Hi,

I am just stuck with a statistical problem. I have series of data and
I would like to see if the average values are showing a decreasing or
increasing tendency, and if so, by how much?


Look forward o hear from you at your convenience. Pls find the data
below

Note: I am using Excel to produce the report

Thanks


June 1 673.8893333
2 697.5722733
3 703.8826556
4 712.5871953
5 738.3132375
6 753.5272175
7 763.9428142
8 745.7208142
9 737.0349564
10 730.5522814
11 709.3650892
12 703.5853458
13 698.3836714
14 689.0700975
15 656.0660975
16 644.9572214
17 636.6490781
18 632.6318496
19 639.4853062
20 636.1294823
21 629.0323298
22 605.6686631
23 594.0867506
24 574.4050339
25 558.4307839
26 545.6389
27 537.4606361
28 534.2531317
29 510.798465
30 506.4449328

RubaiyetSabbir -

"I would like to see if the average values ..." <

An average of how many values?

The first seven values are generally increasing, and the last twenty-four
are generally decreasing.

You may receive more advice if you describe your "statistical problem" in
more detail.

- Mike
www.MikeMiddleton.com


wrote in message
oups.com...
Hi,

I am just stuck with a statistical problem. I have series of data and
I would like to see if the average values are showing a decreasing or
increasing tendency, and if so, by how much?


Look forward o hear from you at your convenience. Pls find the data
below

Note: I am using Excel to produce the report

Thanks


June 1 673.8893333
2 697.5722733
3 703.8826556
4 712.5871953
5 738.3132375
6 753.5272175
7 763.9428142
8 745.7208142
9 737.0349564
10 730.5522814
11 709.3650892
12 703.5853458
13 698.3836714
14 689.0700975
15 656.0660975
16 644.9572214
17 636.6490781
18 632.6318496
19 639.4853062
20 636.1294823
21 629.0323298
22 605.6686631
23 594.0867506
24 574.4050339
25 558.4307839
26 545.6389
27 537.4606361
28 534.2531317
29 510.798465
30 506.4449328

Excel doesn't have an ARIMA model constructor, let alone a constructor
that decides what the best model for you is. (There are dozens of ways
to take this data and extrapolate it into something - based upon
assumptions you make about the data - some observational - some
procedural.)

You must create a model and compute the parameters using code.

If you're looking for something you can justify because it's easy to
explain to others, assign the estimate for the next observation the
avergage of the last three.