A researcher is studying the factors related to the popularity of new music artists. In particular, she is interested in assessing the extent to which the number of albums sold (thousands) is explained by the number of YouTube views (millions) the artists currently have, their number of Facebook fans (millions), and the number of weeks they did not make the coveted JJ's Top 50 list in the last 12 months. A random sample of 40 popular new artists was collected (see partial data in the following table). New YouTube views Facebook fans Weeks off Albums sold Artist (millions) millions) Top 50 (000) 144 7.94 21.6 53.292 2 281.57 18.69 17.5 93.505 3 176.5 3.21 36 14.35 The researcher used software to generate a multiple linear regression output for the data. The results are shown below (some values are missing). a) Determine the missing values. Report answers accurate to at least 5 decimal places where necessary. Do not round intermediate results. Multiple R R Square Adjusted R2 Standard Error Observations df SS MS F Regression 7324.7728 213.8465 Residual 36 34.2525 Total 39 Coefficient Standard Error t-stat p-value Intercept 29.426822 17.527941 0.1018 YouTube 0.262766 21.19764 0.0000 Facebook 0.500078 1.0479 0.3016 Top 50 -0.873502 -2.03038 0.0498 b) What percent of the variation in the dependent variable is not explained by the independent variables in this model? Round percent to 1 decimal place c) Use the regression coefficients from the output to estimate the number of albums sold (units) for a new artist with 29 million YouTube views, having 1,984,362 Facebook fans, and with exactly 8.4 months off JJ's Top 50 list. [Hint: 1 year = 52 weeks] d) Suppose the regression equation from the output is used to predict the number of albums sold by a new artist to be 50,595. If the artist has 139 million YouTube views and has successfully made JJ's Top 50 list 62.5% of the time in the last 12 months, how many Facebook fans (in thousands) does the artist have? Round to the nearest thousand