How do I predict the future? Stats people needed.

[Ziggy12]

HI
I have a series of 5 totals for each year, the slope function over the 5 totals returns 57.6

How do I apply the slope to a subset of the totals
The total are 2106, 2008, 2147,2288. slope = 57.6

My subset is
131,158,138,132,547. I want to predict the next number ignoring the first 3 because something changed in the subset but the over all totals haven't I can expect the same pattern in the subset as the totals.

Maybe there is some other way of doing this?

Any stats people out there?

cheers
Ziggy12

 

[JustynaMK]

Hi Ziggy,

Not sure if I understood your request correctly - if we take your original dataset (131,158,138,132,547) and ignore the first 3 values, we're going to end up with only 2 values: 132,547. Is that what you want to analyse?

One approach might be to use =SLOPE function (assuming that your function is linear), with the result of 415. It means that your next value should equal 962 (547+415).
Another solution might be to use one of =FORECAST() or =GROWTH() function. =FORECAST is used for linear functions, =GROWTH should be applied for exponential models.
=FORECAST() will also give you a result of 962
=GROWTH(), on the other hand, will show the value of 2266.735

Does it help?

P.S. On a side note, it looks like your slope of 57.6 is calculated for the following 5 values: 2106, 2008, 2147, 2288, 2254 (the last value was missing from your post).

 

[Ziggy12]

Thanks for replying.
I want to predict the next number in my subset based on the slope of the main dataset. I think my subset will now grow at the same rate as the main dataset.
I can't predict off the subset dataset as the jump to 547 won't be repeated and now 547 is the new start point norm.

Hope that makes sense.

cheers
Ziggy

 

[JustynaMK]

Hi Ziggy,

Sorry if I'm missing something here. If you assume that your subset will now grow at the same rate as the main dataset (i.e. slope = 57.6), cannot you simply add the slope value to the latest value of your subset (i.e. 547 + 57.6) in order to predict the next future value? After you obtain the new "actual" value, you can re-calculate your slope and adjust your forecast values (this is all assuming that your dataset is based on a linear function).

Hopefully someone else on the forum will also provide some additional input.