4. When you share your concerns from Question 1 with Maria, she quickly says not too worry because the sample taken for the in-depth interviews was probably a much bigger problem. Specifically, Mssr. Tremain took a convenience sample of 16 people affiliated with the company and/or living nearby in a posh urban suburb. Make matters worse was the fact that the women interviewees were significantly older (mean age 45) than the men interviewees (mean age 35). Age, therefore, confounded gender and risked artificially inflating differences between male/female adventure-travel interest. Maria’s point was that although the depth interviews were probably not the best, at least they moved the company forward to a broader piece of research that MarketKnow would surely do better, which it did.Two key questions and results surfaced from MarketKnow’s survey: The first was how much general interest there was in adventure-oriented travel on a scale from 1-7, 7 being a great deal of interest. Although older men were somewhat more interested in adventure travel than older women, the agency was targeting people under the age of 35 where there was very little difference between men and women and where the overall mean was 5.10. Although Chris was hoping for a higher number, MarketKnow pointed out that 5.10 was statistically significantly above the scale’s neutral mid-point of 4.0 (t(299) = 3.48, p = .0003).Sidebar: This was tested with a single-sample or single-mean t-test, which tests if a mean differs significantly from 0. Therefore, to test if 5.10 is significantly larger than 4.0, the researchers subtracted the scale’s mid-point of 4.0 from everyone’s rating such that the new mean was 1.10 (and would have been 0 had it been identical to 4.0). The test showed that 1.10 was significantly larger than 0, such that 5.10 was significantly larger than 4.0, which was the correct way to conduct the test.The second focal question and result involved the degree to which people would seek help from an adventure-travel-oriented agency when considering adventure travel. The mean on this 1-7 scale was only 4.3 and failed to differ significantly from the neutral mid-point, which left the management team discouraged, Chris in particular.To help, you asked MarketKnow how many people were top-box responders (6’s and 7’s) to the question on interest in adventure travel, which turned out to be 80 people or 25% of the sample. You then asked how many of those 80 were top-box responders on the question about interest in getting help with adventure travel from an adventure-travel agency, which turned out to be 24 or 30% of the adventure-travel seekers (8% of the full sample). Among the 220 less travel-oriented respondents, the top-box rates for seeking adventure-oriented travel (if they were interested in such travel) was 50 or 22.7%.What does your contingency table test say about these two proportions (show your tables and results as in Question 3), and how relevant are the results to Aventurier’s decision about whether to stay in the adventure travel business? What would you tell Chris and why?7. It is now five years later and Aventurier’s two ventures are doing well, so much so that the board of directors is seeking novel growth opportunities. One such opportunity under consideration consists of travel-oriented experience films, fictional films with a touch of realism revolving around a band of travels who show up repeatedly across the films in different situations (not the same travelers in each film but some overlap). Although the films are not necessarily sequels, or even series, the concept is related to the idea of seeing a mix of new and known characters over time such that it shares some elements with sequels.To investigate some of the features that determine film and sequel success, your boss asks you to analyze some film data from a former professor of yours (currently sipping Belvedere vodka at Tryst Gastro Lounge, St. Petersburg, Florida; or at least thinking of that). You will find a data set in an Excel spreadsheet in our course Module consisting of the following seven columns:1: Film: Each individual film in the data set numbered from 1-2982: Date: The date the film opened3: Franchise: Number given to each of 89 franchises (each franchise has multiple films/sequels)4: Revenue: Box office revenues from each of the 298 films5: Quality: Film quality as judged by a number of film critics (1-12 scale)6: FilmNum: The number of the film in its franchise’s sequence (e.g., 4 = the third sequel)7: Screens: The number of U.S. movie screens on which the film openedThe Excel file is called Film Data and comes with two pre-programmed features: (1) a bivariate correlation matrix consisting of the dependent measure (Revenues) and three potential predictors you are to assess; Quality, FilmNum, and Screens, and (b) at the bottom of the revenue column is found the mean film revenues, the standard deviation (SD) of the revenues, and the value of the mean +3 SD’s.It turns out that the board of directors is especially interested in the potential effects of the number of screens on which a film opens. The reason is that one of the board members is closely tied (i.e., owns) a large U.S. cinema chain that would be receptive to favorable treatment for a novel film franchise focused on adventure themes. Part of the reason for the study, therefore, is to assess how much additional opening screens might help our films.You now conduct your analyses. As one suggestion, keep your Film Data file separate from your analyses of that file, saving it anew each time under a different name so that you always have a clean place to start in the future.a. Evaluate the bivariate correlation matrix and indicate what it suggests preliminarily about the three predictors’ potential effects on box office sales (revenues)?b. Run and report a standard multiple regression model estimating the effects of Quality, FilmNum, and Screens on Revenues as we did in class and as is outlined in the spreadsheet, EXCEPT PLEASE CHECK THE REGRESSION BOX’S OPTION, “Line Fit Plots.”Please report: (i) the percentage of the variation in film revenues that your model can account for, or statistically explain, plus (ii) the three predictors’ regression coefficients and associated p values? What do they suggest about the predictors’ impact on film sales, especially the potential effect of the number of screens on which the film opens?c. Click on the Screens Line Fit Plot and pull it away from the other two plots so that it has plenty of room to expand. Then click on one or more of the plot’s corners and pull the plot on a diagonal to stretch it to a much larger but proportional size (try not to distort the plot). Do you see those two blue dots high above the 0 and 1 on the x-axis that are around $1billion in sales? They’re a problem.When modeling data, we sometimes focus on and interpret outliers because they can tell the most interesting stories. And that might be the case here too, as we will see in a moment, but for our board of directors’ question about the effects that screens typically have on sales, these two outliers may be a problem because they had very few opening screens combined with unusually high sales.Researchers interested in cleaning up data to identify more typical sorts of patterns and effects sometimes use rules such as, “drop values beyond the mean +/? 3-Standard Deviations. Find the Screen Line Fit Plot’s two outliers at 0 and 1 and compare them to the value found at the bottom of the data’s Revenue column, the mean revenue plus 3 SD’s. Do those two points qualify as outliers as defined by the mean plus/minus three SD’s rule? How many SD’s is the highest grossing film above the mean film revenue (i.e., what is its z-score)?d. Completely delete those two outlier films (rows) from the data set and re-estimate your multiple regression model, being sure to adjust the number of films modeled which is now 296 instead of 298 (re-specify the independent and dependent variable vectors). What is the new percentage of revenues explained by the model and has it gone up or down? How much has the screens’ coefficient changed relative to the original coefficient, and what does it and its p-value say about the Screens’ potential effect? What do you tell the board about that?e. When a board member realizes you dropped two films from the model you reported, he asks how you can justify this and whether the outliers have any meaning for the board. (i) What is your justification? (ii) What were those two outlier films? (iii) And what do those two outlier films teach us about when lots of screens might not be needed for films to be successful?