neap 2012 core

[sam0001]

hey guys, how do you do this question?

[StumbleBum]

I'm not sure about this, so I could be wrong. So if someone could confirm this or not then that would be good.

16 - 24.5 - 22.3 - 31.4

We take the three point moving mean of the top three values to estimate the next point.

So (24.5+22.3+31.4)/3 = 26.1

Now we add that into the original data:

16 - 24.5 - 22.3 - 31.4 - 26.6

And we take the three point moving mean of the top three values again.

So (22.3+31.4+26.1)/3 = 26.6

26.1 and 26.6 respectively, so the answer is C?

[Stick]

One of the weirdest questions I've ever come across. :S StumbleBum's method does make sense, however, so I believe that would be the best way to approach it. That being said, I'm led to think this question doesn't even fit in with the study design, because smoothing time series data is only done for two reasons:

1. To identify an underlying trend
2. To make a regression line more accurate

This question deals with neither of the above, which seriously makes me further doubt the credibility of NEAP papers.

[StumbleBum]

One of the weirdest questions I've ever come across. :S StumbleBum's method does make sense, however, so I believe that would be the best way to approach it. That being said, I'm led to think this question doesn't even fit in with the study design, because smoothing time series data is only done for two reasons:

1. To identify an underlying trend
2. To make a regression line more accurate

This question deals with neither of the above, which seriously makes me further doubt the credibility of NEAP papers.

I totally agree, what I did to estimate the next points in the data follows nothing that is taught within the course. That is simply the only way that I yielded two semi-logical results. Yet I, even after working that out, don't see at all how that is an accurate or correct way to estimate data; even for further which we seem to estimate a lot in.

I got this from the current study design:

• Median smoothing (as a graphical technique) and smoothing using a moving average with
  consideration of the number of terms required and centring where required;
• The terms used to describe standard patterns in time series in qualitative terms, the role of smoothing
  and deseasonalisation in helping to identify these patterns, and some simple techniques for
  quantifying these patterns and enabling forecasting of future values;
• Forecasting using trend lines (with the data deseasonalised where necessary).

These are the only points I found in the study design for Further in relation to smoothing data. This suggests to me that VCAA can only ask you to perform Median, Median (graphically) and Mean smoothing. While also then commenting on the patterns and trends that the original data or smoothed data now reveal. With the only forecasting or estimating that we do, being in relation too the trend lines that we fit to data. That's my take on it anyway.