A sales manager develops a regression model to explain the monthly sales volume of the company using the monthly advertising expenditure and the inflation rate in the economy as the predictor variables. Which of the following hypotheses should be used to test whether the overall model explains any variation in sales at all? a. H0: 1 and 2 are not equal to zero; HA: 1 = 2 = 0 b. H0: 1 2; HA: 1 < 2 c. H0: 1 = 2 = 0; HA: at least one coefficient is not zero d. H0: 1 2 = 0; HA: 1 = 2 An IT project manager wants to predict the average productivity of her software engineers, specifically the relationship between the number of lines of code written by engineers and the time taken to write them (hrs). Using past data on roughly 51 engineers, she establishes the following relationship: Estimated time (hrs) = 8 + 0.05 lines. The manager checks the residual plots and concludes that all assumptions, including that about the distribution of the error term, are satisfied. She also finds that 90% of the residuals lie between 5 hrs. What would be her APPROXIMATE 95% prediction interval for the time taken by one software engineer to write 1700 lines of code? a. [81,87] b. [87,99] c. [74,81] d. [77,89] The COO of a retail pharmacy chain is concerned about inventory record inaccuracy and installs RFID technology in some retail outlets to correct the situation. To understand whether the new technology indeed reduced inaccuracies he asks an analyst to study the dataset of 43 retail stores collected at the end of a fiscal year when physical inventory was checked. The analyst runs the following regression model (at the store level) on the dataset: Response variable (Y): %inaccuracy in inventory records Predictor variables (X): number of SKUs, on-hand inventory for each SKU, total on- hand inventory over all SKUs, RFID (a dummy variable, which equals 1 if the store has RFID technology and equals 0 if not), and number of employees. Which of the following statements is TRUE in this context of this model? a. Multicollinearity is not a concern because there are no interaction variables among the predictor variables b. The coefficient of RFID, gives us a point estimate of the %inaccuracy of inventory records, on average, in stores with RFID technology c. Multicollinearity is a concern and it can be detected by running regressions of one predictor variable at a time against all other predictor variables d. Multicollinearity is not a concern since we have categorical predictor variables Clear my choice An HR manager is attempting to find out the association of the current salary of the employees (in thousands of dollars) with their total full time work experience since college (in months) and their salary in the previous job. Parameter estimates from the regression output based on the data from 42 random employees are: At the 5% significance level, which of the following claims cannot be endorsed? In the population data, a. An additional month of experience is associated with more than $1000 increase in current salary on average keeping previous salary constant b. An additional month of experience is associated with less than $2000 increase in current salary on average keeping previous salary constant c. An additional month of experience is associated with change in current salary on average keeping previous salary constant d. An additional month of experience is associated with increase in current salary on average keeping previous salary constant