Name (Print): Student U#: KEY Signature: STA 3024 — FOURTH TEST — Spring 2023 — Rimbey Instructions: Print your name above and provide your signature and U number. This test consists of 17 problems worth the points indicated by the problem. Please make sure you have all 17 problems. Show all your work on these pages. Do not use scrap paper. Only approved scientific calculators are allowed on this test; all others will be taken away and incur a penalty. Calculators can NOT be shared. Cell phones must be turned off and kept out of sight. Problem | Score TOTAL1. @ points) In linear regression analysis, why do we usually skip hypothesis testing for the y-intereept? The yr bntercegt (ct X20) often makes no Sense ta the context of a Creal world problam : 2. (4 points) How are the explained vari variation related? explained + ner pl ed = tote nthe unexplained variation, and the total 3. (3 points) The general multiple regression equation is. j= b+ mx, +m, x, +--+ m,x, Recall that the StatCrunch output for multiple regression includes a box called Parameter Estimates, How many lines will this box include for the general multiple regression equation shown above? Alternatively, how many parameters are there for the general multiple regression equation? \ yc Answer, ke +4 (qarameters are bymy nay? 4. (4 points) (fill in the blanks) Response variables can also be called Ag penllont arial Explanatory variables can also be called 2 Lg pe dent Vat iables 5. (3 points) (multiple choice) What does it mean for two variables x and y to have a bivariate normal distribution? (@) For any fixed value of y the corresponding values of x are normally distributed. (b) For any fixed value of x the corresponding values of y are normally distributed. (0) For any fixed value of y the corresponding values of x are normally distributed and for any fixed value of x the corresponding values of y are normally distributed. (@) The values of x and y are both normally distributed Answer: C6. (3 points) What type of correlation is indicated in the scatterplot below? ne corcelation 15 acicatsd 7. (5 points) Suppose a regression analysis is run between two variables that have a negative correlation. It is found that 89.9% of the variation can be explained by the ‘felationship between the two variables. What is the value of the correlation coefficient? Explain how you are determining your answer. 59.4 Jo explsrced manns the coeff, of derermaratien (>= 994, {len ME te correlation coef, car Since the corralakion ss neqatile y C= fama =Laas) aa 8. (4 points) Suppose the regression equation $= b +m,x, +m,x, is used to model a certain data set. Ifthe value of 3 decreases by 33 units when the value of x, is increased by three units (while holding the value of x, constant), then what is the value of m,? Explain your answer.9. @ points) (multiple choice) Which of the following is correct? Answer: Py (a) The value of Ry is always less than R°. (b) The value of Ri, is always greater than R* (©) The value of R2, can be greater than or less than R? , depending on the values of n and k. 10. (6 points) In the scatterplot shown below, is the correlation coefficient positive or negative? Why? If the data point at (170, 64) is removed, will the correlation coefficient increase or decrease? Why? UL) + 1S 70 ive since the pints ef] J . trend upwards Gram lt de cig hk i c tt WTF (170,00) 5 removed thea he ast. * Cemoint oats ale mote gal ey ‘ - i ee day cesar C oll ‘ncheast tea 0” Sarin weighs ie llograms) 11. G points) If we have a multiple regression model with five response variables, what should the minimum sample size be to expect reliable results from the multiple regression equation? answer 50. (#10 *(H of dake, a)12. (6 points) The figure below shows five data points and a line. y (a) Looking at the figure, why is it unlikely that the line shown is the regression line for the five data points? for He cearesson lines zcAd&=O0 ose er he qorts shorn, T& (Le drdsl >|d,r4al) (b) To determine the correct regression line for the five data points, what must be minimized? — We nsnenize| TAC13. (9 points) Suppose the value of the correlation coefficient for a set of 20 data pairs is 7 =0,523 Use Table 5 and a hypothesis test to determine if there is a significant linear correlation between the data, Assume a = 0.02. Your test should include the usual ingredients: Statement of null and alternative hypotheses; critical value(s); rejection region(s); standardized test statistic; decision to reject or fail to reject Hy. tpeOe Wt il, £ aa g #0 em The $ with 7.02) we (=n, and 90 ~talod poe ye 2 Ag PSAP 1SSL S82 eyes ico Na cdized test stot is Le Shonda test st * Ayan yieee* $23 ates t- Fast 2.065) 20-2 Saree 2.603 92-552 y dense Ke14, (7 points) The figure below shows a regression line, the prediction intervals, and the confidence intervals for a set of data points. conhi Qe (a) Which curves correspond to the prediction intervals and which correspond to the confidence intervals? You can indicate your answers on the figure. (b) As indicated in the figure, the curves are further apart from the regression line at the left and right ends than they are toward the middle. This is always true, not just for the problem shown. At what point will the curves always be closest to the regression line? (c) If you are trying to estimate the mean of the y-values at a particular value of x, would you use a prediction interval or a confidence interval? Why? =x zr ® GR 3) lcd Wse a conGdence yaterval he cane the mean of de ye eben wa poraneter15. (16 points) The heights (in feet) and the numbers of stories of the nine tallest buildings in Houston, Texas, are shown in the table below. 901 | 780 | 762 | 756 S 8 Height, x | 1002 | 992 741 | 735 47 & + g Stories, y | 75 7 [6 36 | 53 The slope of the regression line for this data is 0,092 and the standard error of estimate is 5, = 2.642. (You do not have to show these.) (a) What are the values of })y, 7x, and "x? ? Show how you are determining your answers. (b) What is the value of the y-intercept (to three decimal places)? Show how you are determining your answer (©) What is the equation of the regression line? (d) Show that the regression line passes throught the point (%,¥). (©) Do a hypothesis test with o = 0.01 to test the claim that the population slope does not equal zero. Your test should include the usual ingredients: Statement of null and alternative hypotheses; critical value(s); rejection region(s); standardized test statistic; decision to reject or fail to reject Hy Cm zy BaP => 44 F44F = SIG Dy = 100% + ALA --- + FHS = P42) 2 he 215,254 e D y= (000- FAZ b= + FAN + F3S - Kriss 4 [) From Fambin Heat, b= 28 -m = Sb (0%) HE Lg, 20) (6) \ mnetl 4pore x 16. 52b\ B24 556 - 2 BAU ‘ (ay a Ex 2a = Bit More Space =>han wxrh=,092(424,5S6) 18.526 ~ 57,333 =) 4 (ed y+ M=0> bo MEO Eran Toble. $, wh £7.01 46 = 7 9-20, fond edo-terled we yr te7E3A94 a tages OSA] Nhe sest 5 tetis te is a fee = or. Jip sr-F(oan = 03482t AGEL. 2UL Smee \0,80 > 3.499 we Aga tte16. (9 points) A multiple linear regression model is used to find how employee salaries at a company are related to the length of employment (in years), previous experience (in ‘years), and education (in years). A sample of eight employees is selected from the company. Part of the output produced by StatCrunch is shown below. Analysis of variance table for multiple regression model: Source DF SS MS Estat P-value Model 3 29231989 97439964 2.403755 0.0058 Enor 4 1739707.7 434926.92 Total = 7-:30971697 What are the values of each of the following four quantities: (a) Explained Variation, (b) Unexplained Variation, (c) Coefficient of Determination, and (d) Standard Error of Estimate? Show (and/or explain) how you are determining your answers. Round to three decimal places, as needed. (©) Exgladoadl Uariection = 35.4.1 = 29231881 (LY Une xploned Ueriaton= Berge 7 ii73ar03. 4 Bross 2 24,884 Aote\ 36,071,644 i Le) Cock. of Determinakion= BOs Unaagbanel {ack (239,707, + 4 = : 2/4344 264 (F35=\_ oH 26AN OStanback Error of Ext = 5n= 1017. (13 points) The final exam scores of six randomly selected statistics students and the numbers of hours they studied for the exam are shown in the table below. Can you conclude there is a correlation between the scores and the time spent studying? Use a =0.05. Hours | 12 4 9 6 10 15 Score | 85 95 100 80 90 B (a) Calculate the test statistic (to three decimal places). Show your work. Use the table below to help with your calculation. [- Hours [Rank Score d 12 5S z 4 as -4 100 3 ¢ 50 oO 19 40 ° 1S 7S o L(say (b) State Hy and H, both mathematically and verbally. Identify the claim. (c) Find the critical value(s) and state the rejection region(s). (@) Should you reject or fail to reject the null hypothesis? Why? ay Drathremadica ty = He @s= O 5 Me ), +O ee U, :Thew io no ac rd on Here on cade Ik is Radon _ (ed Frm Tobe 10 ytheeritieal yalna is Rejection Tego s\Ve lp 886 (a) Wal = |asu3\ =. S434. 886 5 fal coject