Multiple Regression Analysis

4.3.1 Light Filtering Sales in Units and Dollars

Further analysis was done by using multiple regression line analysis and having all these studied variables (including variables from Environment, Demographics and Economics) with the Light Filtering category in sales units. The data shows that R square is 0.89044, the adjusted R square is 0.85873, F is 28.0777 and the P-value is less than 0.01. Thus, the model is statistically significant. Table 9 shows P-values of Non-Hispanic White, Asian and Median Household Income are all less than 0.05. This indicates that these three factors are significant to the effects of sales in units. To further explain, every one thousand population increase in Non-Hispanic White would increase the sales units by 0.059 and every one thousand population increase in Asian would increase the sales units by 0.342. Every dollar increase in Median Household Income would increase the sales units by 0.006 units.

Table 9: Multiple Regression Analysis of Light Filtering in Sales Units

For the Light Filtering sales in dollars, the data shows that R square is 0.89849, the adjusted R square is 0.86911, F is 30.578 and the P-value is less than 0.01 Thus the model is statistically significant. The table 10 shows P-values of Non- Hispanic White, Asian and Median Household Income are all less than 0.05. It indicates that all three independent variables are significant of sales in dollars. By looking at the coefficient of each independent variable, every one thousand increase in population of Non-Hispanic White would increase the sales dollars by $1.273 and every one thousand increase in population of Asian would raise the sales dollars by $7.311. For every dollar increase in Median Household Income, it would increase the sales dollars by $0.128.

American Indian or Alaskan Native 0.138 0.255 0.544 0.590

Asian 0.342 0.105 3.263 0.002

Native Hawaiian or Pacific Islander -1.528 1.077 -1.420 0.164

Mixed race -0.926 0.528 -1.756 0.087

56

Table 10: Multiple Regression Analysis of Light Filtering in Sales Dollars

4.3.2 Room Darkening Sales in Units and Dollars

More analysis was done by having all these studied variables (including variables from Environment, Demographics and Economics) with Room Darkening in sales units and dollars.

For the Room Darkening sales in units, the data shows that R square is 0.86374, the adjusted R square is 0.82429, F is 21.8978 and the P-value is less than 0.01 Thus the model is statistically significant. Table 11 shows that for P-values of Non- Hispanic White and Asian, Home Sales and Median Household Income are all less than 0.05. This shows that these four factors are significant to the effects of sales in units. To further explain, every one thousand population increase in Non-Hispanic White would increase the sales units by 0.083 and every one thousand population increase in Asian would increase the sales units by 0.396. Every one increase of Home Sales would decrease the sales units by 0.001. Every dollar increase on Median Household Income would increase the sales units by 0.008 units.

Coeff's. S.E. t Stat P-value

American Indian or Alaskan Native 2.921 5.428 0.538 0.594

Asian 7.311 2.232 3.276 0.002

Native Hawaiian or Pacific Islander -31.638 22.943 -1.379 0.176

Mixed race -19.604 11.242 -1.744 0.089

57

Table 11: Multiple Regression Analysis of Room Darkening in Sales Units

For the Room Darkening sales in dollars, the data shows that R square is 0.87159, the adjusted R square is 0.83442, F is 23.4487 and the P-value is less than 0.01. Thus the model is statistically significant. Table 12 shows the P-values of Non- Hispanic White, Asian, Home Sales and Median Household Income are all less than 0.05. It displays that all four independent variables are significant of sales in dollars. By looking at the coefficient of each independent variable, every one thousand increase in population of Non-Hispanic White would increase the sales dollars by $2.067 and every one thousand increase in population of Asian would raise the sales dollars by

$10.259. Every one increase of Home Sales would decrease the sales in dollars by

$0.019. Every dollar increase in Median Household Income would increase the sales dollars by $0.191.

Table 12: Multiple Regression Analysis of Room Darkening in Sales Dollars

Coeff's. S.E. t Stat P-value

American Indian or Alaskan Native 0.159 0.342 0.465 0.644

Asian 0.396 0.141 2.819 0.008

Native Hawaiian or Pacific Islander -1.879 1.446 -1.300 0.202

Mixed race -1.127 0.708 -1.590 0.120

American Indian or Alaskan Native 3.928 8.514 0.461 0.647

Asian 10.259 3.500 2.931 0.006

Native Hawaiian or Pacific Islander -46.183 35.987 -1.283 0.207

Mixed race -28.813 17.634 -1.634 0.111

58

4.3.3 Light Blocking Sales in Units and Dollars

The same analysis was done by having all these independent variables (including variables from Environment, Demographics and Economics) with Light Blocking in sales units and dollars.

For the Light Blocking sales in units, the data shows that R square is 0.7653, the adjusted R square is 0.69736, F is 11.2644 and the P-value is less than 0.01 Thus the model is statistically significant. Table 13 shows the P-values of Average Temperature 2014, Non- Hispanic White and Median Household Income are all less than 0.05. This shows that these three factors are significant to the effects of sales in units. To further explain, an increase of one degree in Average Temperature would increase the sales units by 37.341. Every one thousand population increase in Non-Hispanic White would increase the sales units by 0.216. Every dollar increase on Median Household Income would increase the sales units by 0.042 units.

Table 13: Multiple Regression Analysis of Light Blocking in Sales Units

For the Light Blocking sales in dollars, the data shows that R square is 0.81739, the adjusted R square is 0.76453, F is 15.4633 and the P-value is less than 0.01. Thus the model is statistically significant. Table 14 shows the P-values of Average Temperature 2014 for Non- Hispanic White, Black and Median Household Income are all less than 0.05. It indicates that all four independent variables are significant of sales in dollars. By looking at the coefficient of each independent variable, an increase of one degree in Average Temperature would increase the sales by $1797.576. Every one thousand increase in population of Non-Hispanic White would increase the sales dollars by $9.057 and every one thousand increase in population of Black would increase the sales dollars by $14.770. Every dollar increase in Median Household Income would increase the sales dollars by $1.974.

Coeff's. S.E. t Stat P-value

American Indian or Alaskan Native -0.153 1.708 -0.090 0.929

Asian 1.036 0.702 1.476 0.148

Native Hawaiian or Pacific Islander -9.052 7.219 -1.254 0.218

Mixed race -3.825 3.538 -1.081 0.286

59

Table 14: Multiple Regression Analysis of Light Blocking in Sales Dollars

4.3.4 Total Sales in Units and Dollars

Lastly, another analysis was done by measuring all these studied variables against Total Sales in units and dollars of three categories. For the Total Sales in units, the data shows that R square is 0.8328, the adjusted R square is 0.7844, F is 17.2063 and the P-value is less than 0.01. Thus the model is statistically significant. Table 15 shows P-values of Average Temperature 2014 for Non- Hispanic White, Asian and Median Household Income are all less than 0.05. This shows that these four factors are significant to the effects of sales in units. To further explain, an increase of one degree in Average Temperature would increase the sales units by 47.297. Every one thousand population increase in Non-Hispanic White would increase the sales units by 0.358 and every one thousand population increase in Asian would increase the sales units by 1.774. Every dollar increase on Median Household Income would increase the sales units by 0.056 units.

Table 15: Multiple Regression Analysis of Total in Sales Units

Coeff's. S.E. t Stat P-value

American Indian or Alaskan Native -23.265 80.804 -0.288 0.775

Asian 35.575 33.222 1.071 0.291

Native Hawaiian or Pacific Islander -396.934 341.557 -1.162 0.252

Mixed race -167.731 167.368 -1.002 0.323

American Indian or Alaskan Native 0.145 2.054 0.070 0.944

Asian 1.774 0.844 2.101 0.042

Native Hawaiian or Pacific Islander -12.459 8.682 -1.435 0.159

Mixed race -5.877 4.254 -1.381 0.175

60

For the Total Sales in dollars, the data shows that R square is 0.84636, the adjusted R square is 0.80189, F is 19.0307 and the P-value is less than 0.01. Thus the model is statistically significant. Table 16 shows the P-values of Average Temperature 2014 for Non- Hispanic White, Black and Median Household Income are all less than 0.05. It indicates that all four independent variables are significant to sales in dollars.

The coefficient of each independent variable indicates that an increase of one degree in Average Temperature would increase the sales by $2023.363. Every one thousand increase in population of Non-Hispanic White would increase the sales dollars by

$12.397 and every one thousand population increase in Black would increase the sales dollars by $15.428. For every dollar increase in Median Household Income, it would increase the sales dollars by $2.293.

Table 16: Multiple Regression Analysis of Total in Sales Dollars

Coeff's. S.E. t Stat P-value

American Indian or Alaskan Native -16.416 87.404 -0.188 0.852

Asian 53.145 35.935 1.479 0.147

Native Hawaiian or Pacific Islander -474.755 369.455 -1.285 0.207

Mixed race -216.148 181.039 -1.194 0.240

61