The Impact of World Class Great Lakes Water Views On Residential Property Values.
Article Type:
Statistical Data Included
Real property (Valuation)
Views (Valuation)
Real estate industry (Management)
Seiler, Michael J.
Bond, Michael T.
Seiler, Vicky L.
Pub Date:
Name: Appraisal Journal Publisher: The Appraisal Institute Audience: Trade Format: Magazine/Journal Subject: Business; Real estate industry Copyright: COPYRIGHT 2001 The Appraisal Institute ISSN: 0003-7087
Date: July, 2001 Source Volume: 69 Source Issue: 3
Product Code: 6500000 Real Estate NAICS Code: 53 Real Estate and Rental and Leasing
Geographic Scope: United States Geographic Name: Lake Erie Geographic Code: 1USA United States

Accession Number:
Full Text:

This study examines the impact of water views on property values. Half of the samples of homes in this study have a clear water view while the other half, which consist of adjacent properties, do not. The lake under investigation is one of the Great Lakes, Lake Erie, which is fresh water and freezes over in the winter. It provides an ocean-like view because of its size and because it generates waves large enough to surf. The study results show that having a lake view increases home value by $115,000, or by approximately 56%.

This study examines the impact of water views on the value of residential properties. Specifically, the impact of Lake Erie views on residential properties is studied. Cuyahoga County, Ohio, has approximately 1.5 million individuals residing in Cleveland and 31 suburbs. Five of these municipalities are located on Lake Erie. One of the five Great Lakes, Lake Erie, is the second smallest and the shallowest. It covers 9,966 square miles and its maximum depth is 210 feet. Industrialization in the 1800s and 1900s seriously polluted the lake, but conservation and cleanup efforts have significantly improved its quality in the past 40 years. The lake is a prime spring, summer, and fall resort area for boating and fishing. While it is a fresh-water lake, its size and waves give it an ocean-like view. As such, it should be expected to significantly increase property values for homes located on the water.

There has yet to be an empirical examination of the impact of Great Lakes views on residential property values. In addition, there appears to be a paucity of research on the impact of waterfront views on residential property values. In this study, we examine the relationship between the property values of 1,172 lakefront and adjacent residential properties and several factors that are expected to influence home values. The results show that having a lake view does cause significantly higher property values.

Section two of the paper summarizes the existing literature surrounding the degree of waterfront impact on property values. Section three describes the data and research methodology employed in this study, while section four reports the empirical results. Finally, section five summarizes the research findings and offers suggestions for further research in this area.

Literature Review

Previous research concerning the value of various types of views on property values is limited. Darling [1] investigated the impact of distance from three urban lake parks in California. The distance from the lakes was found to be a significant determinant of property value in several of the estimates. Two of the parks (Lake Murray and Santee Lakes) included a dummy variable for the impact of water view on value. This variable showed an increase in value of $2,362 and $2,756, respectively, for having a water view.

Davies [2] studied factors affecting home prices in Nottingham, UK, and found that view, as measured by degrees of panorama, was not a significant determinant of value. Morton [3] examined the determinants of value for approximately 400 single-family homes in Orange County; California. The dummy variable for view was significant. View, on average, added $19,748 to sales prices.

Brown and Pollackowski [4] considered the value of living near water in Seattle. They estimated two separate models. The first was for Green Lake, which is enclosed by a greenbelt. The second model was for Lake Washington and Haller Lake, which have houses built to the lake's shore. The regression models on selling price included a variable for distance from a waterfront as well as a dummy variable for view. Both models found a greater distance from the lake significant reduces selling price. The view dummy was not statistically significant although it had a positive coefficient in both models.

Correll, Lillydahl, and Singell [5] investigated the relationship between the sales price of single-family properties and their distance from a greenbelt in three neighborhoods in Boulder, Colorado. In one of the neighborhoods, they found that sales price decreased by $10.20 for every foot the house was away from the greenbelt. In addition, the authors examined view by classifying properties into excellent, moderate, and no view of the valley. They concluded that the coefficients associated with these view variables were not statistically significant. Apparent reasons why view was found not to be significant in the study were that the sample was small (only 36 properties), views were difficult to define, and some of the best view properties were subject to severe winds.

Plattner and Campbell [6] examined new condominium sales of two developments in western Massachusetts. These projects had units with similar characteristics except that some units had views of adjacent ponds or small lakes and the remaining units did not have these views. By comparing average sales prices of units with and without the water views, the authors found that a water view added 4%-12% to condo values. They also found that the percentage increase was higher for lower-price units.

Gillard [7] used a hedonic pricing model to examine the relationship between property values and view. He examined 392 single-family home sales in Los Angeles in 1970. He used a dummy variable for view in a multiple regression model including nine other hedonic variables and four variables for neighborhood characteristics. The results indicated that a home located on a view lot has about 9% greater value than homes with no view. Rodriguez and Sirmons [8] studied the impact of views on value with a sample of 194 home sales in Fairfax County, Virginia. Using a hedonic pricing model, they found that a "good view" added approximately 8% to home values. The property value variable was obtained from county tax appraisals.

Benson, Hanson, Schwartz, and Smersh [9] examined view as an explanatory variable in a study of 397 residential property sales in Point Roberts, Washington. They used a four-way classification system of ocean front, ocean view, partial ocean view, and no view (determined by personal inspection of the sample properties). They concluded that ocean front views added 147% to the selling price, ocean view added 32%, and partial ocean view added 10% in value, relative to no view.

Benson, Hanson, Schwartz, and Smersh [10] used a detailed classification of water views to examine property value impact. The study was conducted using data for the city of Bellingham, Washington. They found that simple specification of a water view added approximately 22% to home values. When a more detailed classification was used, the percentage increase in value ranged from 6% for a mountain view, 8% for a poor partial ocean view, 10% for a partial lale view, 15% for a lake view, 19% for a good partial ocean view, 27% for a superior ocean view, 53% for an unobstructed ocean view, and 107% for a lakefront property. They also found that distance from the view impacted property values.

These previous studies suggest that views add value to properties and that the quality of a view is also important. This paper extends the existing literature in several ways. It uses a much larger sample of properties (1,172) in a much more controlled setting. There is no combination of mountain, ocean, and lake views-there is simply either a view of Lake Erie from the property or there is not. Finally, this lake is different from the body of waters examined in other studies because it is so large that it has the appearance of an ocean view and can generate large waves like an ocean, yet freezes over in the winter.

Data and Methodology

A hedonic pricing model is utilized to examine the impact of several variables on single-family housing values. The sample properties used in the study are drawn from computer data files provided by the Housing Center at the College of Urban Affairs at Cleveland State University. This file includes tax assessment values from 1998. Ideally, transaction-based data is preferred, however they are not available in this study. Neighborhood effects are controlled by only comparing lakefront to adjacent properties. The hedonic model is shown in Equation 1. All statistical procedures are performed in Statistical Package for Social Sciences (SPSS).

VALUE = f (view, construction quality, age, roof type, basement, A/C, attic, rooms, bedrooms, bathrooms, fireplaces, square footage, lot size, frontage) (1)


VALUE 1998 Cuyahoga County Property Tax Value,

VIEW = 0 - no view (an adjacent property) 1 - lakefront view,

CONSTRQL = a rating of the home's construction quality, AGE the age of the house (in years), ROOFSTYLE = 1 if slate and tile

2 if wood shingle

3 if asphalt shingle,

BASEMENT = 0 - slab or crawl space 1 - basement,

AIRCOND = 1 - central air

2 - window unit air conditioner

3 - none,

ATTIC = 1 - full finished attic

2 - unfinished attic

3 - no attic,

ROOMS total number of rooms in the home,

BEDROOMS = number of bedrooms in the home, BATHS number of bathrooms in the home,

FIREPLACE = number of fireplaces in the home, HOMESQFT = total square feet of living space in the home,

LOTSIZE = lot size (in square feet), and

FRONTAGE = length of linear frontage (in feet).

Table 1 presents descriptive statistics for all the variables used in the analysis. The average assessed home value is $292,530 with roughly half of the sample located on Lake Erie. The average number of fireplaces per home is over one, and approximately half of the homes do not have an air conditioner. These two summary statistics are not surprising due the northern location and the fact that the homes are on or very near the water. On average, the sample included homes with above average construction quality. The average age of homes in the sample is slightly over 61 years. Average living space is almost 2,700 square feet with an average lot size of roughly 23,000 square feet.


The key variable of interest is the effect of view on home values. Table 2 provides some cursory insight into the relationship.

A t-test is performed to measure if there is a difference in the property value of homes located on the lake versus the value of adjacent properties. The difference in home price is a statistically significant $190,811.93 ($395,262.48-$204,450.55) or 93.3%. That is, homes with a lake view are almost twice as valuable as homes without a lake view.

In order to control for the effects of other variables that might cause differences in home values, additional analyses must be performed. Before these variables are introduced into the hedonic model represented by Equation 1, a series of analysis of variance (ANOVA) and least significant difference (LSD) tests have been completed to determine how the categorical variables ATTIC, AIRCOND, and ROOFSTYLE affect property values. ANOVA is a procedure used to determine if the means of more than two variables are significantly different from the overall mean. An LSD test delves further by performing pairwise comparisons of each of the variable's means to the means of each of the other variables to determine which ones are significantly different from each other.

The ANOVA analysis performed in Part A of Table 3 shows that the type of attic does affect property values differently. Specifically, homes with finished attics are significantly more valuable than homes without attics ($341,292 versus $280,512). Similar results are found in Part B where the homes with air conditioners are significantly more valuable than homes without. In fact, homes with central air conditioning are valued $102,667 higher ($359,817 versus $257,150) than homes without air conditioning. This, of course, does not yet consider the additional variables relevant to explaining home values. Part C of Table 3 reflects significant differences as well. Both "slate and tile" and "wood shingle" roofs are associated with far more valuable homes than "asphalt shingle" roofs ($115,824 and $87,227, respectively). Since the subclassifications of these three categorical variables are shown to significantly affect property values differently, they will be converted into dichotomous variables and included as du mmy variables in the hedonic model.

Many of the independent variables provided in the database are suspected to be highly correlated with each other. For example, the number of bedrooms, number of bathrooms, total number of rooms in the house, and the home's square footage should be highly correlated. In fact, many of these variables are subsets of each other. For this reason, it is necessary to identify those variables that will cause multicollinearity in the hedonic model and remove them beforehand. Table 4 provides a correlation matrix of the variables that could possibly be highly related with one another. As expected, several variables are too highly correlated. To avoid misspecifying the hedonic model, the highly correlated variables (number of bedrooms, number of bathrooms, number of total rooms, number of fireplaces, and the home's square footage) will be represented by the single variable "square footage."

Table 5 shows the regression results for the ten remaining independent variables against the dependent variable VALUE. The F-statistic shows the over-all model to be significantly beyond the 1% Level. The individual t-statistics reveal that seven of the ten explanatory variables are significant. Nine of the ten variables have the expected sign. The only one that does not is the insignificant variable LOTSIZE. Since three of the independent variables are not significant, the hedonic regression equation will again be estimated, this time with only the significant variables included.

Table 6 shows the final results of the study. The seven independent variables are all significant determinants of home values. The key variable of interest, VIEW, has a t-statistic over seven times greater than the statistically significant cutoff (t = 14.346 versus 1.960). This robust result demonstrates the significant positive relationship that exists between the view from a property and the property's value. To discuss the results in terms of dollars, the beta coefficient of 115,000.43 shows that after controlling for the relevant determinants of property values, homes with a view are worth $115,000 more than homes without a view.

Not surprisingly, larger homes (HOMESQFT), homes constructed of a higher quality (CONSTRQL), newer homes (AGE), homes with air conditioners (AIRCOND), homes with basements (BASEMENT), and homes using more expensive roofing materials (ROOFTYPE) have higher values.

To verify that multicollinearity is not a problem, advanced diagnostic tests have been performed. There are two sets of figures used to assess multicollinearity; (1) tolerance and variance inflation factor (VIF), and (2) condition index and variance proportions. Tolerance values approaching zero indicate that the variable is too highly predicted, or is collinear, with the other independent variables. The critical cutoff value is 0.10. [11] A related reported value is the VIF. This is simply the reciprocal of the tolerance. For this reason, values below ten show that multicollinearity is not present. As seen in Table 6, none of the variables come close to violating the no multicollinearity regression assumption.

At the bottom of Table 6, the collinearity diagnostics box shows the second set of tests for multicollinearity. The maximum condition index is 17.45. Values below 30 are not problematic. [12] Moreover, the variance proportion values are far below the cutoff of 0.50. For these reasons, the finding of no multicollinearity is again confirmed which allows the results to be interpreted without bias.

Summary and Conclusions

Prime views in general, and prime water views in particular, should positively impact residential property values. While this seems logical, the empirical research necessary to support this claim is somewhat limited. Most existing studies do suggest value enhancement from scenic water views, but many of these suffer from sample size problems or a failure to adjust for the quality of the water view.

In this study, we empirically test a hedonic home valuation model using a large sample of homes where properties on Lake Erie have a water view, but adjacent properties do not. The "tight fit" lakefront homes on Lake Erie allow this rare ability to determine the value of the view holding other variables constant. The results clearly demonstrate that water views have a robust statistically significant impact on home values even after adjusting for other significant home value determinants. The premium paid for homes with a view translates into $115,000, or a premium of 56%. That is, when appraising a home that has a water view of Lake Erie, appraisers should add $115,000 to their appraised valuation. Of course, not all water views around the country are exactly equal to those of Lake Erie. As such, the results of any study-including this one-cannot be used as a cookbook to exactly determine the value of specific home attributes when appraising a residence located in a different geographic region. Instead, we provide a starring point from which appraisers can deviate when assessing premium views in their area of the country.

Michael J. Seiler is an associate professor of finance at Hawaii Pacific University in Honolulu, HI, specializing in real estate portfolio management.

Michael T. Bond is an associate professor in the finance department at Cleveland State University, Ohio.

Vicky L. Seiler is an assistant professor of marketing at Hawaii Pacific University in Honolulu, Hi, specializing in service quality in residentail real estate brokerage.

(1.) A. Darling, "Measuring Benefits Generated by Urban Water Parks." Land Economics (49:1, 1973) 22--34.

(2.) C. Davies, An Econometric Analysis of Residential Amenity." Urban Studies (11, 1974): 217-225.

(3.) T. Morton, "Factor Analysis, Multicollinearity and Regression Appraisal Models" The Appraisal Journal (October 1977): 578-588.

(4.) G. Brown, and H. Pollakowski. "Economic Value of Shoreline." The Review of Economics and Statistics (59:3,1997): 272-278.

(5.) M. Correll, 3. Lillydahl, and L Singell. "The Effects of Greenbelts on Residential Property Values: Same Findings on the Political Economy of Open Space," Land Economics (54:2, 1978): 207-217.

(6.) R. Plattner, and T. Campbell. "A Study of the Effect of Water View on Site Value." The Appraisal Journal (January, 1978): 20-25.

(7.) Q. Gillard, "The Effect of Environmental Amenities on House Values: The Example of a View Lot." Professional Geographer(33:2, 1981): 216-220.

(8.) M. Rodriguez, and C. Sirmons. "Quantifying the Value of a View in Single Family Housing Markets." The Appraisal Journal (October, 1994): 600-603.

(9.) E. Benson, I. Hanson, A. Schwartz, and C. Smersh. "The Influence of 'World Class Water' Views on Residential Property Values." Presented at the American Real Estate Society Meeting, Lake Tahoe, Calif., (April 1996).

(10.) E. Benson, J. Hanson, A. Schwartz, and C. Smersh. "The Influence of Canadian Investment on U.S. Residential Property Values." Journal of Real Estate Research (1 3:3, 1997): 231-249.

(11.) J. Hair, R. Anderson, R. Tatham, and W. Black. Multivariate Data Analysis, 4th Ed. Englewood Cliffs, NJ: Prentice Hall, (1995).

(12.) Ibid.


Benson, E., J. Hanson, A. Schwartz, and G. Smersh. "The Influence of 'World Class Water' Views on Residential Property Values." Presented at the American Real Estate Society Meeting, Lake Tahoe, Calif., April 1996.

Benson, E., J. Hanson, A. Schwartz, and G. Smersh. "The Influence of Canadian Investment on U.S. Residential Property Values." Journal of Real Estate Research (13:3, 1997): 231-249.

Brown, G., and H. Pollakowski. "Economic Value of Shoreline." The Review of Economics and Statistics (59:3, 1997): 272-278.

Clapp, J., and C. Giaccotto. "Estimating Price Trends for Residential Property: A Comparison of Repeat Sales and Assessed Value Methods." Journal of Real Estate Finance and Economics (5, 1992): 357-374.

Correll, M., J. Lillydahl, and L. Singell. "The Effects of Greenbelts on Residential Property Values: Some Findings on the Political Economy of Open Space." Land Economics (54:2, 1978): 207-217.

Darling, A. "Measuring Benefits Generated by Urban Water Parks." Land Economics (49:1, 1973): 22-34.

Davies, G. "An Econometric Analysis of Residential Amenity." Urban Studies (11, 1974): 217-225.

Gillard, Q. "The Effect of Environmental Amenities on House Values: The Example of a View Lot." Professional Geographer (33:2, 1981): 216--220.

Hair, J., R. Anderson, R. Tatham, and W Black. Multivariate Data Analysis, 4th Ed. Englewood Cliffs, NJ: Prentice Hall, 1995.

Morton, T. "Factor Analysis, Multicollinearity and Regression Appraisal Models." The Appraisal Journal (October 1977): 578--588.

Plattner, R. and T. Campbell. "A Study of the Effect of Water View on Site Value." The Appraisal Journal (January 1978): 20--25.

Rodriguez, M. and C. Sirmons. "Quantifying the Value of a View in Single Family Housing Markets." The Appraisal Journal (October 1994): 600-603.
Table 1
Descriptive Statistics for Each Variable
                        N  Minimum  Maximum       Mean
AGE                  1172        3      172      61.29
    Central air       402
    Window unit         6
    No AC             764
    Fully finished    208
    Unfinished        201
    No attic          763
    Basement         1057
    No basement        91
BATHROOM             1171        1        9       1.99
BEDROOMS             1172        1       12       3.75
CONSTRQL             1170        1        6       4.00
FIREPLACE            1172        0        7       1.15
FRONTAGE             1172        0    12750      93.95
HOMESQFT             1172      456    13320    2691.60
LOTSIZE              1170     1742   744004   22986.24
    Slate and tile    264
    Wood shingle       53
    Asphalt shingle   855
ROOMS                1172        3       27       8.06
VALUE                1172    22600  1624100  292530.12
    No                631
    Yes               541
Table 2
T-Test to Determine if Having a Lake View Affects the Value of the Home
            VIEW   N     Mean     Std. Deviation  Std. Error Mean
Home Value   No   631  204450.55    101464.65         4039.24
            Yes   541  395262.48    225397.62         9690.60
            F-test sig.
Home Value      195.062
Table 3
ANOVA and LSD Tests
Part A: Attic
VALUE             Sum of Squares     df      Mean Square       F
Between groups    609470761992.877     2   304735380996.438  8.112
Within groups   43915617024790.390  1169    37566823802.216
Total           44525087786783.270  1171
VALUE           Sig.
Between groups  .000
Within groups
    (I)             (J)
   ATTIC           ATTIC        Mean Difference (I - J)  Std. Error
Fully finished  Unfinished            53601.28 *         19170.533
                No attic              60779.64 *         15160.653
Unfinished      Fully finished        53601.28 *         19170.533
                No attic               7178.36           15366.695
No attic        Fully finished        60779.64 *         15160.653
                Unfinished            -7178.36           15366.695
                      95% Confidence
   ATTIC        Sig.   Lower Bound    Upper Bound
Fully finished  .005     15988.78       91213.78
                .000     31034.51       90524.77
Unfinished      .005    -91213.78      -15988.78
                .640    -22971.02       37327.74
No attic        .000    -90524.77      -31034.51
                .640    -37327.74       22971.02
Part B: Air Conditioning
VALUE             Sum of Squares     df      Mean Square       F
Between groups   2776492217000.565     2  1388246108500.282  38.872
Within groups   41748595569782.690  1169    35713084319.746
Total           44525087786783.250  1171
VALUE           Sig.
Between groups  .000
Within groups
  (I)          (J)
AIRCOND      AIRCOND      Mean Difference (I - J)  Std. Error  Sig.
Central air  Window unit         70400.50          77723.994   .365
             No AC              102667.56 *        11644.028   .000
Window unit  Central Air        -70400.50          77723.994   .365
             No AC               32267.06          77452.731   .677
No AC        Central Air       -102667.56 *        11644.028   .000
             Window unit        -32267.06          77452.731   .677
             95% Confidence
AIRCOND       Lower Bound    Upper Bound
Central air    -82093-62      222894.61
                79822.03      125513.09
Window unit   -222894-61       82093.62
              -119694.84      184228.96
No AC         -125513.09      -79822.03
              -184228.96      119694.84
Part C: Roof Style
VALUE             Sum of Squares     df      Mean Square          F
Between groups   2887643164166.799     2  1443821582083.400  40.536
Within groups   41637444622616.420  1169    35618002243.470
Total           44525087786783.210  1171
VALUE           Sig.
Between groups  .000
Within groups
   (I)              (J)
ROOFSTYLE        ROOFSTYLE        Mean Difference (I - J)  Std. Error
Slate and tile   Wood shingle            28597.06          28406.944
                 Asphalt shingle        115824.26 *        13288.159
Wood shingle     Slate and tile         -28597.06          28406.944
                 Asphalt shingle         87227.20 *        26715.094
Asphalt shingle  Slate and tile        -115824.26 *        13288.159
                 Wood shingle           -87227.20 *        26715.094
                       95% Confidence
ROOFSTYLE        Sig.   Lower Bound    Upper Bound
Slate and tile   .314    -27137.24       84331.35
                 .000     89752.95      141895.56
Wood shingle     .314    -84331.35       27137.24
                 .001     34812.31      139642.09
Asphalt shingle  .000   -141895.56      -89752.95
                 .001   -139642.09      -34812.31
(*)The mean difference is significant at the .05 level.
Table 4
Correlation Matrix
                  Bath    Bed   Construction  Fire              Lot
            Age   Rooms  Rooms    Quality     Place  Frontage  Size
AGE        1.000  -.077   .202     -.071       .018   -.001     .088
BATHROOM   -.077  1.000   .519      .397       .509    .039     .315
BEDROOMS    .202   .519  1.000      .319       .425    .014     .266
CONSTRQL   -.071   .397   .319     1.000       .360   -.004     .146
FIREPLACE   .018   .509   .425      .360      1.000   -.007     .274
FRONTAGE   -.001   .039   .014     -.004      -.007   1.000     .052
HOMESQFT    .039   .685   .609      .465       .543    .033     .493
LOTSIZE     .088   .315   .266      .146       .274    .052    1.000
ROOMS       .167   .624   .784      .419       .510    .040     .408
AGE         .167
BATHROOM    .624
BEDROOMS    .784
CONSTRQL    .419
FRONTAGE    .040
HOMESQFT    .768
LOTSIZE     .408
ROOMS      1.000
Table 5
Regression Results With All Ten Indenpendent Variables
Model Summary a
Model    R     R Square  Adjusted R Square  Std. Error of the Estimate
       .756 b    .572          .568                 127977.85
Model  Durbin-Watson
Model         Sum of Squares     df      Mean Square        F
Regression  24793801644595.170    10  2479380164459.517  151.382
Residual    18589405152263.480  1135    16378330530.629
Total       43383206796858.650  1145
Model        Sig.
Regression  .000 a
Coefficients a
          Unstandardized              Standardized
           Coefficients               Coefficients
Model           B         Std. Error      Beta        t     Sig.
Constant     -7.68.273    23743.659                   -310  .756
VIEW        116815.123     8182.650       .299      14.276  .000
AGE           -697.195      182.037      -.088      -3.830  .000
AC_DUM       45647.310     8880.079       .112       5.140  .000
ATTICDUM      3775.692     9096.161       .009        .415  .678
ROOF_DUM     25530.945     9317.720       .058       2.740  .006
HOMESQFT        67.733        3.728       .472      18.170  .000
CONSTRQL     19962.352     3330.569       .142       5.994  .000
FRONTAGE         1.886       10.032       .004        .188  .851
LOTSIZE     -9.727E-02         .136      -.016       -.713  .476
BASEMENT     32198.147    14414.546       .045       2.234  .026
Model      Tolerance     VIF
VIEW          .860      1.163
AGE           .718      1.393
AC_DUM        .797      1.255
ATTICDUM      .753      1.329
ROOF_DUM      .834      1.199
HOMESQFT      .559      1.789
CONSTRQL      .673      1.487
FRONTAGE      .994      1.006
LOTSIZE       .718      1.392
BASEMENT      .941      1.063
(a)Dependent Variable: VALUE
(b)Predictors: (Constant), BASEMENT,FRONTAGE, VIEW, AC_DUM,ROOF_DUM,
Table 6 Regression Results With the Seven Significant Independent
Model Summary a
Model              R     R Square  Adjusted R Square
1                .755 b    .571          .568
Model Summary a
Model            Std. Error of the Estimate  Durbin-Watson
1                        127838.46               1.169
Model 1       Sum of Squares      df     Mean Square        F
Regression  24762111883259.010     7  3537444554751.287  216.454
Residual    18630647223805.510  1140    16342673003.338
Total       43392759107064.510  1147
Model 1      Sig.
Regression  .000 b
Coefficients *
          Unstandardized              Standardized
           Coefficients               Coefficients
Model           B         Std. Error      Beta        t     Sig.
Constant    -8650.063     23500.380                  -.368  .713
VIEW       115000.428      8016.443       .295      14.346  .000
AGE          -681.947       173.575      -.086      -3.929  .000
AC_DUM      45337.355      8786.058       .111       5.160  .000
ROOF_DUM    26130.126      9253.895       .060       2.824  .005
HOMESQFT       66.695         3.269       .465      20.403  .000
CONSTRQL    20601.787      3258.954       .147       6.322  .000
BASEMENT    33340.238     14274.125       .046       2.336  .020
Model      Tolerance     VIF
VIEW          .892      1.121
AGE           .788      1.270
AC_DUM        .810      1.234
ROOF_DUM      .843      1.186
HOMESQFT      .725      1.379
CONSTRQL      .700      1.428
BASEMENT      .957      1.045
Collinearity Diagnostics a
                                  Variance Proportions
                       Condition                              AC
Dimension  Eigenvalue    Index          Constant        VIEW  AGE
1          6.218         1.000            .00           .01   .00
2           .655         3.080            .00           .00   .02
3           .517         3.467            .00           .69   .00
4           .299         4.560            .00           .24   .01
5           .134         6.800            .00           .05   .21
6          9.084E-02     8.273            .01           .00   .41
7          6.502E-02     9.779            .00           .00   .13
8          2.042E-02    17.448            .99           .00   .21
1          .01   .00    .00       .00       .00
2          .68   .00    .00       .00       .00
3          .03   .07    .00       .00       .00
4          .00   .44    .07       .02       .00
5          .13   .16    .51       .02       .05
6          .09   .01    .26       .26       .12
7          .01   .01    .15       .38       .60
8          .05   .30    .00       .32       .22
(a)Dependent Variable: VALUE.
(b)Predictors: (Constant), BASEMENT, VIEW, AC DUM, ROOF DUM, HOMESQFT,
Gale Copyright:
Copyright 2001 Gale, Cengage Learning. All rights reserved.