The document summarizes the results of a regression analysis conducted to examine the impact of various housing characteristics on the sale price of 1950s-era homes in Louisville, KY from 2001-2009. The regression identified several key characteristics that influenced sale price, including lot size, number of stories, finished basement area, number of bathrooms, and neighborhood location. It was found that having a finished basement added value, but returns diminished after the basement was over 3/4 finished. Neighborhood was also a major determinant of home value. The regression model explained 69% of the variation in home sale prices.
Ray Lucas • Lincoln Financial Advisors Corporation
- Focused on the reality of the market by Linda Ferentchak
- Choppy earnings season draws to a close
- Using an "endowment" investment model (Greg Gann, LPL Financial)
Ray Lucas • Lincoln Financial Advisors Corporation
- Focused on the reality of the market by Linda Ferentchak
- Choppy earnings season draws to a close
- Using an "endowment" investment model (Greg Gann, LPL Financial)
Since becoming actively involved in improving the quality-of-life, public services and homeownership rate in its own West Philadelphia neighborhood, the University of Pennsylvania has seen the performance of this housing stock outpace that of Philadelphia as a whole.
This white paper describes the analysis and models developed to predict house prices in King County. Further, the models are compared and most effective and simple model is recommended.
The objective of the project was to develop an analytical model to predict the house prices at King County. This whitepaper describes in detail the data preprocessing, predictive models developed, recommendations and future plans for improvement.
In this keynote I will give you a business understanding of ML by going through key concepts and concrete use cases that illustrate its possibilities. I'll present new technology that makes ML more accessible, and I'll explain in simple terms the limitations to what can be achieved. Finally, I'll discuss pragmatic considerations of real-world applications and I'll give a sneak peak at the Machine Learning Canvas — a framework for describing a predictive system that uses ML to provide value to its end user.
Our aim was to develop algorithms which use a broad spectrum of features to predict real prices. Algorithm applications rely on a rich dataset that includes housing data and macroeconomic patterns. An accurate forecasting model will allow Sber bank to provide more certainty to their customers in an uncertain economy.
A Study under Prof. Metin Cakanyildirim to understand the various factors involved in pricing of house and perform a Regression Analysis to understand their impact.
For the last number of decades, the real estate market has been broken down into seasons with spring reigning as the best time to sell a home. Traditionally, that’s how it’s been. But there’s a big shift happening now.
Recent years have seen that seasonality blur as more and more people decide to buy or sell a home no matter what time of year it is.
What we do know is that while we’ll probably see more homes hit the market this spring, supply is still too low to keep up with demand.
So, even if more homes do come on the market compared to previous months, there are plenty of willing and ready buyers waiting on the sidelines to scoop them up.
Full story at: www.westernmahomes.net
The presence of a natural gas pipeline does not affect the value of the surrounding property. Integra Realty Resources, a leading provider of real estate valuation and counseling services, conducted a rigorous study of properties in four separate areas of the country in 2015. The report, Pipeline Impact to Property Value and Property Insurability, prepared on behalf of the INGAA Foundation, shows that the presence of pipelines does not affect the value of a property, its insurability, its desirability or the ability to obtain a mortgage.
Qnt. 5040 – Mini Report #1RegressionsDr. Phillip S. RokickiM.docxamrit47
Qnt. 5040 – Mini Report #1
Regressions
Dr. Phillip S. Rokicki
Maximum Points: 5
Excel File Needed: The Prescott Housing Study
The Prescott County Housing Problem
Introduction:
The Prescott county mayor, Robert (Pete) Smith has been worried for some time that housing values in the county have been declining. Pete said to the county commission recently,
“Our housing stock is getting so old and tired, and I’m afraid that if we don’t start building new homes that our children will just move away to Orlando or even to, heaven forbid, to South Florida. I think that we need to study this situation, and do something about it right now!”
What Pete did not say, but each of the commissioners knew, was that his brother-in-law, Bo Bradley is a developer who wants the commission to rezone 350 acres in the north county for a new development. This land is currently envisioned to be a county park, but old Bo want to develop it. Bo said recently, “What this county needs is my development, not some old park for the deer.” Bo it seems is interested in making money more than he is in protecting undeveloped land.
In order to get this study going Pete has asked you to look at some recent sales of homes in the county to understand what is going on with the housing stock, and then to project out what kind of values that five typical housing could bring. What he is hoping is that the values will be so low that the commission will want to rezone those 350 acres for Bo’s development.
Using the Excel data file that has been provided you are to completely answer the following questions:
1. What is the current status of the housing stock in the county?
a. To do this you will create a one-variable summary using StatTools and analyze the age of the recently sold homes, their average price sold, the number of bedrooms, bathrooms and number of cars that can be garaged.
b. What does the skewness and kurtosis tell you about these data?
c. Would it be better to use the Interquartile range to analyze this data (not a yes or no answer) and if so why, or why not?
2. Doing two Q-Q plots, do you consider the data for price and square footage to be normal or not, and why?
3. Doing a correlation in StatTools and using all six of the variables, how are each of these variables correlated to each other. Again be specific.
4. Doing a scatterplot of price versus square footage and adding a trend line to the plot, what does this tell you about the data?
a. Now do a scatterplot of price versus age and adding a trend line, what does this tell you about the question of new homes versus price?
5. Next do a multiple regression using price as the dependent variable, and all other variables as independent variables:
a. Do any of the variables have a t-value that is greater than the alpha (.05) for this assignment? If so, delete them and rerun the regression and compare and contrast the old regression versus the new regression without one or more of the variables.
b. Is the F-ratio ...
Has Milwaukee\'s Riverwest neighborhood reached a condo development saturation point? What is the impact of income and job growth on the sustainability of the condo building boom in this diverse area of Milwaukee?
Since becoming actively involved in improving the quality-of-life, public services and homeownership rate in its own West Philadelphia neighborhood, the University of Pennsylvania has seen the performance of this housing stock outpace that of Philadelphia as a whole.
This white paper describes the analysis and models developed to predict house prices in King County. Further, the models are compared and most effective and simple model is recommended.
The objective of the project was to develop an analytical model to predict the house prices at King County. This whitepaper describes in detail the data preprocessing, predictive models developed, recommendations and future plans for improvement.
In this keynote I will give you a business understanding of ML by going through key concepts and concrete use cases that illustrate its possibilities. I'll present new technology that makes ML more accessible, and I'll explain in simple terms the limitations to what can be achieved. Finally, I'll discuss pragmatic considerations of real-world applications and I'll give a sneak peak at the Machine Learning Canvas — a framework for describing a predictive system that uses ML to provide value to its end user.
Our aim was to develop algorithms which use a broad spectrum of features to predict real prices. Algorithm applications rely on a rich dataset that includes housing data and macroeconomic patterns. An accurate forecasting model will allow Sber bank to provide more certainty to their customers in an uncertain economy.
A Study under Prof. Metin Cakanyildirim to understand the various factors involved in pricing of house and perform a Regression Analysis to understand their impact.
For the last number of decades, the real estate market has been broken down into seasons with spring reigning as the best time to sell a home. Traditionally, that’s how it’s been. But there’s a big shift happening now.
Recent years have seen that seasonality blur as more and more people decide to buy or sell a home no matter what time of year it is.
What we do know is that while we’ll probably see more homes hit the market this spring, supply is still too low to keep up with demand.
So, even if more homes do come on the market compared to previous months, there are plenty of willing and ready buyers waiting on the sidelines to scoop them up.
Full story at: www.westernmahomes.net
The presence of a natural gas pipeline does not affect the value of the surrounding property. Integra Realty Resources, a leading provider of real estate valuation and counseling services, conducted a rigorous study of properties in four separate areas of the country in 2015. The report, Pipeline Impact to Property Value and Property Insurability, prepared on behalf of the INGAA Foundation, shows that the presence of pipelines does not affect the value of a property, its insurability, its desirability or the ability to obtain a mortgage.
Qnt. 5040 – Mini Report #1RegressionsDr. Phillip S. RokickiM.docxamrit47
Qnt. 5040 – Mini Report #1
Regressions
Dr. Phillip S. Rokicki
Maximum Points: 5
Excel File Needed: The Prescott Housing Study
The Prescott County Housing Problem
Introduction:
The Prescott county mayor, Robert (Pete) Smith has been worried for some time that housing values in the county have been declining. Pete said to the county commission recently,
“Our housing stock is getting so old and tired, and I’m afraid that if we don’t start building new homes that our children will just move away to Orlando or even to, heaven forbid, to South Florida. I think that we need to study this situation, and do something about it right now!”
What Pete did not say, but each of the commissioners knew, was that his brother-in-law, Bo Bradley is a developer who wants the commission to rezone 350 acres in the north county for a new development. This land is currently envisioned to be a county park, but old Bo want to develop it. Bo said recently, “What this county needs is my development, not some old park for the deer.” Bo it seems is interested in making money more than he is in protecting undeveloped land.
In order to get this study going Pete has asked you to look at some recent sales of homes in the county to understand what is going on with the housing stock, and then to project out what kind of values that five typical housing could bring. What he is hoping is that the values will be so low that the commission will want to rezone those 350 acres for Bo’s development.
Using the Excel data file that has been provided you are to completely answer the following questions:
1. What is the current status of the housing stock in the county?
a. To do this you will create a one-variable summary using StatTools and analyze the age of the recently sold homes, their average price sold, the number of bedrooms, bathrooms and number of cars that can be garaged.
b. What does the skewness and kurtosis tell you about these data?
c. Would it be better to use the Interquartile range to analyze this data (not a yes or no answer) and if so why, or why not?
2. Doing two Q-Q plots, do you consider the data for price and square footage to be normal or not, and why?
3. Doing a correlation in StatTools and using all six of the variables, how are each of these variables correlated to each other. Again be specific.
4. Doing a scatterplot of price versus square footage and adding a trend line to the plot, what does this tell you about the data?
a. Now do a scatterplot of price versus age and adding a trend line, what does this tell you about the question of new homes versus price?
5. Next do a multiple regression using price as the dependent variable, and all other variables as independent variables:
a. Do any of the variables have a t-value that is greater than the alpha (.05) for this assignment? If so, delete them and rerun the regression and compare and contrast the old regression versus the new regression without one or more of the variables.
b. Is the F-ratio ...
Has Milwaukee\'s Riverwest neighborhood reached a condo development saturation point? What is the impact of income and job growth on the sustainability of the condo building boom in this diverse area of Milwaukee?
Adjusting primitives for graph : SHORT REPORT / NOTESSubhajit Sahu
Graph algorithms, like PageRank Compressed Sparse Row (CSR) is an adjacency-list based graph representation that is
Multiply with different modes (map)
1. Performance of sequential execution based vs OpenMP based vector multiply.
2. Comparing various launch configs for CUDA based vector multiply.
Sum with different storage types (reduce)
1. Performance of vector element sum using float vs bfloat16 as the storage type.
Sum with different modes (reduce)
1. Performance of sequential execution based vs OpenMP based vector element sum.
2. Performance of memcpy vs in-place based CUDA based vector element sum.
3. Comparing various launch configs for CUDA based vector element sum (memcpy).
4. Comparing various launch configs for CUDA based vector element sum (in-place).
Sum with in-place strategies of CUDA mode (reduce)
1. Comparing various launch configs for CUDA based vector element sum (in-place).
The Building Blocks of QuestDB, a Time Series Databasejavier ramirez
Talk Delivered at Valencia Codes Meetup 2024-06.
Traditionally, databases have treated timestamps just as another data type. However, when performing real-time analytics, timestamps should be first class citizens and we need rich time semantics to get the most out of our data. We also need to deal with ever growing datasets while keeping performant, which is as fun as it sounds.
It is no wonder time-series databases are now more popular than ever before. Join me in this session to learn about the internal architecture and building blocks of QuestDB, an open source time-series database designed for speed. We will also review a history of some of the changes we have gone over the past two years to deal with late and unordered data, non-blocking writes, read-replicas, or faster batch ingestion.
06-04-2024 - NYC Tech Week - Discussion on Vector Databases, Unstructured Data and AI
Discussion on Vector Databases, Unstructured Data and AI
https://www.meetup.com/unstructured-data-meetup-new-york/
This meetup is for people working in unstructured data. Speakers will come present about related topics such as vector databases, LLMs, and managing data at scale. The intended audience of this group includes roles like machine learning engineers, data scientists, data engineers, software engineers, and PMs.This meetup was formerly Milvus Meetup, and is sponsored by Zilliz maintainers of Milvus.
Data Centers - Striving Within A Narrow Range - Research Report - MCG - May 2...pchutichetpong
M Capital Group (“MCG”) expects to see demand and the changing evolution of supply, facilitated through institutional investment rotation out of offices and into work from home (“WFH”), while the ever-expanding need for data storage as global internet usage expands, with experts predicting 5.3 billion users by 2023. These market factors will be underpinned by technological changes, such as progressing cloud services and edge sites, allowing the industry to see strong expected annual growth of 13% over the next 4 years.
Whilst competitive headwinds remain, represented through the recent second bankruptcy filing of Sungard, which blames “COVID-19 and other macroeconomic trends including delayed customer spending decisions, insourcing and reductions in IT spending, energy inflation and reduction in demand for certain services”, the industry has seen key adjustments, where MCG believes that engineering cost management and technological innovation will be paramount to success.
MCG reports that the more favorable market conditions expected over the next few years, helped by the winding down of pandemic restrictions and a hybrid working environment will be driving market momentum forward. The continuous injection of capital by alternative investment firms, as well as the growing infrastructural investment from cloud service providers and social media companies, whose revenues are expected to grow over 3.6x larger by value in 2026, will likely help propel center provision and innovation. These factors paint a promising picture for the industry players that offset rising input costs and adapt to new technologies.
According to M Capital Group: “Specifically, the long-term cost-saving opportunities available from the rise of remote managing will likely aid value growth for the industry. Through margin optimization and further availability of capital for reinvestment, strong players will maintain their competitive foothold, while weaker players exit the market to balance supply and demand.”
Adjusting OpenMP PageRank : SHORT REPORT / NOTESSubhajit Sahu
For massive graphs that fit in RAM, but not in GPU memory, it is possible to take
advantage of a shared memory system with multiple CPUs, each with multiple cores, to
accelerate pagerank computation. If the NUMA architecture of the system is properly taken
into account with good vertex partitioning, the speedup can be significant. To take steps in
this direction, experiments are conducted to implement pagerank in OpenMP using two
different approaches, uniform and hybrid. The uniform approach runs all primitives required
for pagerank in OpenMP mode (with multiple threads). On the other hand, the hybrid
approach runs certain primitives in sequential mode (i.e., sumAt, multiply).
1. Pilot Study: Use of Regression to Identify, Quantify and Interpret
Property Values In Louisville, KY
Prepared for
Donna Hunt
Chief Deputy of the Jefferson County PVA
By
Margaret Maginnis
May 2010
An Analysis of Sales in 2000-2009 for
1950s Housing Stock in Jefferson County, Kentucky
12/15/2017
2. The Research Question:
1. What is the effect on sale price of 1950s housing stock in Louisville, KY when a finished basement is
added? Does it matter how much of the basement is finished? Is there a point of diminishing returns? What
is the effect of location in different neighborhoods?
Using a simple linear regression model, sale values of 1950s housing for the period 2001 through 2009 were
examined based on selected characteristics.
Pilot Study: Property Values of 1950s Housing stock
Introduction
The regression analysis identified lot size, number of stories, finished size, number of bathrooms, finished basement area,
garages and neighborhood location to be the most significant characteristics to impact a home’s sale price. We found that
having some portion of a basement finished certainly added value to the home, but that value diminished when the
basement was more than 3/4s finished. The most significant indicator of home value was often the neighborhood in which
the home was located, as the results show in the following write up.
Findings
22/15/2017
Source Data: PVA 3/25/2010
Limitations of the Data and Next Steps
Limitations of the analysis included insufficient data on housing characteristics such as square footage of porches, decks,
and garages. The lack of such detailed information prevents the model from being as robust and reliable it might be
otherwise. Next steps would be to obtain more detailed data from the REMF_CHAR database and rerun the rgressions. In
order to do this, we need to obtain a variable ‘dictionary’ of the codes used in the REMF_CHAR database. Using the
REMF_Master merged with the REMF_char data, we could rerun the regressions and compare predicted values to actual
sales values.
3. Pilot Study: Property Values of 1950s Housing stock
The initial query of parcels from the REMF Master consisted of all single-family houses - a database of
218,376 records with information on PVA assessments and sales, but no information on housing characteristics. These
records were then linked to valid sales for the period 2001 through 2009 with a resulting file of approximately 50,000
records containing detailed information on housing characteristics.
The parameters for the pilot study were 1950s-era single-family housing with full basements. The decision to
use 1950s housing was predicated on the fact that there is a large supply of homes from that era in Louisville and the
sample would be relatively homogeneous. In fact, Louisville Metro has 45,848 homes built during the 1950s. Of these,
approximately 49% (22,504) include some sort of basement, with an average size (basement) of 700 square feet. After
merging the data with valid sales, the number of homes built in the 1950s with full-sized basements comprised slightly over
4,200 records. Selecting for single-family residences with full basements, built in the 1950s, with valid sale transactions
that occurred between 2001-2009, the final study included 1,481 records.
Records were examined first in Excel, then frequencies, comparison of means and regression models were
run in SPSS. The following section highlights the exploratory phase of analysis.
Methodology
32/15/2017
Source Data: PVA 3/25/2010
4. Pilot Study: Property Values of 1950s Housing stock
Exploratory Results
Typical characteristics of 1950s housing stock include:
* quarter acre (or less) lot
* 1 to 1.5 stories
* 1 bathroom
* small front porch or stoop
* 1200 sq ft with full basement
* one-third of the basement area finished
* detached garage
42/15/2017
Source Data: PVA 3/25/2010
5. Pilot Analysis: Property Values of 1950s Housing stock
Exploratory Results
Figure 1. Location of homes included in the analysis.
Most of the 1950s housing used in this analysis was built directly inside the Watterson Expressway or just outside of it. In
the Southwest portion of the city, many homes were built in and just south of Shively. In the East End, 1950-era homes
were built inside the Watterson in the smaller incorporated cities of Kingsley and Wellington, St Mathews, Brownsboro
Village, Indian Hills, Rolling Fields, Beechwood and Woodlawn Park, and in the Louisville neighborhoods of Brownsboro-
Zorn, Clifton, Rock Creek, Gardiner Lane, Highlands-Douglas, Hawthorne, Belknap and Bowman. In addition, homes built in
the 1950s just east of the Watterson in St Regis Park, J-Town and Bon Air are included.
N = 1,481
52/15/2017
Source Data: PVA 3/25/2010
6. Pilot Study: Property Values of 1950s Housing stock
Exploratory Results
Figure 3.
Median Sale Price by Year; 1950s Houses with Full Basement.
Between 2001 and 2009, the median cost of a 1950s home rose approximately 17%.
$115,000
$119,950
$126,275
$127,500
$132,500
$135,100
$137,500
$130,070
$134,000
$100,000
$105,000
$110,000
$115,000
$120,000
$125,000
$130,000
$135,000
$140,000
2001 2002 2003 2004 2005 2006 2007 2008 2009
Median Sale Price by Year of Sale
62/15/2017
Source Data: PVA 3/25/2010
N=546
N=496
N=624
N=539
N=573
N=471
N=299
N=394
N=310
Average number of sales per year = 473
Total number of sales of 1950s housing with basements = 4,252
7. Pilot Study: Property Values of 1950s Housing stock
Exploratory Results
To look further at the data, we divided the
records into 3 categories according to the percent of
finished basement.
We then examined sale price by year based on those three
categories.
Sale Year
Homes with less
than 30% finished
basement
Homes
between 30%
and 75%
finished
basement
Homes with
more than 75%
finished
basements
2001 $117,000.00 $115,000.00 $117,500.00
2002 $118,750.00 $122,900.00 $119,000.00
2003 $127,250.00 $127,450.00 $116,000.00
2004 $123,000.00 $131,000.00 $135,200.00
2005 $130,000.00 $137,500.00 $139,000.00
2006 $132,000.00 $137,500.00 $139,830.00
2007 $134,200.00 $140,000.00 $143,750.00
2008 $123,900.00 $138,500.00 $126,320.00
2009 $129,950.00 $141,750.00 $132,400.00
Median Sale Price
Although sale prices started to fall by 2008, between 2001 and
2009, median home prices rose 23% for homes with 1/3 to 3/4
finished basement, and 13% for those with full-finished
basements (i.e., 75% or more finished).
7
Frequency Percent
1 794 53.6
2 534 36.1
3 153 10.3
Total 1481 100.0
Valid
Above Grade Price per Square Foot, 1950s Homes with Full
Finished Basements
Sale Year
Homes with less
than 30% finished
basement
Homes
between 30%
and 75%
finished
basement
Homes with
more than 75%
finished
basements
2001 $91.71 $95.17 $94.56
2002 $95.87 $102.13 $99.78
2003 $100.05 $102.82 $98.79
2004 $102.34 $106.99 $114.52
2005 $104.67 $112.24 $114.13
2006 $105.47 $111.44 $116.70
2007 $110.86 $113.70 $119.93
2008 $108.20 $115.36 $127.09
2009 $107.92 $112.32 $110.80
Average Sale Price Per Square Foot (Above Grade)
Homes with 1/3 to 3/4s of the basement finished did better in
the marketplace in 2001-2003, but this trend began to change in
2004 when finished basements fetched a higher price per square
foot (until 2009).
2/15/2017
Source Data: PVA 3/25/2010
8. Pilot Study: Property Values of 1950s Housing stock
Exploratory Results
Average Sale Price per Square Foot in 1950s Homes with Full Basements,
Varying by Percent of Basement Finished
Figure 4.
Average sale price per square foot in 1950s homes with full basements.
While sales
were higher for
houses with
75% or more of
the basement
finished, Sale
prices for 1950s
homes with a
third to three-
quarters
finished
basement rose
at a higher rate,
on average 3%
annually until
2008.
Homes with less
than 30% of the
basement
finished-the
majority of the
study sample-
rose steadily
until 2008 when
prices began to
fall.
$95
$100 $99
$115 $114
$117
$120
$127
$111
$95
$102
$103
$107
$112 $111
$114 $115 $112
$92
$96
$100 $102
$105 $105
$111 $108 $108
$60
$70
$80
$90
$100
$110
$120
$130
$140
2001 2002 2003 2004 2005 2006 2007 2008 2009
> 75% finishedbasement
30%-75% finishedbasement
< 30% finishedbasement
N = 1,481
82/15/2017
Source Data: PVA 3/25/2010
9. Pilot Study: Property Values of 1950s Housing stock
Statistical Modeling in Valuation
AVMs incorporating mass appraisal models use data based on geographical/neighborhood identity and finds appropriate
‘comparable’ properties. In the case of assessor’s models, often more than 20 independent variables are compared
against the selling price of a home. Regression is often run to derive values relative to a specific home. The process
usually includes all properties in the given area because the assessor’s job is to properly value all properties and distribute
the tax burden evenly.
Some AVMs on the other hand value properties one at a time, like a fee appraisor. The process may be progressive
wherein the valuation algorithm is data-driven and starts with the identification of the subject property. Or the process
can be retrospective, based on predetermined valuation equations (much like assessor’s models.)
Pros and Cons to both:
1) Prospective method is cumbersome and blind to problems, but is also dynamic and can be more current than the
retrospective method.
2) The retrospective method has the advantage that it can be verified up to a point. Outliers can be seen in advance
before the information is released to the public.
The following things need to be explained in detail in the appraisal report whenever AVM output is used:
1. number of sales
2. sales not used and reasons why
3. sample size(does the sample represent the whole population or market?
4. method used to derive value---regression, artificial intelligence, expert system, etc.
5. independent variables tested, used and not used in the model
6. area analyzed
7. statistics that measure model accuracy
8. outcome measures (independent/dependant values)
9. clear rationale of the model
10. any other information that may affect reliability of the model. ( I.e., source of sales data, source of property data,
description of editing process)
NOTE: It’s more important to have the simplest most straightforward model with only a small set of variables that can
reliably predict value.
92/15/2017
Source Data: PVA 3/25/2010
Automated Valuation Models (AVMs) and Appraisal
10. 102/15/2017
Source Data: PVA 3/25/2010
Pilot Study: Property Values of 1950s Housing stock
Statistical Modeling in Valuation
Location, size of home, number of bedrooms and bathrooms are the most important variables used to determine sale
value. Other primary variables may include year built, house style, subdivision, number of car spaces, lot size and
basement finished square footage. However, the particular set of primary variables will differ from market to
market.
Most buyers have an upper-limit price constraint, AND a certain minimum level of amenities they prefer. Examples would
be a preference for quality, view, new kitchen and/or yard size.
Primary Variables
Secondary Variables
Fireplaces, garage type, pool and air conditioning may be considered secondary variables important enough to the
homebuyer to include in a model. These variables have some market impact but are less often significant in a
regression model. They can be important though when there is little difference in variation among the primary
variables.
For example, if a neighborhood has homes all built within a 2- year period and all between 1400-1600 square feet of living
area , then 2 primary variables can be excluded from the model ---age and gross living area---and other factors such
as size of garage, fireplaces, or floor plan may be included.
Other Variables
A third set of variables such as location of the laundry area, guest closet, fencing, flooring, patio or deck may influence
some buyers or have small value relative to the overall decision to buy. Variables at this third level tend to be subjective
or calculated by another method: construction quality, physical condition or functional utility for example.
11. Pilot Study: Property Values of 1950s Housing stock
Regression Results
Following the exploratory analysis, two regression models were run. The first model included a
dependent variable of sale price, and independent variables of lot size, above-grade square footage of the home, and
square footage of enclosed porches, open porches, decks, basements, finished basements, and attached and detached
garages.
The descriptive statistics in Model 1 indicate that the average sale price of a 1950s home with full basement over the
years 2001 through 2009 was around $135,000. Lot sizes are almost ¼ acre (0.205), basements are roughly 1/3 finished.
Note: the data is currently insufficient in measuring square footage.
Model 1.
112/15/2017
Source Data: PVA 3/25/2010
12. Pilot Study: Property Values of 1950s Housing stock
Regression Results
The independent variables in Model 1 explain approximately 69% of the dependent variable-median sale price.
Model 1.
The variables in Model
1 that are significant
and help to explain the
sale price of 1950s
homes include those
circled in column 1 at
right, with t-scores
greater than 1.96
122/15/2017
Source Data: PVA 3/25/2010
13. Pilot Study: Property Values of 1950s Housing stock
Regression Results
Model 1. Interpretation of the data:
Independent variables that influence market price. Market Value
Lot size $66,501
Number of stories $12,412
Finished size of home per square foot (above grade) $56
Half-Bathroom $6,373
Full Bathroom $11,037
Basement area per square foot $37
Finished basement area per square foot $5
Attached garage per square foot $16
Detached garage per square foot $7
SaleYear $3,334
The results of the regression in Model 1 suggest that the lot size contributes $66,500 per acre to the total
property value. The above grade finished size of the home is worth $56 per square foot, a half-bathroom
would be $6,000, with additional half baths at $1500-$2000. A full bathroom is worth $11,000, with
additional full baths roughly $3000-$4000. Basement area is worth $37 per square foot and finished
basements add $5-$10 per square foot to the overall cost. An attached garage is worth $16 per square
foot, and a detached garage worth $7 need to recalculate the model to get better numbers.
132/15/2017
Source Data: PVA 3/25/2010
According to Rick, this would
be more like $15-$16 per
square foot.
Land value usually accounts for
~20% of total cost so the $66,000
is very high for a median price of
%135,000.
14. Pilot Study: Property Values of 1950s Housing stock
Regression Results
Model 2.
142/15/2017
Model 2 was run with the same dependent variable of sale price, and the same independent variables, but with urban
neighborhoods added to the model as independent variables. With neighborhoods added the model is a better fit for
the data and explains 82% of results.
Those neighborhoods that significantly impacted sale price include the following:
Algonquin, Auburndale, Audubon, Avondale, Bashford Manor, Beechmont, Belknap, Bon Air, Bowman, Brownsboro,
Camp Taylor, Cherokee Gardens, Cherokee Seneca, Chickasaw, Cloverleaf, Hazelwood, Highland-Douglas, Jacobs,
Kenwood Hill, Klondike, Park DuValle, Poplar Level, Portland, Prestonia, Rock creek, St Joseph, Shawnee, South
Louisville, Southland Park, Southside, Taylor-Berry and Tyler Park.
The Bar Chart in Figure 5 shows the neighborhoods listed according to the number of homes sold between 2001 and
2009. The Table included in Figure 5 shows the value added or subtracted from a given house based on neighborhood
location.
Source Data: PVA 3/25/2010
15. Source Data: PVA 3/25/2010
Pilot Study: Property Values of 1950s Housing stock
Regression Results
15
1
1
2
2
2
2
2
3
4
5
5
6
8
8
9
9
9
10
10
11
11
12
12
14
16
16
16
22
22
25
26
30
31
31
37
39
49
50
53
69
70
87
95
117
136
259
CHSENECA
MERRIWETHER
ALGONGUIN
PARKDUVALLE
PORTLAND
SCHNITZELBURG
SOUTHLOUISVILLE
TYLERPARK
CHGARDENS
DEERPARK
GERMANTOWN
AUBURNDALE
JACOBS
SHAWNEE
CAMPTAYLOR
CRESCENTHILL
WYANDOTTE
HAZELWOOD
TAYLORBERRY
REMAINDERCITY
SAINTJOSEPH
CLIFTONHEIGHTS
HAYFIELDDUNDEE
HILANDDOUGLAS
CHICKASAW
PRESTONIA
SOUTHSIDE
IROQUOIS
SOUTHLANDPARK
BROWNSBORO
BEECHMONT
GARDINERLANE
BELKNAP
HAWTHORNE
BOWMAN
ROCKCREEK
KENWOODHILL
POPLARLEVEL
IROQUOISPARK
AUDUBON
CLOVERLEAF
BASHFORD
HIKESPOINT
KLONDIKE
AVONDALE
BONAIR
N= 1,454
2/15/2017
Figure 5.
Number of Homes Sold by Neighborhood 2001-2009
Neighborhood
Value Added to
Sale Based on
Neighborhood
LocationsAlgonquin $50,180
Auburndale $33,662
Audubon $10,711
Avondale $14,581
Bashford Manor $28,058
Beechmont $34,952
Belknap $12,575
Bon Air $18,829
Bowman $11,511
Brownsboro $16,323
Camp Taylor $25,422
Cherokee Gardens $63,693
Cherokee Seneca $71,350
Chickasaw $51,233
Cloverleaf $28,574
Hazelwood $39,948
Highland-Douglas $62,944
Jacobs $45,009
Kenwood Hill $36,746
Klondike $20,104
Park DuValle $67,488
Poplar Level $17,973
Portland $50,710
Prestonia $33,477
Rock Creek $31,934
St Joseph $15,549
Shawnee $50,212
South Louisville $42,269
Southland Park $35,217
Southside $38,684
Taylor-Berry $45,528
Tyler Park $67,223
The Bar Chart (left) in
Figure 5 shows the
neighborhoods listed
according to the
number of homes sold
between 2001 and
2009. The Table (right)
shows the value added
or subtracted from a
given house based on
neighborhood location.
(Numbers in red
indicate negative
values.)
Given houses
with identical
characteristics,
the sale price
could be
roughly
$60,000 more
than the
median of
$135,000 in
Cherokee
Gardens, and
$50,000 less
than the
median in
neighborhoods
such as
Algonquin,
Chickasaw,
Portland or
Shawnee.
16. 2/15/2017 16
Conclusions
This study found that having some portion of the basement finished certainly added value to the home, but that value diminished
when the basement was more than 3/4s finished. The more significant indicator of home value was neighborhood.