Test 3D AP Statistics Name:

Directions: Work on these sheets.

Part 1: Multiple Choice Circle the letter corresponding the best answer.
1. For children between the ages of 18 months and 29 months there is a!!ro"imatel# a linear
relationshi! between height and age$ The relationshi! can be re!resented b#
% y
& '($93 ) *$'3x where y re!resents height +in centimeters, and x re!resents age +in months,$
-ose!h is 22$. months old and is 8* centimeters tall$ /hat is -ose!h0s resid1al2
3 +a, 49$1 +b,

*$9 +c, *$9 +d, .'$' +e, '($93

2. A st1d# e"amined the relationshi! between the se!al length and se!al width for two 5arieties of an
e"otic tro!ical !lant$ 6arieties A and 7 are re!resented b# "8s and o8s res!ecti5el# in the following
scatter!lot$ /hich of the following statements is FA9S:2
+a, ;onsidering 6ariet# A onl# there is a negati5e
correlation between se!al length and width$
+b, ;onsidering 6ariet# 7 onl# the least<s=1ares
regression line for !redicting se!al length from se!al
width has a negati5e slo!e$
+c, ;onsidering both 5arieties there is a !ositi5e
correlation between se!al length and width$
+d, ;onsidering each 5ariet# se!aratel# there is a !ositi5e correlation between se!al length and
width$
+e, ;onsidering both 5arieties the least<s=1ares regression line for !redicting se!al length from
se!al width has a !ositi5e slo!e$
3. >n ?a# 11 .* randoml# selected s1b@ects had their s#stolic blood !ress1re +S7P, recorded twice
Athe first time at abo1t 9:** a$m$ and the second time at abo1t 2:** !$m$ Bf one were to e"amine
the relationshi! between the morning and afternoon readings then one might e"!ect the correlation
to be
+a, near Cero as morning and afternoon readings sho1ld be inde!endent$
+b, high and !ositi5e as those with relati5el# high readings in the morning will tend to ha5e
relati5el# high readings in the afternoon$
+c, high and negati5e as those with relati5el# high readings in the morning will tend to ha5e
relati5el# low readings in the afternoon$
+d, near Cero as correlation meas1res the strength of the linear association$
+e, near Cero as blood !ress1re readings sho1ld follow a!!ro"imatel# a Normal distrib1tion$
4. S1!!ose we fit a least<s=1ares regression line to a set of data$ /hat is tr1e if a !lot of the resid1als
shows a c1r5ed !attern2
+a, A straight line is not a good model for the data$
+b, The correlation m1st be *$
+c, The correlation m1st be !ositi5e$
+d, >1tliers m1st be !resent$
+e, The regression line might or might not be a good model for the data de!ending on the e"tent of
the c1r5e$
>ne concern abo1t the de!letion of the oCone la#er is that the increase in 1ltra5iolet +D6, light will
decrease cro! #ields$ An e"!eriment was cond1cted in a green ho1se where so#bean !lants were
;ha!ter 3 1 Test 3D
e"!osed to 5ar#ing le5els of D6 meas1red in Dobson 1nits$ At the end of the e"!eriment the #ield +Eg,
was meas1red$ A regression anal#sis was !erformed with the following res1lts:
5. The least<s=1ares regression line is the line that
+a, minimiCes the s1m of the s=1ared differences between the act1al D6 5al1es and the !redicted
D6 5al1es$
+b, minimiCes the s1m of the s=1ared resid1als between the act1al #ield and the !redicted #ield$
+c, minimiCes the s1m of the s=1ared differences between the act1al #ield and the !redicted D6$
+d, minimiCes the s1m of the s=1ared resid1als between the act1al D6 reading and the !redicted
D6 reading$
+e, minimiCes the total 5ariation in the data$
6. /hich of the following is correct2
+a, Bf the D6 reading increases b# 1 Dobson 1nit the #ield is e"!ected to increase b# *$*('3 Eg$
+b, Bf the #ield increases b# 1 Eg the D6 reading is e"!ected to decline b# *$*('3 Dobson 1nits$
+c, The estimated #ield is 3$98 Eg when the D6 reading is * Dobson 1nits$
+d, The !redicted #ield is ($3 Eg when the D6 reading is 2* Dobson 1nits$
+e, None of these
7. The following are resistant:
+a, 9east<s=1ares regression line +b, ;orrelation coefficient
+c, 7oth +a, and +b, +d, Neither +a, nor +b, +e, Bt de!ends
8. ?r$ Nerdl# asEed the st1dents in his AP Statistics class to re!ort their o5erall grade !oint a5erages
and their SAT ?ath scores$ The scatter!lot below !ro5ides information abo1t his st1dents8 data$
The darE line is the least<s=1ares regression line for the data and its e=1ation is
x y (2 $ 1*9 (. $ 24* + =
$
/hich of the following statements abo1t the highlighted !oint is FA9S:2
+a, This st1dent has a grade !oint a5erage of 2$9 and an SAT ?ath score of '4*$
+b, Bf we 1sed the least<s=1ares line to !redict this st1dent8s SAT
?ath score we wo1ld maEe a !rediction that is too low$
+c, This st1dent8s resid1al is F82$23$
+d, Gemo5ing this data !oint wo1ld ca1se the correlation
coefficient to increase$
+e, Gemo5ing this st1dent8s data !oint wo1ld increase the slo!e of
the least<s=1ares line$
;ha!ter 3 2 Test 3D
Part 2. Free Response
Answer completely, but be concise. Show your work.
7eca1se elderl# !eo!le ma# ha5e diffic1lt# standing to ha5e their heights meas1red a st1d# looEed at
!redicting o5erall height from height to the Enee$ Here are data +in centimeters, for fi5e elderl# men:
Inee height x 57.7 47.4 43.5 44.8 55.2
Height y 192.1 153.3 146.4 162.7 169.1
9. ;onstr1ct a scatter!lot on #o1r calc1lator$ Describe the form direction and strength of the
relationshi! that #o1 see$
1. Dse #o1r calc1lator to determine the least<s=1ares regression line$ /rite the e=1ation below$ 7e s1re
to define an# 5ariables #o1 1se$
11. Bnter!ret the slo!e of the regression line in the conte"t of the !roblem$
12. Bnter!ret the 5al1e of
2
r
in the conte"t of the !roblem$
13. /e define the standard de5iation of the resid1als s, b#
2
resid1als
2
s
n
=

$ ;alc1late s and inter!ret

its 5al1e in the conte"t of the !roblem$
14. Sho1ld #o1 1se #o1r regression line from J1estion 1* to !redict the height of an elderl# man
whose Enee height is 4* centimeters2 Bf so do it$ Bf not e"!lain wh# not$
;ha!ter 3 3 Test 3D
15. Keologists collected data on the d1ration of er1!tions of the >ld Faithf1l ge#ser and the s1bse=1ent
inter5als between er1!tions$ The scatter!lot below is constr1cted with d1ration time of the
!re5io1s er1!tion as the e"!lanator# 5ariable and time 1ntil the ne"t er1!tion as the res!onse
5ariable$ Se5eral other s1mmar# statistics are !ro5ided below$
D1ration times Time between er1!tions

r & *$844
Bf #o1 ha5e @1st witnessed an er1!tion that lasted ($. min1tes 1se a least<s=1ares regression model
to estimate the time 1ntil the ne"t er1!tion$ Show #o1r method clearl#$
16. ?embers of a high school foreign<lang1age cl1b belie5e that the st1d# of a foreign lang1age
im!ro5es a st1dent8s command of :nglish$ From school records the# obtain the scores on an
:nglish achie5ement test gi5en to all seniors$ The mean score of seniors who st1died a foreign
lang1age for at least two #ears is m1ch higher than the mean score of seniors who st1died no
foreign lang1age$
+a, Bdentif# the e"!lanator# and res!onse 5ariables in this st1d#$
+b, :"!lain what l1rEing 5ariable !re5ents the concl1sion that lang1age st1d# im!ro5es st1dents8
:nglish scores$
I pledge that I have neither given nor received aid on this test LLLLLLLLLLLLLLLLLLLLLLLLLLLL
;ha!ter 3 ( Test 3D