Chapter 13: Dummy and Interaction Variables

Size: px
Start display at page:

Download "Chapter 13: Dummy and Interaction Variables"

Transcription

1 Chapter 13: Dummy and eraction Variables Chapter 13 Outline Preliminary Mathematics: Averages and Regressions Including Only a Constant An Example: Discrimination in Academia o Average Salaries o Dummy Variables o Models Type 1 Models: No explanatory variables; only a constant. Type 2 Models: A constant and a single dummy explanatory variable denoting sex. Type 3 Models: A constant, a dummy explanatory variable denoting sex, and other explanatory variable(s). o Beware of Implicit Assumptions o eraction Variables o Conclusions Beware of Averages Power of Multiple Regression Analysis Flexibility of Multiple Regression Analysis An Example: ernet and Television Use o Similarities and Differences o eraction Variable: Economic and Political eraction Chapter 13 Prep Questions 1. Recall our first regression example, Professor Lord s quiz: Minutes Quiz Studied Student Score (y) (x) Consider the most simple of all possible models, one that does not include even a single explanatory variable: Model: y t = Const + e t b Const denotes the estimate of Const : The Estimates: Esty t = b Const Residuals: Res t = y t Esty t The sum of squared residuals equals: SSR = Res + Res + Res = ( y b ) + ( y b ) + ( y b ) Const 2 Const 3 Const

2 2 Using calculus derive the equation for b Const that minimizes the sum of squared residuals by expressing b Const in terms of y 1, y 2, and y Consider the following faculty salary data: 1 Faculty Salary Data: Artificially generated cross section salary data and characteristics for 200 faculty members. Salary t Salary of faculty member t (dollars) Experience t Teaching experience for faculty member t (years) Articles t Number of articles published by faculty member t SexM1 t 1 if faculty member t is male; 0 if female You can access these data by clicking on the following link: [Link to MIT-FacultySalaries.wf1 goes here.] a. What is the average salary for all 200 faculty members? b. What is the average salary for the men? c. What is the average salary for the women? Getting Started in EViews For all faculty members: In the Workfile window: double click Salary. In the Workfile window: click View, then click Descriptive Statistics, then click Histogram and Stats. For men only: In the Workfile window: click Sample. To include only men, enter SexM1 = 1 in the If condition window. Click OK. For women only: In the Workfile window: click Sample. To include only women, enter SexM1 = 0 in the If condition window. Click OK. NB: Do not forget to turn off the sample. In the Workfile window: click Sample. Clear the If condition window. Click OK. d. Consider the following model: Salary = Const + e t What is the value of the estimated constant? Getting Started in EViews

3 3 To estimate the model, you must trick EViews into running the appropriate regression: In the Workfile window: highlight Salary and then while depressing <Ctrl> highlight one other variable, say SexM1. In the Workfile window: double click a highlighted variable. Click Open Equation. In the Equation Specification window delete SexM1 so that the line specifying the equation looks like this: salary c Click OK. e. Now consider a second model: Salary t = Const + SexM1 SexM1 t + e t Run the appropriate regression to estimate the values of the constant and coefficient. What is the estimated salary for men? What is the estimated salary for women? f. Compare your answers to d and e with your answers to a, b, and c. What conclusions can you draw concerning averages and the regression estimates? 3. Consider the following model explaining ernet use in various countries: LogUsersernet = + Year + CapitalHuman + t Const Year t CapHum t CapitalPhysical + Gdp + Auth + e CapPhy t GDP t Auth t t where LogUsersernet t Logarithm of ernet users per 1,000 people for observation t Year t Year for observation t CapitalHuman t Literacy rate for observation t (percent of population 15 and over) CapitalPhysical t Telephone mainlines per 10,000 people for observation t GdpPC t Per capita real GDP in nation t (1,000 s of international dollars) Auth t The Freedom House measures of political authoritarianism for observation t normalized to a 0 to 10 scale. 0 represents the most democratic rating and 10 the most authoritarian. During the period, Canada and the U.S. had a 0 rating; Iraq and the Democratic Republic of Korea (North Korea) rated 10. a. Note that the dependent variable is the logarithm of ernet users. erpret the coefficient of Year,. Year

4 4 b. Develop a theory that explains how each explanatory variable affects ernet use. What do your theories suggest about the sign of each coefficient? 4. Consider a similar model explaining television use in various countries: LogUsers = + Year + CapitalHuman + where LogUsers t t Const Year t CapHum t CapitalPhysical + Gdp + Auth + e CapPhy t GDP t Auth t t Logarithm of television users per 1,000 people for observation t a. Develop a theory that explains how each explanatory variable affects television use. b. Based on your theories which coefficients should be qualitatively similar (have the same sign) as those in the ernet use model and which may be qualitatively different?

5 5 Preliminary Mathematics: Averages and Regressions Including Only a Constant Before investigating the possibility of discrimination in academia, we shall consider a technical issue that will prove useful. While a regression that includes only a constant (that is, a regression with no explanatory variables) is not interesting in itself, it teaches us an important lesson. When a regression includes only a constant, the ordinary least squares (OLS) estimate of the constant equals the average of the dependent variable s values. A little calculus allows us to prove this: Model: y t = Const + e t Estimates: Esty t = b Const Residuals: Res t = y t Esty t Now, compute the sum of the squared residuals: SSR = Res1 + Res2 + Res3 = ( y1 Esty1) + ( y2 Esty2) + ( y3 Esty3) = ( y b ) + ( y b ) + ( y b ) 1 Const 2 Const 3 Const To minimize the sum of squared residuals, differentiate with respect to b Const and set the derivative equal to 0: dssr dbconst = 2( y1 bconst ) 2( y2 bconst ) 2( y3 bconst ) = 0 Divide by 2 y1 bconst + y2 bconst + y3 bconst = 0 Rearranging terms. y1 + y2 + y3 = 3b Const Dividing by 3. y + y + y = b y1+ y2 + y3 equals the mean of y, y. 3 y = b Const We have just shown that when a regression includes only a constant the ordinary least squares (OLS) estimate of the constant equals the average value of the dependent variable, y. Const

6 6 An Example: Discrimination in Academia Now, we consider faculty salary data. It is important to keep in mind that these data were artificially generated; the data are not real. Artificially generated, rather than real, data are used as a consequence of privacy concerns. Faculty Salary Data: Artificially generated cross section salary data and characteristics for 200 faculty members. [Link to MIT-FacultySalaries.wf1 goes here.] Salary t Salary of faculty member t (dollars) Experience t Teaching experience for faculty member t (years) Articles t Number of articles published by faculty member t SexM1 t 1 if faculty member t is male; 0 if female Project: Assess the possibility of discrimination in academia. We begin by examining the average salaries of men and women. Average Salaries First, let us report the average salaries: Both males and females $82,802 Males only 91,841 Females only 63,148 Difference 28,693 On average, males earn nearly $30,000 more than females. This certainly raises the possibility that gender discrimination exists, does it not? Dummy Variables A dummy variable separates the observations into two disjoint groups; a dummy variable equals 1 for one group and 0 for the other group. The variable SexM1 is a dummy variable; SexM1 denotes whether a faculty member is a male of female; SexM1 equals 1 if the faculty member is a male and 0 if female. We shall now show that dummy variables prove very useful in exploring the possibility of discrimination by considering three types of models: Type 1 Models: No explanatory variables; only a constant. Type 2 Models: A constant and a single dummy explanatory variable denoting sex. Type 3 Models: A constant, a dummy explanatory variable denoting sex, and other explanatory variable(s).

7 7 Models Type 1 Models: No explanatory variables; only a constant. We begin with a model that includes no explanatory variables; that is, a model that includes only a constant. We are doing this to confirm our conclusion that a regression with only a constant is equivalent to computing an average of the dependent variable. Step 0: Formulate a model reflecting the theory to be tested. Model: Salary t = Const + e t Since this model includes only a constant, we are theorizing that except for random influences each faculty member earns the same salary. That is, this model attributes all variations in income to random influences. Step 1: Collect data, run the regression, and interpret the estimates. Getting Started in EViews To estimate the model, you must trick EViews into running the appropriate regression: In the Workfile window: highlight Salary and then while depressing <Ctrl> highlight one other variable, say SexM1. In the Workfile window: double click a highlighted variable. Click Open Equation. In the Equation Specification window delete SexM1 so that the window looks like this: salary c Click OK. Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob Const Number of Observations 200 Table 13.1: Discrimination Regression Results All Observations This confirms the fact that when a regression only includes a constant, the ordinary least squares (OLS) estimate of the constant is just the average of the dependent variable. To emphasize this fact, we shall now run two more regressions with only a constant: one regression including only men and one including only women:

8 8 Dependent variable: Salary Explanatory variables: None Sample: SexM1 = 1 Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob Const Number of Observations 137 Sample SexM1 = 1 Table 13.2: Discrimination Regression Results Males Only Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob Const Number of Observations 63 Sample SexM1 = 0 Table 13.3: Discrimination Regression Results Females Only Compare the regression results to the salary averages: Both males and females $82,802 Males only 91,841 Females only 63,148 Tables 13.1, 13.2, and 13.3 illustrate the important lesson that Type 1 models teach us. In a regression that includes only a constant, the ordinary least squares (OLS) estimate of the constant is the average of the dependent variable. Next, let us consider a slightly more complicated model.

9 9 Type 2 Models: A constant and a single dummy explanatory variable denoting sex. Step 0: Formulate a model reflecting the theory to be tested. Salary t = Const + SexM1 SexM1 t + e t where SexM1 equals 1 for males and 0 for females Discrimination Theory: Women are discriminated against in the job market; hence, men earn higher salaries than women. Since SexM1 equals 1 for males and 0 for females, SexM1 should be positive indicating that men will earn more than women: SexM1 > 0. Step 1: Collect data, run the regression, and interpret the estimates. Using the ordinary least squares (OLS) estimation procedure to estimate the parameters: Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob SexM Const Number of Observations 200 Estimated Equation: EstSalary = 63, ,693SexM1 erpretation of Estimates: b SexM1 = 28,693: Men earn $28,693 more than women. Critical Result: The SexM1 coefficient estimate equals 28,693. This evidence, the positive sign of the coefficient estimate, suggests that men earn more than women thereby supporting the discrimination theory. Table 13.4: Discrimination Regression Results Male Sex Dummy For emphasis, let us apply the estimated equation to men and then to women by plugging in their values for SexM1: Estimated Equation: EstSalary = 63, ,693SexM1 We can now compute the estimated salary for men and women: For men For women SexM1 = 1 SexM1 = 0 EstSalary Men = 63, ,693= 91,841 EstSalary Women = 63, = 63,148

10 10 Next, note something very interesting by comparing the regression results to the salary averages: Males only 91,841 Females only 63,148 Difference 28,693 An ordinary least squares (OLS) regression that includes only a constant and a dummy variable is equivalent to comparing averages. The conclusions are precisely the same: men earn $28,693 more than women. The dummy variable s coefficient estimate equals the difference of the averages. Step 2: Play the cynic and challenge the results; construct the null and alternative hypotheses. Cynic s view: Despite the results, there is no discrimination. H 0 : SexM1 = 0 Cynic is correct: No discrimination H 1 : SexM1 > 0 Cynic is incorrect: Discrimination in favor of men, against women Step 3: Formulate the question to assess the cynic s view and the null hypothesis. Generic Question for Discrimination Hypothesis: What is the probability that the results would be like those we obtained (or even stronger), if the cynic is correct and no discrimination were present? Specific Question for Discrimination Hypothesis: What is the probability that the coefficient estimate, b SexM1, in one regression would be 2,240 or more, if H 0 were true (if the actual coefficient, SexM1, equals 0)? Steps 4 and 5 It is easy to calculate the Prob[Results IF H 0 True] by using the tails probability reported in the regression printout. Since this is a one-tailed test, we divide the tails probability by 2: < Prob[Results IF H 0 True] = =< Clearly, the Prob[Results IF H 0 True] is very small. We can reject the null hypothesis which asserts that no discrimination exists.

11 11 Before we continue, let us point out that our dummy variable, SexM1, assigned 1 to males and 0 to females. This was an arbitrary choice. We could just as easily assigned 0 to males and 1 to females, could we not? To see what happens when we switch the assignments, generate a new variable, SexF1: SexF1 = 1 SexM1 For men For women SexM1 = 1 SexM1 = 0 SexF1 = 1 1 = 0 SexF1 = 1 0 = 1 Step 0: Formulate a model reflecting the theory to be tested. Salary t = Const + SexF1 SexF1 t + e t where SexF1 t = 1 if faculty member t is female; 0 if male Discrimination Theory: Women are discriminated against in the job market; hence, women earn lower salaries than men. Since SexF1 equals 1 for females and 0 for males, SexF1 should be negative indicating that women will earn less than men: SexF1 < 0. Step 1: Collect data, run the regression, and interpret the estimates. After we generate the new dummy variable, SexF1, we can easily run the regression: Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob SexF Const Number of Observations 200 Estimated Equation: EstSalary = 91,841 28,693SexF1 erpretation of Estimates: b SexF1 = 28,693: Women earn $28,693 less than men. Critical Result: The SexF1 coefficient estimate equals 28,683. This evidence, the negative sign of the coefficient estimate, suggests that women earn less than men thereby supporting the theory. Table 13.5: Discrimination Regression Results Female Sex Dummy

12 12 Let us apply this estimated equation to men and then to women by plugging in their values for SexF1: For men For women SexF1 = 0 SexF1 = 1 EstSalary Men = 91,841 0 = 91,841 EstSalary Women = 91,841 28,693 = 63,148 The results are precisely the same as before. This is reassuring. The decision to assign 1 to one group and 0 to the other group is completely arbitrary. It would be very discomforting if this arbitrary decision affected our conclusions. The fact that the arbitrary decision does not affect the results is crucial. Step 2: Play the cynic and challenge the results; construct the null and alternative hypotheses. Cynic s view: Despite the results, there is no discrimination. H 0 : SexF1 = 0 Cynic is correct: No discrimination H 1 : SexF1 < 0 Cynic is incorrect: Discrimination in favor of men, against women The null hypothesis, like the cynic, challenges the evidence. The alternative hypothesis is consistent with the evidence. Steps 3, 4, and 5 It is easy to calculate the Prob[Results IF H 0 True] by using the tails probability reported in the regression printout. Since this is a one-tailed test, we divide the tails probability by 2: <.0001 Prob[Results IF H 0 True] = =< Since the probability is so small, we reject the null hypothesis that no discrimination exists. Bottom Line Our choice of the base group for the dummy variable (that is, the group that is assigned a value of 0 for the dummy variable) does not influence the results. Type 2 models, models that include only a constant and a dummy variable, are equivalent to comparing averages.

13 13 Question: Do Type 2 models provide convincing evidence of gender discrimination? On the one hand, yes: o The dummy variable coefficients suggest that women earn less than men. o The dummy variable coefficients are very significant the probability of obtaining results like we obtained if no discrimination exists is less than On the other hand, what implicit assumption is this discrimination model making? The model implicitly assumes that the only relevant factor in determining faculty salaries is gender. Is this reasonable? Well, very few individuals contend that gender is the only factor. Many individuals believe that gender is one factor, perhaps an important factor, affecting salaries, but they believe that other factors such as education, experience, etc. also play a role. Type 3 Models: A constant, a dummy explanatory variable denoting sex, and other explanatory variable(s). While these models allow the possibility of gender discrimination, they also permit us to explore the possibility that other factors affect salaries too. To explore such models, let us include both a dummy variable and the number of years of experience as explanatory variables. Step 0: Formulate a model reflecting the theory to be tested. Salary t = Const + SexF1 SexF1 t + Exper Experience t + e t Theories: Discrimination: As before, we theorize that women are discriminated against: SexF1 < 0. Experience: It is generally believed that in most occupations, employees with more experience earn more than employees with less experience. Consequently, we theorize that the experience coefficient should be positive: Exper > 0.

14 14 Step 1: Collect data, run the regression, and interpret the estimates. Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob SexF Experience Const Number of Observations 200 Estimated Equation: EstSalary = 42,238 2,240SexF1 + 2,447Experience erpretation of Estimates: b SexF1 = 2,240: Women earn about $2,240 less than men after accounting for experience. b Exper = 2,447: Each additional year of experience results in a $2,447 increase in salary for both men and women. Critical Result: The SexF1 coefficient estimate equals 2,240. The negative sign of the coefficient estimate suggests that women earn less than men. This evidence supports the discrimination theory. The Experience coefficient estimate equals 2,447. The positive sign of the coefficient estimate suggests that additional experience increases salaries. This evidence supports the experience theory. Table 13.6: Discrimination Regression Results Female Sex Dummy and Experience We can now compute the estimated salary for men and women: EstSalary = 42,238 2,240SexF1 + 2,447Experience For men, SexF1 = 0: EstSalary Men = 42, ,447Experience = 42, ,447Experience For women, SexF1 = 1: EstSalary Women = 42,238 2, ,447Experience 39, ,447Experience

15 15 We can illustrate the estimated salaries of men and women graphically: EstSalary EstSalary Men = 42, ,447Experience 42,238 2,240 Slope = 2,447 EstSalary Women = 39, ,447Experience 39,998 Figure 13.1: Salary Discrimination Experience Step 2: Play the cynic and challenge the results; construct the null and alternative hypotheses. Cynic s View on Discrimination: Despite the results, there is no discrimination. Cynic s View on Experience: Despite the results, experience does not increase salary. Discrimination Hypotheses Experience Hypotheses H 0 : SexF1 = 0 No discrimination H 0 : Exper = 0 Experience has no effect on salary H 1 : SexF1 < 0 Discrimination H 1 : Exper > 0 Experience increases salary The null hypothesis, like the cynic, challenges the evidence. The alternative hypothesis is consistent with the evidence. We shall proceed by focusing on discrimination.

16 16 Step 3: Formulate the question to assess the cynic s view and the null hypothesis. Generic Question for Discrimination Hypothesis: What is the probability that the results would be like those we obtained (or even stronger), if the cynic is correct and no discrimination were present? Specific Question for Discrimination Hypothesis: The regression s coefficient estimate was 2,240. What is the probability that the coefficient estimate in one regression would be 2,240 or less, if H 0 were true (if the actual coefficient, SexF1, equals 0; that is, if no discrimination existed)? Answer: Prob[Results IF H 0 True].4638/2 = ,240 0 Figure 13.2: Probability Distribution of Coefficient Estimate b SexF1 Step 4 and 5: Use the general properties of the estimation procedure, the probability distribution of the estimate, to calculate Prob[Results IF H 0 True] Prob[Results IF H 0 True] = =.23 2 At the traditional significance levels of 1, 5, and 10 percent, we cannot reject the null hypothesis that no discrimination exists. What should we make of this dramatic change?

17 17 Beware of Implicit Assumptions Focus on our last model: Salary t = Const + SexF1 SexF1 t + Exper Experience t + e t EstSalary EstSalary Men = 42, ,447Experience 42,238 2,240 Slope = 2,447 EstSalary Women = 39, ,447Experience 39,998 Experience Figure 13.3: Estimated Discrimination Equations with Lump Sum Discrimination Implicit Assumption: One year of added experience increases the salary of men and women by equal amounts. In other words, this model implicitly assumes that women start behind men by a certain amount and then remain behind men by that same amount for each level of experience. We shall call this lump sum discrimination. Figure 13.3 illustrates this well; the slopes of the lines representing the estimated salaries for men and women are equal. Might gender discrimination take another form? Yes. Experience could affect the salaries of men and women differently. It is possible that a man could receive more for an additional year of experience than a woman. In other words, might men be more highly rewarded for experience than women? Our last model excludes this possibility because it implicitly assumes that a year of added experience increases the salary of men and women by equal amounts. To explore the possibility of this second type of discrimination we shall introduce interaction variables. We shall refer to this type of discrimination as raise discrimination.

18 18 eraction Variables An interaction variable allows us to explore the possibility that one explanatory variable influences the effect that a second explanatory variable has on the dependent variable. We generate an interaction variable by multiplying the two variables together. We shall focus on the interaction of Experience and SexF1 by generating the variable Exper_SexF1: Exper_SexF1 = Experience SexF1 We shall now add the interaction variable, Exper_SexF1, to our last model. Step 0: Formulate a model reflecting the theory to be tested. Salary t = Const + SexF1 SexF1 t + Exper Experience t + Exper_SexF1 Exper_SexF1 t + e t Theories: Lump Sum Discrimination: As before, we theorize that women are discriminated against: SexF1 < 0. Experience: As before, we theorize that the experience coefficient should be positive: Exper > 0. Raise Discrimination: One year of additional experience should increase the salary of women by less than their male counterparts. Hence, we theorize that the coefficient of the interaction variable is negative: Exper_SexF1 < 0. (If it is not clear why you should expect this coefficient to be negative, be patient. It should become clear shortly.) Step 1: Collect data, run the regression, and interpret the estimates. Ordinary Least Squares (OLS) Dependent Variable: Salary Explanatory Variable(s): Estimate SE t-statistic Prob SexF Experience Exper_SexF Const Number of Observations 200 Estimated Equation: EstSalary = 37, ,970SexF1 + 2,676Experience 1,135Exper_SexF1 Table 13.7: Discrimination Regression Results Female Sex Dummy, Experience, and Female Sex Dummy-Experience eraction Variable

19 19 Now, let use apply the estimated equation to men and women: For men For women SexF1 = 0 SexF1 = 1 Exper_SexF1 = 0 Exper_SexF1 = Experience For men, EstSalary Men = 37, ,970SexF1 + 2,676Experience 1,135Exper_SexF1 = 37, ,676Experience 0 = 37, ,676Experience For women, EstSalary Women = 37, ,970SexF1 + 2,676Experience 1,135Exper_SexF1 = 37, , ,676Experience 1,135Experience = 48, ,541Experience Plotting the estimated salary for men and women: EstSalary EstSalary Men = 37, ,676Experience EstSalary Women = 48, ,541Experience 48,565 37,595 Experience Figure 13.4: Estimated Discrimination Equations with Lump Sum and Raise Discrimination

20 20 We can use this regression to assess the possibility of two different types of discrimination. One of the estimates is a little surprising: Lump Sum Discrimination: As before, the coefficient of the sex dummy variable, SexF1, assesses the possibility of lump sum discrimination. The coefficient estimate is positive. This is unexpected. It suggests that when faculty members are hired from graduate school with no experience, women receive about $10,970 more than men. The positive coefficient estimate suggests that reverse discrimination exists at the entry level. Raise Discrimination: The coefficient of the interaction variable, Exper_SexF1, assesses the possibility of this more subtle type of discrimination, raise discrimination. The coefficient estimate is negative. It suggests that a woman is receives $1,135 less than a man for an additional year of experience. The negative coefficient estimate suggests that women receive smaller annual raises than their male counterparts. These regression results paint a more complex picture of possible discrimination than is often contemplated. Again, recall that as a consequence of privacy concerns these data were artificially generated. Consequently, do not conclude that the conclusions we have suggested here necessarily reflect the real world. This example was used because it illustrates how multiple regression analysis can exploit dummy variables and interaction variables to investigate important issues, such as the presence of discrimination. Conclusions Beware of Averages: We should not consider differences in averages, by themselves, as evidence of discrimination. When we just consider average salaries, we are implicitly adopting a model of salary determination that few, if anyone, consider realistic. We implicitly assume that the only factor that determines an individual s salary is his/her sex. While many would argue that gender is one factor, very few would argue that gender is the only factor. Power of Multiple Regression Analysis: Since is it naïve to consider just averages what quantitative tools should we use to assess the presence of discrimination? Multiple regression analysis is an appropriate tool. It allows us to consider the roles played by several factors in the determination of salary and separates out the individual influence of each. Multiple regression analysis allows us to consider not only the role of gender, but also the role that the other factors may play. Multiple regression analysis sorts out the impact that each individual explanatory variable has on the dependent variable.

21 21 Flexibility of Multiple Regression Analysis: Not only does multiple regression analysis allow us to consider the roles played by various factors in salary determination, but also it allows us to consider various types of potential discrimination. The above example illustrates how we can assess the possible presence lump sum discrimination and/or raise discrimination. An Example: ernet and Television Use Next, we consider ernet and television use: Project: Assess the determinants of ernet and television use internationally. ernet and Data: Panel data of ernet,, economic, and political statistics for 208 countries from 1995 to [Link to MIT-ernetFlat wf1 goes here.] LogUsersernet t LogUsers t Year t CapitalHuman t CapitalPhysical t GdpPC t Auth t Logarithm of ernet users per 1,000 people for observation t Logarithm of television users per 1,000 people for observation t Year for observation t Literacy rate for observation t (percent of literate population 15 and over) Telephone mainlines per 10,000 people for observation t Per capita real GDP in nation t (1,000 s of international dollars) The Freedom House measures of political authoritarianism for observation t normalized to a 0 to 10 scale. 0 represents the most democratic rating and 10 the most authoritarian. During the period, Canada and the U.S. had a 0 rating; Iraq and the Democratic Republic of Korea (North Korea) rated 10.

22 22 Step 0: Formulate a model reflecting the theory to be tested. ernet Model: LogUsersernet = + Year + CapitalHuman + t Const Year t CapHum t CapitalPhysical + GdpPC + Auth + e CapPhy t GDP t Auth t t Television Model: LogUsers = + Year + CapitalHuman + t Const Year t CapHum t CapitalPhysical + GdpPC + Auth + e CapPhy t GDP t Auth t t The dependent variable in both the ernet and television models is the logarithm of users. This is done so that the coefficients can be interpreted as percentages. Similarities and Differences The theory behind the effect of human capital, physical capital, and per capita GDP on both ernet and television use is straightforward: Additional human capital, physical capital, and per capita GDP should stimulate both ernet and television use. We postulate that the impact of time and political factors should be different for the two media, however: As an emerging technology, we theorize that there should be substantial growth of ernet use over time even after accounting for all the other factors that may affect ernet use. Television, on the other hand, is a mature technology. After accounting for all the other factors, time should play little or no role in explaining television use. We postulate that the political factors should affect ernet and television use differently. Since authoritarian nations control the content of television, we would expect authoritarian nations to promote television; television provides the authoritarian nation the means to get the government s message out. On the other hand, since it is difficult to control ernet content, we would expect authoritarian nations to suppress ernet use.

23 23 Table 13.8 summarizes our theories and presents the appropriate null and alternative hypotheses: LogUsersernet LogUsers Variable Theory Hypotheses Theory Hypotheses Year Year > 0 H 0 : Year = 0 Year = 0 H 0 : Year = 0 Emerging versus H 1 : Year > 0 H 1 : Year 0 mature technology CapitalHuman CapHum > 0 H 0 : Literacy rate H 1 : CapitalPhysical CapPhy > 0 H 0 : Telephone (main) H 1 : lines per 1,000 people GdpPC GDP > 0 H 0 : Per capita real GDP H 1 : CapHum = 0 CapHum > 0 H 0 : CapHum > 0 H 1 : CapPhy = 0 CapPhy > 0 H 0 : CapPhy > 0 H 1 : GDP = 0 GDP > 0 H 0 : GDP > 0 H 1 : Auth Auth < 0 H 0 : Auth = 0 Auth > 0 H 0 : 0-democratic H 1 : Auth < 0 H 1 : 10-authoritarian Table 13.8: Theories and Hypotheses for ernet and Television Use CapHum = 0 CapHum > 0 CapPhy = 0 CapPhy > 0 GDP = 0 GDP > 0 Auth = 0 Auth > 0

24 24 As Table 13.8 reports, all the hypothesis tests are one-tailed tests with the exception of the Year coefficient in the television use model. Let us begin by focusing on ernet use. Step 1: Collect data, run the regression, and interpret the estimates. Ordinary Least Squares (OLS) Dependent Variable: LogUsersernet Explanatory Variable(s): Estimate SE t-statistic Prob Year CapitalHuman CapitalPhysical GdpPC Auth Const Number of Observations 566 Estimated Equation: EstLogUsersernet = Year +.024CapitalHuman +.002CapitalPhysical +.118GdpPC.096Auth erpretation of Estimates: After accounting for all other explanatory variables b Year =.450: A one unit increase in Year increases ernet use by 45 percent. That is, after accounting for all other explanatory variables, the estimated annual rate of ernet use is 45 percent per year. This is just the type of rapid growth we would expect for an emerging technology. b =.024: A one percentage point increase in the literacy rate, CapHum CapitalHuman, increases ernet use by 2.4 percent. b =.002: A one unit increase in telephone mainlines per 10,000 persons, CapPhy CapitalPhysical, increases ernet use by.2 percent. b GDP =.118: A 1,000 international dollar increase in gross domestic product, GdpPC, increases ernet use by 11.8 percent. b Auth =.096: A 1 unit increase in the authoritarian index, Auth, decreases ernet use by 9.6 percent. Table 13.9: ernet Regression Results Since the dependent variables are logarithms, we interpret the coefficient estimates in terms of percentages. The signs of all the coefficient estimates support our theories.

25 25 Next, estimate television use model. Ordinary Least Squares (OLS) Dependent Variable: LogUsers Explanatory Variable(s): Estimate SE t-statistic Prob Year CapitalHuman CapitalPhysical GdpPC Auth Const Number of Observations 742 Estimated Equation: EstLogUsers = Year +.036CapitalHuman +.002CapitalPhysical +.059GdpPC +.063Auth erpretation of Estimates: After accounting for all other explanatory variables b Year =.023: A one unit increase in Year increases television use by 2.3 percent. The tails probability indicates that after accounting for all other explanatory variables, we cannot reject the null hypothesis that there is no growth in television use at the traditional significance levels. This is what we would expect for a mature technology. b =.036: A one percentage point increase in the literacy rate, CapHum CapitalHuman, increases television use by 3.6 percent. b =.002: A one unit increase in telephone mainlines per 10,000 persons, CapPhy CapitalPhysical, increases television use by.2 percent. b GDP =.058: A 1,000 international dollar increase in gross domestic product, GdpPC, increases television use by 5.9 percent. b Auth =.063: A 1 unit increase in the authoritarian index, Auth, increases television use by 6.3 percent. Table 13.10: Television Regression Results

26 26 Steps 2, 3, 4, and 5: Table summarizes the remaining steps: LogUsersernet LogUsers Year.450*.023 (<.0001) (.1487) CapitalHuman.024*.036* (<.0001) (<.0001) CapitalPhysical.002*.002* (<.0001) (.0001) GdpPC.118*.059* (<.0001) (<.0001) Auth.096*.064* (<.0001) (<.0001) Prob[Results IF H 0 True] in parentheses. * indicates significance at the 1 percent level. Table 13.11: Coefficient Estimates and Prob[Results IF H 0 True] Note that all the results support the theories and all the coefficients except for the Year coefficient in the television regression are significant at the 1 percent level. It is noteworthy that the regression results suggest that the impact of Year and Auth differ for the two media just as we postulated. Our results suggest that after accounting for all other explanatory variables: ernet use grows by an estimated 45 percent whereas the growth rate of television use does not differ significantly from 0. Increases in the authoritarian index results to a significant decrease ernet use, but a significant increase television use.

27 27 eraction Variable: Economic and Political eraction Next, let us investigate the following question: Question: Does per capita GDP have a greater impact on ernet use in authoritarian nations than non-authoritarian ones? Some argue that the answer to this question is yes; that is, that per capita GDP has a greater impact on ernet use in authoritarian nations. Their rationale is based on the following logic: In authoritarian nations, citizens have few sources of uncensored information. There are few, if any, uncensored newspapers, news magazines, etc. available. The only source of uncensored information is the ernet. Consequently, the effect of per capita GDP on ernet use will be large. In non-authoritarian nations, citizens have many sources of uncensored information. Higher per capita GDP will no doubt stimulate ernet use, but it will also stimulate the purchase of uncensored newspapers, news magazines, etc. Consequently, the effect on ernet use will be modest. An authoritarian index-gdp interaction variable can be used to explore this issue. To do so, generate the interaction variable Auth_GdpPC, the product of the authoritarian index and per capita GDP: Auth_GdpPC = Auth GdpPC Step 0: Formulate a model reflecting the theory to be tested. Add this interaction variable to the ernet model: LogUsersernet = + Year + CapitalHuman + t Const Year t CapHum t CapitalPhysical + GdpPC + Auth + Auth _ GdpPC + e CapPhy t GDP t Auth t Auth _ GDP t t If the theory regarding the interaction of authoritarianism and per capita GDP is correct, the coefficient of the interaction variable, Auth_GdpPC, should positive: Auth _ GDP > 0. (If you are not certain why, it should become clear shortly.) The null and alternative hypotheses are: H 0 : Auth _ GDP = 0 H 1 : _ > 0 Auth GDP

28 28 Step 1: Collect data, run the regression, and interpret the estimates. Ordinary Least Squares (OLS) Dependent Variable: LogUsersernet Explanatory Variable(s): Estimate SE t-statistic Prob Year CapitalHuman CapitalPhysical GdpPC Auth Auth_GdpPC Const Number of Observations 566 Estimated Equation: EstLogUsersernet = Year +.022CapitalHuman +.003CapitalPhysical +.033GdpPC.230Auth Auth_GdpPC Table 13.12: ernet Regression Results With eraction Variable Focus attention on the estimated effect of GDP. To do so, consider both the GDP and Auth_GDP terms in the estimated equation:.033gdppc Auth_GdpPC.033GdpPC Auth GdpPC

29 29 We shall now estimate the impact of GDP for several values of the authoritarian index: Authoritarian Index Estimated Effect of Per Capita GDP 0.033GdpPC GdpPC.033GdpPC + 0 =.033GdpPC 2.033GdpPC GdpPC.033GdpPC +.035GdpPC =.068GdpPC 4.033GdpPC GdpPC.033GdpPC +.070GdpPC =.103GdpPC 6.033GdpPC GdpPC.033GdpPC +.105GdpPC =.138GdpPC 8.033GdpPC GdpPC.033GdpPC +.140GdpPC =.173GdpPC Table 13.13: eraction Variable Estimate Calculations Recall that as the authoritarian index increases, the level of authoritarianism rises. Therefore, the estimates suggest that as a nation becomes more authoritarian, a $1,000 increase in per capita GDP increases ernet use by larger amounts. This supports the position of those who believe that citizens of all nations seek out uncensored information. In authoritarian nations, citizens have few sources of uncensored information; therefore, as per capita GDP rises, they embrace the uncensored information the ernet provides more enthusiastically than do citizens of non-authoritarian nation in which other sources of uncensored information are available. 1 As a consequence of privacy concerns, these data were artificially generated. 2 Note that even though the tails probability is reported as.0000, the probability can never precisely equal 0. It will always exceed 0. Consequently, instead of writing.0000, we write <.0001 to emphasize the fact that the probability can never equal precisely 0.

Chapter 15: Other Regression Statistics and Pitfalls

Chapter 15: Other Regression Statistics and Pitfalls Chapter 15: Other Regression Statistics and Pitfalls Chapter 15 Outline Two-Tailed Confidence Intervals o Confidence Interval Approach: Which Theories Are Consistent with the Data? o A Confidence Interval

More information

Chapter 5: Ordinary Least Squares Estimation Procedure The Mechanics Chapter 5 Outline Best Fitting Line Clint s Assignment Simple Regression Model o

Chapter 5: Ordinary Least Squares Estimation Procedure The Mechanics Chapter 5 Outline Best Fitting Line Clint s Assignment Simple Regression Model o Chapter 5: Ordinary Least Squares Estimation Procedure The Mechanics Chapter 5 Outline Best Fitting Line Clint s Assignment Simple Regression Model o Parameters of the Model o Error Term and Random Influences

More information

Chapter 12: Model Specification and Development

Chapter 12: Model Specification and Development Chapter 12: Model Specification and Development Chapter 12 Outline Model Specification: Ramsey REgression Specification Error Test (RESET) o RESET Logic o Linear Demand Model o Constant Elasticity Demand

More information

Solutions: Monday, October 15

Solutions: Monday, October 15 Amherst College Department of Economics Economics 360 Fall 2012 1. Consider Nebraska petroleum consumption. Solutions: Monday, October 15 Petroleum Consumption Data for Nebraska: Annual time series data

More information

Chapter 11 Handout: Hypothesis Testing and the Wald Test

Chapter 11 Handout: Hypothesis Testing and the Wald Test Chapter 11 Handout: Hypothesis Testing and the Wald Test Preview No Money Illusion Theory: Calculating True] o Clever Algebraic Manipulation o Wald Test Restricted Regression Reflects Unrestricted Regression

More information

Chapter 8 Handout: Interval Estimates and Hypothesis Testing

Chapter 8 Handout: Interval Estimates and Hypothesis Testing Chapter 8 Handout: Interval Estimates and Hypothesis esting Preview Clint s Assignment: aking Stock General Properties of the Ordinary Least Squares (OLS) Estimation Procedure Estimate Reliability: Interval

More information

Wednesday, October 17 Handout: Hypothesis Testing and the Wald Test

Wednesday, October 17 Handout: Hypothesis Testing and the Wald Test Amherst College Department of Economics Economics 360 Fall 2012 Wednesday, October 17 Handout: Hypothesis Testing and the Wald Test Preview No Money Illusion Theory: Calculating True] o Clever Algebraic

More information

Chapter 10: Multiple Regression Analysis Introduction

Chapter 10: Multiple Regression Analysis Introduction Chapter 10: Multiple Regression Analysis Introduction Chapter 10 Outline Simple versus Multiple Regression Analysis Goal of Multiple Regression Analysis A One-Tailed Test: Downward Sloping Demand Theory

More information

Wednesday, September 19 Handout: Ordinary Least Squares Estimation Procedure The Mechanics

Wednesday, September 19 Handout: Ordinary Least Squares Estimation Procedure The Mechanics Amherst College Department of Economics Economics Fall 2012 Wednesday, September 19 Handout: Ordinary Least Squares Estimation Procedure he Mechanics Preview Best Fitting Line: Income and Savings Clint

More information

[Mean[e j ] Mean[e i ]]

[Mean[e j ] Mean[e i ]] Amherst College Department of Economics Economics 360 Fall 202 Solutions: Wednesday, September 26. Assume that the standard ordinary least square (OLS) premises are met. Let (x i, y i ) and (, y j ) be

More information

Wednesday, October 10 Handout: One-Tailed Tests, Two-Tailed Tests, and Logarithms

Wednesday, October 10 Handout: One-Tailed Tests, Two-Tailed Tests, and Logarithms Amherst College Department of Economics Economics 360 Fall 2012 Wednesday, October 10 Handout: One-Tailed Tests, Two-Tailed Tests, and Logarithms Preview A One-Tailed Hypothesis Test: The Downward Sloping

More information

Hint: The following equation converts Celsius to Fahrenheit: F = C where C = degrees Celsius F = degrees Fahrenheit

Hint: The following equation converts Celsius to Fahrenheit: F = C where C = degrees Celsius F = degrees Fahrenheit Amherst College Department of Economics Economics 360 Fall 2014 Exam 1: Solutions 1. (10 points) The following table in reports the summary statistics for high and low temperatures in Key West, FL from

More information

Monday, October 15 Handout: Multiple Regression Analysis Introduction

Monday, October 15 Handout: Multiple Regression Analysis Introduction Amherst College Department of Economics Economics 360 Fall 2012 Monday, October 15 Handout: Multiple Regression Analysis Introduction Review Simple and Multiple Regression Analysis o Distinction between

More information

where Female = 0 for males, = 1 for females Age is measured in years (22, 23, ) GPA is measured in units on a four-point scale (0, 1.22, 3.45, etc.

where Female = 0 for males, = 1 for females Age is measured in years (22, 23, ) GPA is measured in units on a four-point scale (0, 1.22, 3.45, etc. Notes on regression analysis 1. Basics in regression analysis key concepts (actual implementation is more complicated) A. Collect data B. Plot data on graph, draw a line through the middle of the scatter

More information

In order to carry out a study on employees wages, a company collects information from its 500 employees 1 as follows:

In order to carry out a study on employees wages, a company collects information from its 500 employees 1 as follows: INTRODUCTORY ECONOMETRICS Dpt of Econometrics & Statistics (EA3) University of the Basque Country UPV/EHU OCW Self Evaluation answers Time: 21/2 hours SURNAME: NAME: ID#: Specific competences to be evaluated

More information

Marketing Research Session 10 Hypothesis Testing with Simple Random samples (Chapter 12)

Marketing Research Session 10 Hypothesis Testing with Simple Random samples (Chapter 12) Marketing Research Session 10 Hypothesis Testing with Simple Random samples (Chapter 12) Remember: Z.05 = 1.645, Z.01 = 2.33 We will only cover one-sided hypothesis testing (cases 12.3, 12.4.2, 12.5.2,

More information

Hypothesis testing. Data to decisions

Hypothesis testing. Data to decisions Hypothesis testing Data to decisions The idea Null hypothesis: H 0 : the DGP/population has property P Under the null, a sample statistic has a known distribution If, under that that distribution, the

More information

CHAPTER 7. + ˆ δ. (1 nopc) + ˆ β1. =.157, so the new intercept is = The coefficient on nopc is.157.

CHAPTER 7. + ˆ δ. (1 nopc) + ˆ β1. =.157, so the new intercept is = The coefficient on nopc is.157. CHAPTER 7 SOLUTIONS TO PROBLEMS 7. (i) The coefficient on male is 87.75, so a man is estimated to sleep almost one and one-half hours more per week than a comparable woman. Further, t male = 87.75/34.33

More information

Midterm 2 - Solutions

Midterm 2 - Solutions Ecn 102 - Analysis of Economic Data University of California - Davis February 23, 2010 Instructor: John Parman Midterm 2 - Solutions You have until 10:20am to complete this exam. Please remember to put

More information

Solutions to Exercises in Chapter 9

Solutions to Exercises in Chapter 9 in 9. (a) When a GPA is increased by one unit, and other variables are held constant, average starting salary will increase by the amount $643. Students who take econometrics will have a starting salary

More information

4. Nonlinear regression functions

4. Nonlinear regression functions 4. Nonlinear regression functions Up to now: Population regression function was assumed to be linear The slope(s) of the population regression function is (are) constant The effect on Y of a unit-change

More information

Chapter 9: The Regression Model with Qualitative Information: Binary Variables (Dummies)

Chapter 9: The Regression Model with Qualitative Information: Binary Variables (Dummies) Chapter 9: The Regression Model with Qualitative Information: Binary Variables (Dummies) Statistics and Introduction to Econometrics M. Angeles Carnero Departamento de Fundamentos del Análisis Económico

More information

CHAPTER 4. > 0, where β

CHAPTER 4. > 0, where β CHAPTER 4 SOLUTIONS TO PROBLEMS 4. (i) and (iii) generally cause the t statistics not to have a t distribution under H. Homoskedasticity is one of the CLM assumptions. An important omitted variable violates

More information

Ch 7: Dummy (binary, indicator) variables

Ch 7: Dummy (binary, indicator) variables Ch 7: Dummy (binary, indicator) variables :Examples Dummy variable are used to indicate the presence or absence of a characteristic. For example, define female i 1 if obs i is female 0 otherwise or male

More information

Making sense of Econometrics: Basics

Making sense of Econometrics: Basics Making sense of Econometrics: Basics Lecture 4: Qualitative influences and Heteroskedasticity Egypt Scholars Economic Society November 1, 2014 Assignment & feedback enter classroom at http://b.socrative.com/login/student/

More information

Statistics and Quantitative Analysis U4320

Statistics and Quantitative Analysis U4320 Statistics and Quantitative Analysis U3 Lecture 13: Explaining Variation Prof. Sharyn O Halloran Explaining Variation: Adjusted R (cont) Definition of Adjusted R So we'd like a measure like R, but one

More information

CHAPTER 5 FUNCTIONAL FORMS OF REGRESSION MODELS

CHAPTER 5 FUNCTIONAL FORMS OF REGRESSION MODELS CHAPTER 5 FUNCTIONAL FORMS OF REGRESSION MODELS QUESTIONS 5.1. (a) In a log-log model the dependent and all explanatory variables are in the logarithmic form. (b) In the log-lin model the dependent variable

More information

ECON 482 / WH Hong Binary or Dummy Variables 1. Qualitative Information

ECON 482 / WH Hong Binary or Dummy Variables 1. Qualitative Information 1. Qualitative Information Qualitative Information Up to now, we assume that all the variables has quantitative meaning. But often in empirical work, we must incorporate qualitative factor into regression

More information

An Introduction to Econometrics. A Self-contained Approach. Frank Westhoff. The MIT Press Cambridge, Massachusetts London, England

An Introduction to Econometrics. A Self-contained Approach. Frank Westhoff. The MIT Press Cambridge, Massachusetts London, England An Introduction to Econometrics A Self-contained Approach Frank Westhoff The MIT Press Cambridge, Massachusetts London, England How to Use This Book xvii 1 Descriptive Statistics 1 Chapter 1 Prep Questions

More information

9. Linear Regression and Correlation

9. Linear Regression and Correlation 9. Linear Regression and Correlation Data: y a quantitative response variable x a quantitative explanatory variable (Chap. 8: Recall that both variables were categorical) For example, y = annual income,

More information

2) For a normal distribution, the skewness and kurtosis measures are as follows: A) 1.96 and 4 B) 1 and 2 C) 0 and 3 D) 0 and 0

2) For a normal distribution, the skewness and kurtosis measures are as follows: A) 1.96 and 4 B) 1 and 2 C) 0 and 3 D) 0 and 0 Introduction to Econometrics Midterm April 26, 2011 Name Student ID MULTIPLE CHOICE. Choose the one alternative that best completes the statement or answers the question. (5,000 credit for each correct

More information

One-Way ANOVA. Some examples of when ANOVA would be appropriate include:

One-Way ANOVA. Some examples of when ANOVA would be appropriate include: One-Way ANOVA 1. Purpose Analysis of variance (ANOVA) is used when one wishes to determine whether two or more groups (e.g., classes A, B, and C) differ on some outcome of interest (e.g., an achievement

More information

Binary Logistic Regression

Binary Logistic Regression The coefficients of the multiple regression model are estimated using sample data with k independent variables Estimated (or predicted) value of Y Estimated intercept Estimated slope coefficients Ŷ = b

More information

Ecn Analysis of Economic Data University of California - Davis February 23, 2010 Instructor: John Parman. Midterm 2. Name: ID Number: Section:

Ecn Analysis of Economic Data University of California - Davis February 23, 2010 Instructor: John Parman. Midterm 2. Name: ID Number: Section: Ecn 102 - Analysis of Economic Data University of California - Davis February 23, 2010 Instructor: John Parman Midterm 2 You have until 10:20am to complete this exam. Please remember to put your name,

More information

PBAF 528 Week 8. B. Regression Residuals These properties have implications for the residuals of the regression.

PBAF 528 Week 8. B. Regression Residuals These properties have implications for the residuals of the regression. PBAF 528 Week 8 What are some problems with our model? Regression models are used to represent relationships between a dependent variable and one or more predictors. In order to make inference from the

More information

Regression Models REVISED TEACHING SUGGESTIONS ALTERNATIVE EXAMPLES

Regression Models REVISED TEACHING SUGGESTIONS ALTERNATIVE EXAMPLES M04_REND6289_10_IM_C04.QXD 5/7/08 2:49 PM Page 46 4 C H A P T E R Regression Models TEACHING SUGGESTIONS Teaching Suggestion 4.1: Which Is the Independent Variable? We find that students are often confused

More information

Sociology Research Statistics I Final Exam Answer Key December 15, 1993

Sociology Research Statistics I Final Exam Answer Key December 15, 1993 Sociology 592 - Research Statistics I Final Exam Answer Key December 15, 1993 Where appropriate, show your work - partial credit may be given. (On the other hand, don't waste a lot of time on excess verbiage.)

More information

Solutions to Problem Set 5 (Due November 22) Maximum number of points for Problem set 5 is: 220. Problem 7.3

Solutions to Problem Set 5 (Due November 22) Maximum number of points for Problem set 5 is: 220. Problem 7.3 Solutions to Problem Set 5 (Due November 22) EC 228 02, Fall 2010 Prof. Baum, Ms Hristakeva Maximum number of points for Problem set 5 is: 220 Problem 7.3 (i) (5 points) The t statistic on hsize 2 is over

More information

Simple Linear Regression: One Qualitative IV

Simple Linear Regression: One Qualitative IV Simple Linear Regression: One Qualitative IV 1. Purpose As noted before regression is used both to explain and predict variation in DVs, and adding to the equation categorical variables extends regression

More information

Is economic freedom related to economic growth?

Is economic freedom related to economic growth? Is economic freedom related to economic growth? It is an article of faith among supporters of capitalism: economic freedom leads to economic growth. The publication Economic Freedom of the World: 2003

More information

Economics Introduction to Econometrics - Fall 2007 Final Exam - Answers

Economics Introduction to Econometrics - Fall 2007 Final Exam - Answers Student Name: Economics 4818 - Introduction to Econometrics - Fall 2007 Final Exam - Answers SHOW ALL WORK! Evaluation: Problems: 3, 4C, 5C and 5F are worth 4 points. All other questions are worth 3 points.

More information

ECON 450 Development Economics

ECON 450 Development Economics ECON 450 Development Economics Statistics Background University of Illinois at Urbana-Champaign Summer 2017 Outline 1 Introduction 2 3 4 5 Introduction Regression analysis is one of the most important

More information

B. Weaver (24-Mar-2005) Multiple Regression Chapter 5: Multiple Regression Y ) (5.1) Deviation score = (Y i

B. Weaver (24-Mar-2005) Multiple Regression Chapter 5: Multiple Regression Y ) (5.1) Deviation score = (Y i B. Weaver (24-Mar-2005) Multiple Regression... 1 Chapter 5: Multiple Regression 5.1 Partial and semi-partial correlation Before starting on multiple regression per se, we need to consider the concepts

More information

MGEC11H3Y L01 Introduction to Regression Analysis Term Test Friday July 5, PM Instructor: Victor Yu

MGEC11H3Y L01 Introduction to Regression Analysis Term Test Friday July 5, PM Instructor: Victor Yu Last Name (Print): Solution First Name (Print): Student Number: MGECHY L Introduction to Regression Analysis Term Test Friday July, PM Instructor: Victor Yu Aids allowed: Time allowed: Calculator and one

More information

Sociology 593 Exam 2 Answer Key March 28, 2002

Sociology 593 Exam 2 Answer Key March 28, 2002 Sociology 59 Exam Answer Key March 8, 00 I. True-False. (0 points) Indicate whether the following statements are true or false. If false, briefly explain why.. A variable is called CATHOLIC. This probably

More information

ECO220Y Simple Regression: Testing the Slope

ECO220Y Simple Regression: Testing the Slope ECO220Y Simple Regression: Testing the Slope Readings: Chapter 18 (Sections 18.3-18.5) Winter 2012 Lecture 19 (Winter 2012) Simple Regression Lecture 19 1 / 32 Simple Regression Model y i = β 0 + β 1 x

More information

(quantitative or categorical variables) Numerical descriptions of center, variability, position (quantitative variables)

(quantitative or categorical variables) Numerical descriptions of center, variability, position (quantitative variables) 3. Descriptive Statistics Describing data with tables and graphs (quantitative or categorical variables) Numerical descriptions of center, variability, position (quantitative variables) Bivariate descriptions

More information

The Empirical Rule, z-scores, and the Rare Event Approach

The Empirical Rule, z-scores, and the Rare Event Approach Overview The Empirical Rule, z-scores, and the Rare Event Approach Look at Chebyshev s Rule and the Empirical Rule Explore some applications of the Empirical Rule How to calculate and use z-scores Introducing

More information

5. Let W follow a normal distribution with mean of μ and the variance of 1. Then, the pdf of W is

5. Let W follow a normal distribution with mean of μ and the variance of 1. Then, the pdf of W is Practice Final Exam Last Name:, First Name:. Please write LEGIBLY. Answer all questions on this exam in the space provided (you may use the back of any page if you need more space). Show all work but do

More information

Wooldridge, Introductory Econometrics, 4th ed. Chapter 6: Multiple regression analysis: Further issues

Wooldridge, Introductory Econometrics, 4th ed. Chapter 6: Multiple regression analysis: Further issues Wooldridge, Introductory Econometrics, 4th ed. Chapter 6: Multiple regression analysis: Further issues What effects will the scale of the X and y variables have upon multiple regression? The coefficients

More information

ECONOMETRIC MODEL WITH QUALITATIVE VARIABLES

ECONOMETRIC MODEL WITH QUALITATIVE VARIABLES ECONOMETRIC MODEL WITH QUALITATIVE VARIABLES How to quantify qualitative variables to quantitative variables? Why do we need to do this? Econometric model needs quantitative variables to estimate its parameters

More information

Final Exam - Solutions

Final Exam - Solutions Ecn 102 - Analysis of Economic Data University of California - Davis March 19, 2010 Instructor: John Parman Final Exam - Solutions You have until 5:30pm to complete this exam. Please remember to put your

More information

Intermediate Econometrics

Intermediate Econometrics Intermediate Econometrics Markus Haas LMU München Summer term 2011 15. Mai 2011 The Simple Linear Regression Model Considering variables x and y in a specific population (e.g., years of education and wage

More information

Ordinary Least Squares Regression Explained: Vartanian

Ordinary Least Squares Regression Explained: Vartanian Ordinary Least Squares Regression Explained: Vartanian When to Use Ordinary Least Squares Regression Analysis A. Variable types. When you have an interval/ratio scale dependent variable.. When your independent

More information

Chapter 14: Omitted Explanatory Variables, Multicollinearity, and Irrelevant Explanatory Variables

Chapter 14: Omitted Explanatory Variables, Multicollinearity, and Irrelevant Explanatory Variables Chapter 14: Omitted Explanatory Variables, Multicollinearity, and Irrelevant Explanatory Variables Chapter 14 Outline Review o Unbiased Estimation Procedures Estimates and Random Variables Mean of the

More information

Correlation & Simple Regression

Correlation & Simple Regression Chapter 11 Correlation & Simple Regression The previous chapter dealt with inference for two categorical variables. In this chapter, we would like to examine the relationship between two quantitative variables.

More information

Contest Quiz 3. Question Sheet. In this quiz we will review concepts of linear regression covered in lecture 2.

Contest Quiz 3. Question Sheet. In this quiz we will review concepts of linear regression covered in lecture 2. Updated: November 17, 2011 Lecturer: Thilo Klein Contact: tk375@cam.ac.uk Contest Quiz 3 Question Sheet In this quiz we will review concepts of linear regression covered in lecture 2. NOTE: Please round

More information

11.5 Regression Linear Relationships

11.5 Regression Linear Relationships Contents 11.5 Regression............................. 835 11.5.1 Linear Relationships................... 835 11.5.2 The Least Squares Regression Line........... 837 11.5.3 Using the Regression Line................

More information

CHAPTER 4 & 5 Linear Regression with One Regressor. Kazu Matsuda IBEC PHBU 430 Econometrics

CHAPTER 4 & 5 Linear Regression with One Regressor. Kazu Matsuda IBEC PHBU 430 Econometrics CHAPTER 4 & 5 Linear Regression with One Regressor Kazu Matsuda IBEC PHBU 430 Econometrics Introduction Simple linear regression model = Linear model with one independent variable. y = dependent variable

More information

Answer Key. 9.1 Scatter Plots and Linear Correlation. Chapter 9 Regression and Correlation. CK-12 Advanced Probability and Statistics Concepts 1

Answer Key. 9.1 Scatter Plots and Linear Correlation. Chapter 9 Regression and Correlation. CK-12 Advanced Probability and Statistics Concepts 1 9.1 Scatter Plots and Linear Correlation Answers 1. A high school psychologist wants to conduct a survey to answer the question: Is there a relationship between a student s athletic ability and his/her

More information

Regression with Qualitative Information. Part VI. Regression with Qualitative Information

Regression with Qualitative Information. Part VI. Regression with Qualitative Information Part VI Regression with Qualitative Information As of Oct 17, 2017 1 Regression with Qualitative Information Single Dummy Independent Variable Multiple Categories Ordinal Information Interaction Involving

More information

Intermediate Econometrics

Intermediate Econometrics Intermediate Econometrics Heteroskedasticity Text: Wooldridge, 8 July 17, 2011 Heteroskedasticity Assumption of homoskedasticity, Var(u i x i1,..., x ik ) = E(u 2 i x i1,..., x ik ) = σ 2. That is, the

More information

Chapter 9. Dummy (Binary) Variables. 9.1 Introduction The multiple regression model (9.1.1) Assumption MR1 is

Chapter 9. Dummy (Binary) Variables. 9.1 Introduction The multiple regression model (9.1.1) Assumption MR1 is Chapter 9 Dummy (Binary) Variables 9.1 Introduction The multiple regression model y = β+β x +β x + +β x + e (9.1.1) t 1 2 t2 3 t3 K tk t Assumption MR1 is 1. yt =β 1+β 2xt2 + L+β KxtK + et, t = 1, K, T

More information

ECON 5350 Class Notes Functional Form and Structural Change

ECON 5350 Class Notes Functional Form and Structural Change ECON 5350 Class Notes Functional Form and Structural Change 1 Introduction Although OLS is considered a linear estimator, it does not mean that the relationship between Y and X needs to be linear. In this

More information

Regression #8: Loose Ends

Regression #8: Loose Ends Regression #8: Loose Ends Econ 671 Purdue University Justin L. Tobias (Purdue) Regression #8 1 / 30 In this lecture we investigate a variety of topics that you are probably familiar with, but need to touch

More information

Chapter 10 Nonlinear Models

Chapter 10 Nonlinear Models Chapter 10 Nonlinear Models Nonlinear models can be classified into two categories. In the first category are models that are nonlinear in the variables, but still linear in terms of the unknown parameters.

More information

Lecture 3: Multiple Regression. Prof. Sharyn O Halloran Sustainable Development U9611 Econometrics II

Lecture 3: Multiple Regression. Prof. Sharyn O Halloran Sustainable Development U9611 Econometrics II Lecture 3: Multiple Regression Prof. Sharyn O Halloran Sustainable Development Econometrics II Outline Basics of Multiple Regression Dummy Variables Interactive terms Curvilinear models Review Strategies

More information

Answer Key: Problem Set 5

Answer Key: Problem Set 5 : Problem Set 5. Let nopc be a dummy variable equal to one if the student does not own a PC, and zero otherwise. i. If nopc is used instead of PC in the model of: colgpa = β + δ PC + β hsgpa + β ACT +

More information

Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares

Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares Many economic models involve endogeneity: that is, a theoretical relationship does not fit

More information

[ ESS ESS ] / 2 [ ] / ,019.6 / Lab 10 Key. Regression Analysis: wage versus yrsed, ex

[ ESS ESS ] / 2 [ ] / ,019.6 / Lab 10 Key. Regression Analysis: wage versus yrsed, ex Lab 1 Key Regression Analysis: wage versus yrsed, ex wage = - 4.78 + 1.46 yrsed +.126 ex Constant -4.78 2.146-2.23.26 yrsed 1.4623.153 9.73. ex.12635.2739 4.61. S = 8.9851 R-Sq = 11.9% R-Sq(adj) = 11.7%

More information

Review of Multiple Regression

Review of Multiple Regression Ronald H. Heck 1 Let s begin with a little review of multiple regression this week. Linear models [e.g., correlation, t-tests, analysis of variance (ANOVA), multiple regression, path analysis, multivariate

More information

Chapte The McGraw-Hill Companies, Inc. All rights reserved.

Chapte The McGraw-Hill Companies, Inc. All rights reserved. 12er12 Chapte Bivariate i Regression (Part 1) Bivariate Regression Visual Displays Begin the analysis of bivariate data (i.e., two variables) with a scatter plot. A scatter plot - displays each observed

More information

Sociology Exam 2 Answer Key March 30, 2012

Sociology Exam 2 Answer Key March 30, 2012 Sociology 63993 Exam 2 Answer Key March 30, 2012 I. True-False. (20 points) Indicate whether the following statements are true or false. If false, briefly explain why. 1. A researcher has constructed scales

More information

Practice exam questions

Practice exam questions Practice exam questions Nathaniel Higgins nhiggins@jhu.edu, nhiggins@ers.usda.gov 1. The following question is based on the model y = β 0 + β 1 x 1 + β 2 x 2 + β 3 x 3 + u. Discuss the following two hypotheses.

More information

Note on Bivariate Regression: Connecting Practice and Theory. Konstantin Kashin

Note on Bivariate Regression: Connecting Practice and Theory. Konstantin Kashin Note on Bivariate Regression: Connecting Practice and Theory Konstantin Kashin Fall 2012 1 This note will explain - in less theoretical terms - the basics of a bivariate linear regression, including testing

More information

Do not copy, post, or distribute

Do not copy, post, or distribute 14 CORRELATION ANALYSIS AND LINEAR REGRESSION Assessing the Covariability of Two Quantitative Properties 14.0 LEARNING OBJECTIVES In this chapter, we discuss two related techniques for assessing a possible

More information

ECON 497 Midterm Spring

ECON 497 Midterm Spring ECON 497 Midterm Spring 2009 1 ECON 497: Economic Research and Forecasting Name: Spring 2009 Bellas Midterm You have three hours and twenty minutes to complete this exam. Answer all questions and explain

More information

Econ 444, class 11. Robert de Jong 1. Monday November 6. Ohio State University. Econ 444, Wednesday November 1, class Department of Economics

Econ 444, class 11. Robert de Jong 1. Monday November 6. Ohio State University. Econ 444, Wednesday November 1, class Department of Economics Econ 444, class 11 Robert de Jong 1 1 Department of Economics Ohio State University Monday November 6 Monday November 6 1 Exercise for today 2 New material: 1 dummy variables 2 multicollinearity Exercise

More information

y response variable x 1, x 2,, x k -- a set of explanatory variables

y response variable x 1, x 2,, x k -- a set of explanatory variables 11. Multiple Regression and Correlation y response variable x 1, x 2,, x k -- a set of explanatory variables In this chapter, all variables are assumed to be quantitative. Chapters 12-14 show how to incorporate

More information

Technical Appendix C: Methods

Technical Appendix C: Methods Technical Appendix C: Methods As not all readers may be familiar with the multilevel analytical methods used in this study, a brief note helps to clarify the techniques. The general theory developed in

More information

LI EAR REGRESSIO A D CORRELATIO

LI EAR REGRESSIO A D CORRELATIO CHAPTER 6 LI EAR REGRESSIO A D CORRELATIO Page Contents 6.1 Introduction 10 6. Curve Fitting 10 6.3 Fitting a Simple Linear Regression Line 103 6.4 Linear Correlation Analysis 107 6.5 Spearman s Rank Correlation

More information

Answer all questions from part I. Answer two question from part II.a, and one question from part II.b.

Answer all questions from part I. Answer two question from part II.a, and one question from part II.b. B203: Quantitative Methods Answer all questions from part I. Answer two question from part II.a, and one question from part II.b. Part I: Compulsory Questions. Answer all questions. Each question carries

More information

Properties of estimator Functional Form. Econometrics. Lecture 8. Nathaniel Higgins JHU. Nathaniel Higgins Lecture 8

Properties of estimator Functional Form. Econometrics. Lecture 8. Nathaniel Higgins JHU. Nathaniel Higgins Lecture 8 Econometrics Lecture 8 Nathaniel Higgins JHU Homework Next class: GDP, population, temperature, energy, and mortality data together by Nov. 9 (next class) If you have questions / need help, let Rob or

More information

Multiple regression: Model building. Topics. Correlation Matrix. CQMS 202 Business Statistics II Prepared by Moez Hababou

Multiple regression: Model building. Topics. Correlation Matrix. CQMS 202 Business Statistics II Prepared by Moez Hababou Multiple regression: Model building CQMS 202 Business Statistics II Prepared by Moez Hababou Topics Forward versus backward model building approach Using the correlation matrix Testing for multicolinearity

More information

STAT 135 Lab 11 Tests for Categorical Data (Fisher s Exact test, χ 2 tests for Homogeneity and Independence) and Linear Regression

STAT 135 Lab 11 Tests for Categorical Data (Fisher s Exact test, χ 2 tests for Homogeneity and Independence) and Linear Regression STAT 135 Lab 11 Tests for Categorical Data (Fisher s Exact test, χ 2 tests for Homogeneity and Independence) and Linear Regression Rebecca Barter April 20, 2015 Fisher s Exact Test Fisher s Exact Test

More information

Outline for Today. Review of In-class Exercise Bivariate Hypothesis Test 2: Difference of Means Bivariate Hypothesis Testing 3: Correla

Outline for Today. Review of In-class Exercise Bivariate Hypothesis Test 2: Difference of Means Bivariate Hypothesis Testing 3: Correla Outline for Today 1 Review of In-class Exercise 2 Bivariate hypothesis testing 2: difference of means 3 Bivariate hypothesis testing 3: correlation 2 / 51 Task for ext Week Any questions? 3 / 51 In-class

More information

Module 03 Lecture 14 Inferential Statistics ANOVA and TOI

Module 03 Lecture 14 Inferential Statistics ANOVA and TOI Introduction of Data Analytics Prof. Nandan Sudarsanam and Prof. B Ravindran Department of Management Studies and Department of Computer Science and Engineering Indian Institute of Technology, Madras Module

More information

Psych 230. Psychological Measurement and Statistics

Psych 230. Psychological Measurement and Statistics Psych 230 Psychological Measurement and Statistics Pedro Wolf December 9, 2009 This Time. Non-Parametric statistics Chi-Square test One-way Two-way Statistical Testing 1. Decide which test to use 2. State

More information

STAT 350 Final (new Material) Review Problems Key Spring 2016

STAT 350 Final (new Material) Review Problems Key Spring 2016 1. The editor of a statistics textbook would like to plan for the next edition. A key variable is the number of pages that will be in the final version. Text files are prepared by the authors using LaTeX,

More information

18.0 Multiple and Nonlinear Regression

18.0 Multiple and Nonlinear Regression 18.0 Multiple and Nonlinear Regression 1 Answer Questions Multiple Regression Nonlinear Regression 18.1 Multiple Regression Recall the regression assumptions: 1. Each point (X i,y i ) in the scatterplot

More information

Unit 1 Review of BIOSTATS 540 Practice Problems SOLUTIONS - Stata Users

Unit 1 Review of BIOSTATS 540 Practice Problems SOLUTIONS - Stata Users BIOSTATS 640 Spring 2017 Review of Introductory Biostatistics STATA solutions Page 1 of 16 Unit 1 Review of BIOSTATS 540 Practice Problems SOLUTIONS - Stata Users #1. The following table lists length of

More information

Inference with Simple Regression

Inference with Simple Regression 1 Introduction Inference with Simple Regression Alan B. Gelder 06E:071, The University of Iowa 1 Moving to infinite means: In this course we have seen one-mean problems, twomean problems, and problems

More information

Mathematics for Economics MA course

Mathematics for Economics MA course Mathematics for Economics MA course Simple Linear Regression Dr. Seetha Bandara Simple Regression Simple linear regression is a statistical method that allows us to summarize and study relationships between

More information

Example. χ 2 = Continued on the next page. All cells

Example. χ 2 = Continued on the next page. All cells Section 11.1 Chi Square Statistic k Categories 1 st 2 nd 3 rd k th Total Observed Frequencies O 1 O 2 O 3 O k n Expected Frequencies E 1 E 2 E 3 E k n O 1 + O 2 + O 3 + + O k = n E 1 + E 2 + E 3 + + E

More information

Recitation 1: Regression Review. Christina Patterson

Recitation 1: Regression Review. Christina Patterson Recitation 1: Regression Review Christina Patterson Outline For Recitation 1. Statistics. Bias, sampling variance and hypothesis testing.. Two important statistical theorems: Law of large numbers (LLN)

More information

download instant at

download instant at Answers to Odd-Numbered Exercises Chapter One: An Overview of Regression Analysis 1-3. (a) Positive, (b) negative, (c) positive, (d) negative, (e) ambiguous, (f) negative. 1-5. (a) The coefficients in

More information

Linear Regression With Special Variables

Linear Regression With Special Variables Linear Regression With Special Variables Junhui Qian December 21, 2014 Outline Standardized Scores Quadratic Terms Interaction Terms Binary Explanatory Variables Binary Choice Models Standardized Scores:

More information

Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page!

Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page! Econometrics - Exam May 11, 2011 1 Exam Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page! Problem 1: (15 points) A researcher has data for the year 2000 from

More information

An Introduction to Mplus and Path Analysis

An Introduction to Mplus and Path Analysis An Introduction to Mplus and Path Analysis PSYC 943: Fundamentals of Multivariate Modeling Lecture 10: October 30, 2013 PSYC 943: Lecture 10 Today s Lecture Path analysis starting with multivariate regression

More information

Econometrics I Lecture 7: Dummy Variables

Econometrics I Lecture 7: Dummy Variables Econometrics I Lecture 7: Dummy Variables Mohammad Vesal Graduate School of Management and Economics Sharif University of Technology 44716 Fall 1397 1 / 27 Introduction Dummy variable: d i is a dummy variable

More information