Question

Use the data below to answer this questions.

Period	Employment
Jan-01	1,879.50
Feb-01	1,901.00
Mar-01	1,925.30
Apr-01	1,914.60
May-01	1,961.50
Jun-01	1,960.60
Jul-01	1,953.40
Aug-01	1,940.20
Sep-01	1,928.00
Oct-01	1,909.20
Nov-01	1,896.40
Dec-01	1,881.40
Jan-02	1,880.20
Feb-02	1,884.00
Mar-02	1,902.60
Apr-02	1,913.40
May-02	1,937.40
Jun-02	1,990.90
Jul-02	1,994.80
Aug-02	2,013.10
Sep-02	2,002.30
Oct-02	1,982.50
Nov-02	1,969.00
Dec-02	1,959.20
Jan-03	1,928.20
Feb-03	1,952.40
Mar-03	1,980.40
Apr-03	1,972.00
May-03	1,987.80
Jun-03	2,018.70
Jul-03	2,027.80
Aug-03	2,030.20
Sep-03	2,012.20
Oct-03	2,032.30
Nov-03	2,008.30
Dec-03	2,026.80
Jan-04	1,982.30
Feb-04	1,967.20
Mar-04	1,981.60
Apr-04	1,977.40
May-04	2,020.50
Jun-04	2,074.90
Jul-04	2,069.50
Aug-04	2,064.10
Sep-04	2,050.80
Oct-04	2,057.30
Nov-04	2,037.30
Dec-04	2,056.50
Jan-05	2,017.60
Feb-05	2,030.60
Mar-05	2,039.10
Apr-05	2,051.70
May-05	2,097.70
Jun-05	2,115.90
Jul-05	2,120.40
Aug-05	2,128.80
Sep-05	2,103.10
Oct-05	2,122.40
Nov-05	2,124.50
Dec-05	2,124.60

a) Generate a scatter of the data

b) Report the monthly averages (January for all years, February for all years etc.)

c) Is there seasonality? Is there a trend?

d) How can you forecast the value for January 2006? Generate that forecast.

e.) (Not technical) This forecast will for sure be wrong. Why?

Answer 1

1. SCATTER PLOT OF THE DATA

(2) Monthly Averages

Variable N N* Mean
janua 5 0 1937.6
februa 5 0 1947.0
march 5 0 1965.8
april 5 0 1965.8
may 5 0 2001.0
june 5 0 2032.2
july 5 0 2033.2
august 5 0 2035.3
sept 5 0 2019.3
oct 5 0 2020.7
nov 5 0 2007.1
dec 5 0 2009.7

(3)

Fitted Trend Equation

Yt = 1888.14 + 3.598×t

Accuracy Measures

MAPE 1.355
MAD 26.939
MSD 990.441

(4) Forecast for January 2006

Forecasts

Period Forecast
61 2107.64

(5)The forecasting will sure be wrong as we do not check the autocorrelation before starting.