅLooking at the regression scatterplot shown below, which line represents the single augmented model applied to the data. öTop red line
Top blue line
Middle black line
Bottom blue line
Bottom red line
@ ð scatter.gif Top red line
Top blue line Middle black line Bottom blue line Bottom red line 0 0 1 0 0 ]Sorry, this is the upper boundary of the regression line plus the standard error of estimate. lSorry, this is the upper boundary of the regression line itself using the confidence interval for the slope. Correct lSorry, this is the lower boundary of the regression line itself using the confidence interval for the slope. ^Sorry, this is the lower boundary of the regression line minus the standard error of estimate. <1 points awarded for selecting Middle black line
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ Î ¦Looking at the scatterplot shown below, which of these lines represent the compact model. —The red lines
The black line
The blue lines
None of these
@ ð scatter.gif
The red lines The black line The blue lines
None of these 0 0 0 1 =Sorry, these set the confidence intervals around predictions. "Sorry, this is the augmented model BSorry, these set conficence intervals around the augmented model. UCorrect, the compact model would be a line with no slope drawn at the mean of Achiev. 81 points awarded for selecting None of these
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ Ø ¾Looking at the Regression ANOVA tab output shown below, what are the sum of squared errors for the compact model? ƒ0.0295587
0.00994133
0.0023667
0.0395
@ ð
output.gif 0.0295587
0.00994133 0.0023667 0.0395 0 0 0 1 PSorry, this is the difference in error between the compact and augmented models. 1Sorry, this is the error for the augmented model. 3Sorry, we haven't talked about this error in class. Correct 11 points awarded for selecting 0.0395
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ â ÕLooking at the ANOVA tab output from the simple regression output shown below, what is the sum of squared error for the augmented model? „0.0295587
0.00994133
0.00236667
0.0395
@ ð
output.gif 0.0295587
0.00994133
0.00236667 0.0395 0 1 0 0 USorry, this is the difference between the errors in the augmented and compact models. Correct 'Sorry, we haven't discussed this error. /Sorry, this is the error for the compact model. 51 points awarded for selecting 0.00994133
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ ì ¿Looking at the ANOVA tab from the simple regression procedure above, how many people were evaluated in this study? i42
43
16
44
@ ð
output.gif 42 43 16 44 0 0 0 1 -1 points awarded for selecting 44
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ ö ¾Looking at the multiple regression output shown below, how many parameters are estimated in this augmented model? e2
4
5
7
@ ð
multipler.gif 2 4 5 7 0 0 1 0 ,1 points awarded for selecting 5
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ HLooking at the output below, what is the approximate F statistic if a researcher wanted to compare a compact model which used a constant, miles, carpool, and starttime to estimate the dependent variable against an augmented model that is shown below? w108.7
190.72
0.7921
63.36
@ ð
multipler.gif 108.7 190.72 0.7921 63.36 0 0 1 0 9Sorry, this is the F using the mean as the compact model. MSorry, this would have a compact model with everything except miles included. Correct WSorry, this would have a compact model with all the predictors except carpool included. 11 points awarded for selecting 0.7921
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ
Looking at the multiple regression output shown below, if the mean on the dependent variable is the compact model, what is the PRE found for determining whether this augmented model is better? m60%
70%
90%
96%
@ ð
multipler.gif 60% 70% 90% 96% 0 0 0 1 .1 points awarded for selecting 96%
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ Looking at the multiple regression output shown below, which of the following compact models could be directly tested against this full augmented model by clicking the Further ANOVA tab? Íd = bo + b1miles + b2lights + b3carpool
d = bo + b1lights + b2carpool + b3starttime
d = bo + b1starttime
d = bo
@ ð
multipler.gif 'd = bo + b1miles + b2lights + b3carpool +d = bo + b1lights + b2carpool + b3starttime d = bo + b1starttime d = bo 1 0 0 0 R1 points awarded for selecting d = bo + b1miles + b2lights + b3carpool
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ ÿThere are times (although the circumstances are unusual) when a model with more estimated parameters is poorer (produces more error) than a model with fewer estimated parameters? ëTrue
False
True False 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Sorry, this can never happen. 01 points awarded for selecting False
ÿÿÿÿÿÿÿÿÿÿÿÿ ( ÇError in a model can always be reduced to zero when the number of parameters estimated are equal to the number of________. ’predictor variables
subjects
constants
none of these
predictor variables subjects constants
none of these 0 1 0 0 Sorry 31 points awarded for selecting subjects
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ 2 ¡Tolerance is a measure of the redundancy between a predictor and criterion variable. ëTrue
False
True False 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 cSorry, tolerance is a redundancy measure between predictor variables not a predictor and criterion. 01 points awarded for selecting False
ÿÿÿÿÿÿÿÿÿÿÿÿ < “Match the following statistics with the type of outlier they identify. Ë
1. Studentized deleted residual
Unusual predictor value
2. Lever
Unusual criterion value
3. Cooks D
Unusual influence on regression
Lever Studentized deleted residual Cooks D 1 1 1 >1 points awarded for Lever::Unusual predictor value U1 points awarded for Studentized deleted residual::Unusual criterion value H1 points awarded for Cooks D::Unusual influence on regression ÿÿÿÿÿÿÿÿ Unusual predictor value Unusual criterion value Unusual influence on regression 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Hale fanny1 h12@psu.edu