What makes a polynomial trend line curve up or down?
 

I added a polynomial trend line with an order of 2 and it is higher in the
middle than on the ends. I did this with a different data series and it was
lower in the middle than on the ends. Can anyone tell me why it curves up or
down?

What makes a polynomial trend line curve up or down?
 

The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U shape
It's all in the math. Google on the word "parabola" to learn more
best wishes
--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in the
middle than on the ends. I did this with a different data series and it
was
lower in the middle than on the ends. Can anyone tell me why it curves up
or
down?

What makes a polynomial trend line curve up or down?
 

Both of those are straight lines, not parabolas. If you want parabolas, try
y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in the
middle than on the ends. I did this with a different data series and it
was
lower in the middle than on the ends. Can anyone tell me why it curves
up or
down?

What makes a polynomial trend line curve up or down?
 

Wow! You guys sound like you really know the subject so I appreciate your
help! Here is my situation... I am a real estate appraiser who downloaded
data from my local Multiple Listing Service into Excel. I charted two sets
of data over time. One set shows the selling prices of homes in a particular
neighborhood; the other shows the Living Areas of those homes. Linear trend
lines for both sets of data indicate a downward trend; both appear to be
decreasing at the same rate. I interpreted this to mean that, while the
sales prices were declining, this may be due (in part) to the smaller sizes
of the houses that sold. I then changed the trend lines from linear to
polynomial with an order of 2. (the default order). The trend line for the
SALES PRICES was an inverted 'u'; the trend line for the SIZE of the homes
was not inverted (looks more like a 'u'). This causes me to wonder if there
is a relationship between two polynomial trend lines that oppose one another
in the direction of their curve. I can only guess that, for the 'u' shaped
trend line, there may be more points at either end that are somewhat higher
(bigger houses) than points nearer the center of the timeline; the opposite
being true for the sales prices. Of course, I am just guessing here and it
appears that both of you have far superior statistical skills. Do you have
any additional input?

Thank you for your help,
Appraiser in Mission Viejo, CA

"David Biddulph" wrote:

Both of those are straight lines, not parabolas. If you want parabolas, try
y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in the
middle than on the ends. I did this with a different data series and it
was
lower in the middle than on the ends. Can anyone tell me why it curves
up or
down?

What makes a polynomial trend line curve up or down?
 

Probably nobody would want to hazard a guess without seeing the actual data.

- Jon
-------
Jon Peltier, Microsoft Excel MVP
Tutorials and Custom Solutions
http://PeltierTech.com
_______


"jimshoe40" wrote in message
...
Wow! You guys sound like you really know the subject so I appreciate your
help! Here is my situation... I am a real estate appraiser who downloaded
data from my local Multiple Listing Service into Excel. I charted two
sets
of data over time. One set shows the selling prices of homes in a
particular
neighborhood; the other shows the Living Areas of those homes. Linear
trend
lines for both sets of data indicate a downward trend; both appear to be
decreasing at the same rate. I interpreted this to mean that, while the
sales prices were declining, this may be due (in part) to the smaller
sizes
of the houses that sold. I then changed the trend lines from linear to
polynomial with an order of 2. (the default order). The trend line for
the
SALES PRICES was an inverted 'u'; the trend line for the SIZE of the homes
was not inverted (looks more like a 'u'). This causes me to wonder if
there
is a relationship between two polynomial trend lines that oppose one
another
in the direction of their curve. I can only guess that, for the 'u'
shaped
trend line, there may be more points at either end that are somewhat
higher
(bigger houses) than points nearer the center of the timeline; the
opposite
being true for the sales prices. Of course, I am just guessing here and
it
appears that both of you have far superior statistical skills. Do you
have
any additional input?

Thank you for your help,
Appraiser in Mission Viejo, CA

"David Biddulph" wrote:

Both of those are straight lines, not parabolas. If you want parabolas,
try
y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in
the
middle than on the ends. I did this with a different data series and
it
was
lower in the middle than on the ends. Can anyone tell me why it
curves
up or
down?

What makes a polynomial trend line curve up or down?
 

That was a typo for y=2x^2+3 !!!!

--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email

"David Biddulph" wrote in message
...
Both of those are straight lines, not parabolas. If you want parabolas,
try y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in
the
middle than on the ends. I did this with a different data series and it
was
lower in the middle than on the ends. Can anyone tell me why it curves
up or
down?

What makes a polynomial trend line curve up or down?
 

Hi Jim,

Intuitively, there will be a relationship between living areas and prices.
If you test the living area and price for a correlation, you might get an
indication as to how much one relates to the other.

But other things can affect prices (eg distance from city centre), and user
preferences as to living areas (eg # children). So, while a relationship
might exist, it may not be a deterministic one.

Cheers

--
macropod
[MVP - Microsoft Word]


"jimshoe40" wrote in message
...
Wow! You guys sound like you really know the subject so I appreciate your
help! Here is my situation... I am a real estate appraiser who downloaded
data from my local Multiple Listing Service into Excel. I charted two
sets
of data over time. One set shows the selling prices of homes in a
particular
neighborhood; the other shows the Living Areas of those homes. Linear
trend
lines for both sets of data indicate a downward trend; both appear to be
decreasing at the same rate. I interpreted this to mean that, while the
sales prices were declining, this may be due (in part) to the smaller
sizes
of the houses that sold. I then changed the trend lines from linear to
polynomial with an order of 2. (the default order). The trend line for
the
SALES PRICES was an inverted 'u'; the trend line for the SIZE of the homes
was not inverted (looks more like a 'u'). This causes me to wonder if
there
is a relationship between two polynomial trend lines that oppose one
another
in the direction of their curve. I can only guess that, for the 'u'
shaped
trend line, there may be more points at either end that are somewhat
higher
(bigger houses) than points nearer the center of the timeline; the
opposite
being true for the sales prices. Of course, I am just guessing here and
it
appears that both of you have far superior statistical skills. Do you
have
any additional input?

Thank you for your help,
Appraiser in Mission Viejo, CA

"David Biddulph" wrote:

Both of those are straight lines, not parabolas. If you want parabolas,
try
y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in
the
middle than on the ends. I did this with a different data series and
it
was
lower in the middle than on the ends. Can anyone tell me why it
curves
up or
down?

What makes a polynomial trend line curve up or down?
 

Thank you for your reply. You may not have read my post completely. The
correlation and relation are, "Linear trend lines for both sets of data
indicate a downward trend; both appear to be decreasing at the same rate."

As an appraiser, I am aware that "other things can affect prices"; that is
why I said, "while the sales prices were declining, this may be due (in part)
to the smaller sizes of the houses that sold".

Having said that, the size of homes in this neighborhood is a large
determinant of how much a buyer is willing to pay. However, the polynomial
trend lines are drawn with opposing curves. Having tested the price against
other property characteristics such as site size, it appears that the
polynomial curve with an order of two curves up or down based upon how high
the points are at each end as compared to the middle. Points that are
relatively higher on the ends create a 'u' shaped curve; points that are
relatively low on the ends create an upside-down 'u'.

However, this is still speculation. I am welcome to comments from anyone
who is an expert in charting.

Jim Shoe

"macropod" wrote:

Hi Jim,

Intuitively, there will be a relationship between living areas and prices.
If you test the living area and price for a correlation, you might get an
indication as to how much one relates to the other.

But other things can affect prices (eg distance from city centre), and user
preferences as to living areas (eg # children). So, while a relationship
might exist, it may not be a deterministic one.

Cheers

--
macropod
[MVP - Microsoft Word]


"jimshoe40" wrote in message
...
Wow! You guys sound like you really know the subject so I appreciate your
help! Here is my situation... I am a real estate appraiser who downloaded
data from my local Multiple Listing Service into Excel. I charted two
sets
of data over time. One set shows the selling prices of homes in a
particular
neighborhood; the other shows the Living Areas of those homes. Linear
trend
lines for both sets of data indicate a downward trend; both appear to be
decreasing at the same rate. I interpreted this to mean that, while the
sales prices were declining, this may be due (in part) to the smaller
sizes
of the houses that sold. I then changed the trend lines from linear to
polynomial with an order of 2. (the default order). The trend line for
the
SALES PRICES was an inverted 'u'; the trend line for the SIZE of the homes
was not inverted (looks more like a 'u'). This causes me to wonder if
there
is a relationship between two polynomial trend lines that oppose one
another
in the direction of their curve. I can only guess that, for the 'u'
shaped
trend line, there may be more points at either end that are somewhat
higher
(bigger houses) than points nearer the center of the timeline; the
opposite
being true for the sales prices. Of course, I am just guessing here and
it
appears that both of you have far superior statistical skills. Do you
have
any additional input?

Thank you for your help,
Appraiser in Mission Viejo, CA

"David Biddulph" wrote:

Both of those are straight lines, not parabolas. If you want parabolas,
try
y = 2*x^2+3 or y = -2*x^2+3
--
David Biddulph

"Bernard Liengme" wrote in message
...
The curve y = 2x+3 is U shaped, the curve y = -2x+3 has an inverted U
shape
It's all in the math. Google on the word "parabola" to learn more
best wishes

"jimshoe40" wrote in message
...
I added a polynomial trend line with an order of 2 and it is higher in
the
middle than on the ends. I did this with a different data series and
it
was
lower in the middle than on the ends. Can anyone tell me why it
curves
up or
down?