Documente Academic
Documente Profesional
Documente Cultură
A Case Study On
Spending
QT - I
&
Student
Achievement
Submited To : Prof. Mahesh K C
TABLE OF CONTENTS
Case study data____________________________________________________________________3
Question 1_________________________________________________________________________4
Graph Presentation for spending per Pupil________________________________________4
Graph Presentation for Composite Score__________________________________________5
Numerical Summary______________________________________________________________6
Question 2_________________________________________________________________________7
Simple linear regression:_________________________________________________________8
Course:
Question 3_________________________________________________________________________9
Question 4________________________________________________________________________10
Graph Explanation______________________________________________________________10
Question 5________________________________________________________________________12
Question 6________________________________________________________________________13
Question 6
2|Page
State
Composite Score
Louisiana
4,049
581
Mississippi
3,423
582
California
4,917
580
Hawaii
5,532
580
South Carolina
4,304
603
Alabama
3,777
604
Georgia
4,663
611
Florida
4,934
611
New Mexico
4,097
614
Arkansas
4,060
615
Delaware
6,208
615
Tennessee
3,800
618
Arizona
4,041
618
West Virginia
5,247
625
Maryland
6,100
625
Kentucky
5,020
626
Texas
4,520
627
New York
8,162
628
North Carolina
4,521
629
Rhode Island
6,554
638
Washington
5,338
639
Missouri
4,483
641
Colorado
4,772
644
Indiana
5,128
649
Utah
3,280
650
Wyoming
5,515
657
Connecticut
7,629
657
Massachusetts
6,413
658
Nebraska
5,410
660
Minnesota
5,477
661
Iowa
5,060
665
Montana
4,985
667
Wisconsin
6,055
667
North Dakota
4,374
671
Maine
5,561
675
3|Page
QUESTION 1
DEVELOP NUMERICAL GRAPHICAL SUMMARIES OF THE DATA
GRAPH PRESENTATION FOR SPENDING PER PUPIL
NUMERICAL SUMMARY
Name
Mean
Median
Mode
Largest
Smallest
Skewness
4|Page
Value
Spending per
Composite
Pupil ($)
Score
5068.83
631.17
4985
628
N/A
580
8162
675
3280
580
0.93
-0.28
QUESTION 2
USE REGRESSION ANALYSIS TO INVESTIGATE THE RELATIONSHIP BETWEEN
THE AMOUNTS SPENT PER PUPIL AND THE COMPOSITE SCORE ON THE NAEP
TEST. DISCUSS YOUR FINDINGS.
Spending
Composite
pre pupil
X
4049
3423
4917
5532
4304
3777
4663
4934
4097
4060
6208
3800
4041
5247
6100
5020
4520
8162
4521
6554
5338
4483
4772
5128
3280
5515
7629
6413
5410
5477
5060
4985
6055
4374
5561
177409
test score
Y
581
582
580
580
603
604
611
611
614
615
615
618
618
625
625
626
627
628
629
638
639
641
644
649
650
657
657
658
660
661
665
667
667
671
675
22091
5|Page
xi - x
-1019.83
-1645.83
-151.83
463.17
-764.83
-1291.83
-405.83
-134.83
-971.83
-1008.83
1139.17
-1268.83
-1027.83
178.17
1031.17
-48.83
-548.83
3093.17
-547.83
1485.17
269.17
-585.83
-296.83
59.17
-1788.83
446.17
2560.17
1344.17
341.17
408.17
-8.83
-83.83
986.17
-694.83
492.17
yi - y
-50.17
-49.17
-51.17
-51.17
-28.17
-27.17
-20.17
-20.17
-17.17
-16.17
-16.17
-13.17
-13.17
-6.17
-6.17
-5.17
-4.17
-3.17
-2.17
6.83
7.83
9.83
12.83
17.83
18.83
25.83
25.83
26.83
28.83
29.83
33.83
35.83
35.83
39.83
43.83
(xi - x)
(xi - x)
*(yi - y)
1040050.32
2708751.69
23051.92
214527.77
584962.74
1668821.06
164696.83
18178.74
944450.77
1017735.09
1297711.54
1609925.94
1056431.57
31745.06
1063314.52
2384.23
301212.80
9567709.49
300116.14
2205734.17
72453.26
343195.12
88107.20
3501.26
3199907.66
199068.94
6554477.74
1806796.83
116397.94
166603.92
77.94
7027.23
972534.09
482786.74
242232.72
40076680.97
51166.26
80927.74
7769.28
-23701.14
21546.31
35100.83
8186.14
2719.68
16687.68
16314.20
-18422.03
16712.28
13537.97
-1099.57
-6363.80
252.51
2289.40
-9809.77
1189.57
10141.60
2107.23
-5757.86
-3807.89
1054.94
-33681.09
11523.97
66125.57
36062.20
9835.48
12175.17
-298.66
-3003.46
35333.11
-27674.03
21571.17
346711.03
yi^
yi^
(yi^ )
(yi - )
622.35
616.93
629.86
635.18
624.55
620.00
627.66
630.01
622.76
622.44
641.03
620.19
622.28
632.71
640.09
630.75
626.42
657.93
626.43
644.02
633.50
626.10
628.60
631.68
615.70
635.03
653.32
642.80
634.12
634.70
631.10
630.45
639.70
625.16
635.43
-8.82
-14.24
-1.31
4.01
-6.62
-11.18
-3.51
-1.17
-8.41
-8.73
9.86
-10.98
-8.89
1.54
8.92
-0.42
-4.75
26.76
-4.74
12.85
2.33
-5.07
-2.57
0.51
-15.48
3.86
22.15
11.63
2.95
3.53
-0.08
-0.73
8.53
-6.01
4.26
77.84
202.73
1.73
16.06
43.78
124.90
12.33
1.36
70.69
76.17
97.12
120.49
79.07
2.38
79.58
0.18
22.54
716.08
22.46
165.08
5.42
25.69
6.59
0.26
239.49
14.90
490.56
135.23
8.71
12.47
0.01
0.53
72.79
36.13
18.13
2999.46
2517.17
2417.83
2618.52
2618.52
793.63
738.29
406.89
406.89
294.86
261.52
261.52
173.49
173.49
38.09
38.09
26.74
17.40
10.06
4.72
46.63
61.29
96.60
164.57
317.86
354.52
667.12
667.12
719.77
831.09
889.74
1144.37
1283.69
1283.69
1586.32
1920.94
25852.97
6|Page
35
5068.828571
631.1714286
b1
0.008651191
b0
587.3200236
Y^
587.32 + 0.008651 x
SSR
2999.463389
SST
25852.97143
0.116020064
0.34061718
QUESTION 3
DO YOU THINK THAT THE ESTIMATED REGRESSION EQUATION DEVELOPED
FOR THESE DATA COULD BE USED TO ESTIMATE THE COMPOSITE TEST
SCORES FOR THE STATES THAT DID NOT PARTICIPATE IN THE NAEP
PROGRAM?
8|Page
QUESTION 4
SUPPOSE THAT YOU ONLY CONSIDERED STATES THAT SPEND AT LEAST $4000
PER PUPIL BUT NOT MORE THAN $6000 PER PUPIL. FOR THESE STATES, DOES
THE RELATIONSHIP BETWEEN THE TWO VARIABLES APPEAR TO BE ANY
DIFFERENT THAN FOR THE COMPLETED DATA SET? DISCUSS THE RESULTS OF
YOUR FINDINGS AND WHETHER YOU THINK DELETING STATES WITH
SPENDING LESS THAN $4000 PER YEAR AND MORE THAN $6000 PER PUPIL IS
APPROPRIATE.
GRAPH EXPLANATION
9|Page
With regression line of the whole data and of the data only
for spending between $4000 and $6000, the difference in
the values of their coefficient determination is just about
0.4
The value of the r2 for the whole data is 0.116 and that for
the selected range is 0.156, thus in any case r2 remains less
than 0.5 and hence the regression model is a poor fit in
both the cases
Thus deleting states with spending less than $4000 per
year and more than $6000 per pupil does not make any
difference
Also if talk about the correlation coefficient r, r = 0.34 for
the whole data and r = 0.39 for the selected data
We know that if r = 0.25 then there exist highly negative
correlation
In both the cases, r is very near to 0.25 and thus we can
conclude that the two data are highly negatively correlated
10 | P a g e
QUESTION 5
DEVELOP ESTIMATES OF THE COMPOSITE TEST SCORES FOR THE STATES
THAT DID NOT PARTICIPATE IN THE NAEP PROGRAM.
State
Spending per
pupil
estimates of y^
Idaho
3602
618.48
South Dakota
4067
622.50
Oklahoma
4265
624.22
Nevada
4658
627.62
Kansas
5164
631.99
Illinois
5297
633.15
New
Hampshire
5387
633.92
Ohio
5438
634.37
Oregon
5588
635.66
Vermont
6269
641.55
Michigan
6391
642.61
Pennsylvania
6579
644.24
Alaska
7890
655.58
11 | P a g e
QUESTION 6
BASED UPON YOUR ANALYSES DO YOU THINK THAT THE EDUCATIONAL
ACHIEVEMENT LEVEL OF STUDENTS IS RELATED TO HOW MUCH THE
STATE SPENDS ON EDUCATION?
Since r2 < 0.5, this regression model is poor fit with the
data. Also since our r = 0.34 (is more near to 0.25 and
lesser to 1) we can say that the two data are highly
negatively correlate. We also found that, since the data are
positive, as the value of X (spending per pupil) increases,
y^ also increases. Thus, based upon our analysis, we can
conclude that the educational achievement of the students
is not related to how much the state in which they reside
spends on education
12 | P a g e