Anda di halaman 1dari 3

From:

To:
Date:
Subject:

Fis Malsori
Prof. Barb Downey
April 18th, 2016
University of Missouri Fall 2016 Enrollment Forecast

Introduction
This memo serves the purpose of presenting an enrollment estimate for the Fall 2016 admissions process at the
University of Missouri. Additionally, this memo will recommend a forecasting technique to be used by the
admissions team in future estimation of enrollment.
Analysis
The data provided by Mizzou Magazine showed a positive trend and relationship between time & enrollment, that
is, enrollment constantly increased with time (exhibit 1.0). However, the time increments in the dataset were old
and irregular starting from the year 1948 to 2003, when the enrollment counts became available on a year by year
basis. Therefore, a decision to only keep the last 12 years of data was made.
First, after plotting the data points a nearly perfect positive linear relationship (R2=.9778) between time and
enrollment was observed. Hence, a time series forecast of was performed both in two and three year moving
averages as well as weighted moving averages of the same period structure. All four of these forecasting techniques
yielded a slight decrease in enrollment, however the least subjective to error being the weighted two year moving
average with a Mean Absolute Deviation of 1,003 which resulted in a forecast of roughly two students less than
last years actual. When the weighted moving average was performed, 70% of the weight was assigned to recent
periods and the remaining percentage was distributed to earlier periods (see exhibit 2.0).
Second, to resolve the issue of delay in forecast that the above mentioned method presents, Exponential Smoothing
was found to be significantly less subject to error and delay. The Mean Absolut Deviation of this technique
(MAD=880) was significantly lower than any of the moving averages, and thus proved to be more appropriate in
forecasting enrollment. Again, the weight of the Most recent Actual value was twice that of the previous forecast,
and thus yielded a forecast of 35,386 students to be enrolled at MU next fall which indicates a decrease of 62
students from last years actual enrollment of 35,448 (See exhibit 3.0).
Finally, a regression model was found to be appropriate since the models as well as the dependent variables pvalues were statistically significant at =0.05. The regression model yielded the most accurate estimate, with an R2
of 0.9778 and Mean Absolute Deviation of only 358. The model predicts that in Fall 2016, the University of
Missouri should expect 36,945 students to enroll, roughly 1500 more students than last years actual enrollment.
This estimate is significantly higher than any of the previously used methods, however, given that it has the least
exposure to error and delay, it should be used in priority of any other method used in this analysis (exhibit 3.0 &3.2).
Recommendation
After analyzing the results of this forecasting effort, I recommend that a simple or multiple regression model be
used to forecast future enrollment, given its very low error exposure as well as absence of delay. Researching for
other variables closely related to enrollment that could improve the estimation of the regression model should also
be considered.

Malsori Capital
Tel 573 554 6669
Fax 573 554 6668

1700 Caniff Cir


Columbia MO 65201

www.malesoricapital.com
fkm6x2@mail.missouri.edu

APPENDIX
Exhibit 1.0

FALL ENROLLMENT
1 9 4 8 - 2 0 15
40,000

y = 348.08x - 668179
R = 0.9024

35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
1940

1950

1960

1970

1980

1990

2000

2010

2020

Exhibit 2.0
Historical Data
Year

Moving Average (2 Period)

Fall Enrollment Forecast

Error

Abs Value

Moving Average (3 Period)

Error^2

Forecast

Error

Abs Value

Weighted Moving Average (2 Period)

Error^2

Forecast

Error

Abs Value

Error^2

Weighted Moving Average (3 Period)


Forecast

Error

Abs Value

Error^2

2003

26,805

2004

27,003

2005

27,985

26,904 -1,081

2006

28,253

27,494

-759

759

576,081 27,264

-989

989

977,462 27,690

-563

563

316,519 27,671

-582

582

339,190

2007

28,472

28,119

-353

353

124,609 27,747

-725

725

525,625 28,173

-299

299

89,640 28,074

-398

398

158,086

2008

30,200

28,363 -1,838

1,838 3,376,406 28,237 -1,963

1,963 3,854,678 28,406 -1,794

1,794 3,217,360 28,380 -1,821

1,821 3,314,220

2009

31,314

29,336 -1,978

1,978 3,912,484 28,975 -2,339

2,339 5,470,921 29,682 -1,632

1,632 2,664,730 29,660 -1,654

1,654 2,736,708

2010

32,009

30,757 -1,252

1,252 1,567,504 29,995 -2,014

2,014 4,054,853 30,980 -1,029

1,029 1,059,253 30,807 -1,202

1,202 1,444,804

2011

33,318

31,662 -1,657

1,657 2,743,992 31,174 -2,144

2,144 4,595,307 31,801 -1,518

1,518 2,302,806 31,689 -1,629

1,629 2,653,315

2012

34,255

32,664 -1,592

1,592 2,532,872 32,214 -2,041

2,041 4,167,042 32,925 -1,330

1,330 1,768,102 32,856 -1,399

1,399 1,957,761

2013

34,658

33,787

-872

872

759,512 33,194 -1,464

1,464 2,143,296 33,974

-684

684

467,993 33,843

-815

815

664,225

2014

35,441

34,457

-985

985

969,240 34,077 -1,364

1,364 1,860,496 34,537

-904

904

817,035 34,443

-998

998

995,206

2015

35,448

35,050

-399

399

158,802 34,785

-242

242

58,516 35,166

-282

282

79,637

2016

1,081 1,168,561

35,445

26,944 -1,041

-663

663

440,011 35,206

35,182

1,041 1,084,514

35,446

35,368

MA2

MA3

WMA2

MAD

1,160

1,571

1,003

1,078

MSE

1,626,370

2,808,969

1,258,770

1,434,315

Malsori Capital
Tel 573 554 6669
Fax 573 554 6668

1700 Caniff Cir


Columbia MO 65201

www.malesoricapital.com
fkm6x2@mail.missouri.edu

WMA3

Exhibit 3.0
Exponential Smoothing (Simple)
Forecast
Error
Abs Value
Error^2
Forecast
26,805
0
0
0
26,213
26,805
-198
198
39,204
27,039
26,954
-1,032
1,032
1,063,992
27,864
27,727
-526
526
276,545
28,690
28,122
-350
350
122,828
29,515
28,384
-1,816
1,816
3,296,466
30,341
29,746
-1,568
1,568
2,458,324
31,166
30,922
-1,087
1,087
1,181,517
31,992
31,737
-1,581
1,581
2,498,752
32,817
32,923
-1,332
1,332
1,774,720
33,643
33,922
-736
736
541,764
34,468
34,474
-967
967
935,111
35,294
35,199
-249
249
61,878
36,119
35,386
36,945
ES
MAD
880 MAD
MSE
1,096,239 MSE

Year
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016

Regression
Error
Abs Value
-592
592
36
36
-121
121
437
437
1,043
1,043
141
141
-148
148
-17
17
-501
501
-612
612
-190
190
-147
147
671
671

Error^2
350,425
1,265
14,617
190,647
1,088,193
19,796
21,836
297
250,704
374,753
35,962
21,640
450,816
Regression
358
216,996

Exhibit 3.2

FALL ENROLLMENT
39,000
37,000
35,000
33,000
31,000

2003-2016

Weights
Old
0.10
Mid
0.20
Recent
0.70
Sum of Weights
1.00
Exponential Smoothing
Alpha (Recent)
0.75
Regression Equation
Intercept
1,627,329.5
Year
825.5

29,000

27,985
26,805

28,253

y = 825.53x - 2E+06
R = 0.9778
35,441
34,255

35,448

34,658

33,318
32,009
31,314
30,200

28,472

27,003

27,000
25,000
2002

2004

2006

2008

2010

2012

Malsori Capital
Tel 573 554 6669
Fax 573 554 6668

1700 Caniff Cir


Columbia MO 65201

www.malesoricapital.com
fkm6x2@mail.missouri.edu

2014

2016

Anda mungkin juga menyukai