ISUP - Undergrad econometrics prep.
Terms : Hide Images
| 1576675683 | What is the estimated slope of the OLS? |  | 0 | |
| 1576675684 | What does the ^ in equations mean? | The Hat means that this is the PREDICTED values. Hence, not the actual values for the population | 1 | |
| 1576675685 | What is this equations? ûi = Yi-Yhat.i | The error term - Being the deviation from the regression line. | 2 | |
| 1576675686 | What are the 5 Simple linear regression assumption? | ••Assumption SLR.1 (Linear in parameters) ••Assumption SLR.2 (Random sampling) ••Assumption SLR.3 (Sample variation in explanatory variable) ••Assumption SLR.4 (Zero conditional mean) ••Assumption SLR.5 (Homoskedasticity) | 3 | |
| 1576675687 | What is the total variation and how is it meassured? |  | 4 | |
| 1576675688 | How do you meassure goodness of fit, and what is the formula? |  | 5 | |
| 1576675689 | What is the definition and formula for SST? |  | 6 | |
| 1576675690 | What is the definition and formula for SSE? |  | 7 | |
| 1576675691 | What is the definition and formula for SSR? |  | 8 | |
| 1576677294 | If H0: β1 = 0 and β2 = 0, how many restrictions are there? | Two | 9 | |
| 1576677295 | Homoskedasticity defn. | Variance of error term is constant, Var(Yi I Xi') = Var(Yi I Xi'') | 10 | |
| 1576677296 | Heteroskedasticity defn. | Variance of error term is not constant, Var(Yi I Xi') < Var(Yi I Xi'') | 11 | |
| 1576677297 | How does adding additional regressors affect R2? | It inflates it, (R2 = 1 - SSR/TSS)...SSR increases. We need the adj. R2 | 12 | |
| 1576677298 | Pitfalls of R2 and adj. R2 with regards to significance, correlation and causality | 1. An increase in R2 or adj. R2 does not mean the added variable is statistically significant 2. Correlation does not mean causality 3. A high R2 or adj. R2 does not mean there is not OVB (the reverse is also true) 4. A high R2 or adj. R2 does not mean I | 13 | |
| 1576677299 | How do we know if a regressor is statistically significant? | Perform a t-test ( or look at) | 14 | |
| 1576677300 | If R2 is low then we know what about the goodness of fit? | Goodness of fit = (SSE/SST); if it is low then there are a lot of extraneous factors effecting Y other than X. | 15 | |
| 1576677301 | What is the regressor ? | The X variable | 16 | |
| 1576677302 | What relation do the fitted residuals and the regressor have to satisfy? | The values of X must not contain any information about the mean of the residuals: Σ ÛiXi = 0 | 17 | |
| 1576677303 | What is the equation for the "true" regression model corresponding with the linear empirical model of Y and X? | Yi = α + β(Xi) + εi where Xi and Yi need to be specifically defined (say where Xi = age and Yi = earnings) | 18 | |
| 1576677304 | R^2 formula | 1- SSR/TSS | 19 | |
| 1576677305 | Adj. R^2 formula | 1 - [(n -1)/(n - k -1)] - SSR/TSS | 20 | |
| 1576701555 | Please list 3 synonyms for the Y-variable | The dependant variable, the explained variable, the response variable, regressant | 21 | |
| 1576701556 | Please list 3 synonyms for the X-variable | Independant variable, explanatory variable, regressor | 22 | |
| 1576701557 | What is the difference between a parameter and a variable? | The parameters are the Beta's and the variables are the X's in connection with each beta. Each parameter is the estimated relationship between the independant and dependant variable for the population. | 23 | |
| 1576701558 | Which values can R^2 be? | 0| 24 | | |
| 1576727579 | Explain "Exogenous variable" and what it means for MLR 4 | X is uncorrelated with Ui, meaning that no variation in the variable can be explained by the error terms. MLR.4 holds if all explanatory variables (x's) are exogenous. Exogeneity is the key assumption for a causal interpretation of the regression, and for unbiasedness of the OLS estimators | 25 | |
| 1576727580 | Explain "Endogenous variable" and what it means for MLR 4 | X is correlated with Ui, meaning that variation in the variable can be explained by the error terms. (bad!). ; endogeneity is a violation of assumption MLR.4 | 26 | |
| 1576727581 | How can you Fix multicollinarity in a regression? | Drop one of the multicollinear variables or a constant and run the regression again. | 27 | |
| 1576727582 | What is the formula for degrees of freedom? | (N-k-1) | 28 | |
| 1576727583 | What does zero conditional mean, mean? | The value of the explanatory variables must contain no information about the mean of the unobserved factors. | 29 | |
| 1576728035 | Explain the "Unbiasedness of OLS" theorem in words: | On average the coefficients will be the same if the drawing is repeated multiple times. | 30 | |
| 1576749781 | What are the standard assumptions for the multiple regression model? | Assumption MLR.1 (Linear in parameters) Assumption MLR.2 (Random sampling) Assumption MLR.3 (No perfect collinearity) Assumption MLR.4 (Zero conditional mean) Assumption MLR.5 (Homoscedasticity) Assumption MLR.6 (Normality in error terms) | 31 | |
| 1576751967 | Which equation describes Assumption MLR.1 (Linear in parameters)? | Y= B0 + B2X2 +...+ BkXk + u | 32 | |
| 1576751968 | Which equation describes Assumption MLR.4 (Zero conditional mean)? | E(Ui I Xi) = 0 | 33 | |
| 1576749782 | Is Assumption MLR.3 (No perfect collinearity) hard to meet? | No. It only rules out PERFECT collinearity. Hence, close to perfect is ok. Constants are ruled out (collinear with intercept) | 34 | |
| 1576751969 | Is Assumption MLR.4 (Zero conditional mean) E(Ui I Xi) = 0 hard to meet? Comparing simple to multiple regression | No. In a multiple regression model, the zero conditional mean assumption is much more likely to hold because fewer things end up in the error | 35 | |
| 1576755753 | Which equation describes Assumption MLR.5 (Homoscedasticity)? | Var( Ui I Xi1, Xi2,..., Xik) = s^2 | 36 | |
| 1576755754 | When is OLS the best estimator? | OLS is only the best estimator (best linear unbiased estimator BLUE) if MLR.1 - MLR.5 hold; if there is heteroscedasticity for example, there are better estimators. | 37 | |
| 1581273609 | What is an interaction term and what is the interpretation? | A combination term, allowing the change in Y to be a function of a combination of sex, race etc. | 38 | |
| 1581275287 | What is the equation used to find the T-value? t(calc) |  | 39 | |
| 1581277450 | With n>120, what is the critical value for the 5% confidence level? | 1,96 | 40 | |
| 1581277451 | With n>120, what is the critical value for the 10% confidence level? | 1,65 | 41 | |
| 1581277452 | With n>120, what is the critical value for the 1% confidence level? | 2,58 | 42 | |
| 1581278882 | According to Carol, should you drop parameters that are not statistically significant? | No. While it may not be significant in the population, it will still belong to the model formulated and should be left in. Exception is "^2"-terms. You include both the normal and the quadratic term. If the quadratic term is not significant, you drop it, but keep the normal | 43 | |
| 1581281179 | What is the formula for the T-statistic for other values than 0? |  | 44 | |
| 1584654780 | What is the equation for normality in error terms? |  | 45 | |
| 1584663655 | What does the central limit theorem state? | Central limit theorem (CLT) states that, the arithmetic mean of a sufficiently large number of iterates of independent random variables, each with a well-defined expected value and well-defined variance, will be approximately normally distributed | 46 | |
| 1584663656 | What is a type 1 error? | That you reject a true Ho | 47 | |
| 1584663657 | What is a type 2 error? | That you fail to reject a false Ho | 48 | |
| 1584663658 | What is the probability of making a type 1 error? | The p-value is the probability of making a type 1 error, GIVEN THAT THE NULL HYPOTHESIS IS TRUE. If the null hypothesis is false, then it is impossible to make a type 1 error. | 49 | |
| 1584676198 | In the population, there is no difference between men and women on a certain test. However, you found a difference in your sample. The probability value for the data was .03, so you rejected the null hypothesis. What type of error did you make? | Type 1 | 50 | |
| 1584676199 | It has been shown many times that on a certain memory test, recognition is substantially better than recall. However, the probability value for the data from your sample was .12, so you were unable to reject the null hypothesis that recall and recognition produce the same results. What type of error did you make? | Type 2 | 51 | |
| 1584676200 | As the p-value gets lower, which error rate also gets lower? | Type 1 | 52 | |
| 1584676201 | If the null hypothesis is false, you cannot make which kind of error? | Type 1 | 53 | |
| 1584696096 | What is the "R. Fisher approach" to conducting significance tests? | If the probability is below 0.01, the data provide strong evidence that the null hypothesis is false. If the probability value is below 0.05 but larger than 0.01, then the null hypothesis is typically rejected, but not with as much confidence as it would be if the probability value were below 0.01. Probability values between 0.05 and 0.10 provide weak evidence against the null hypothesis and, by convention, are not considered low enough to justify rejecting it | 54 | |
| 1584696097 | What is the "Neyman and Pearson approach" to conducting significance tests? | The analyst specify an α level before analyzing the data. If the data analysis results in a probability value below the α level, then the null hypothesis is rejected; if it is not, then the null hypothesis is not rejected. According to this perspective, if a result is significant, then it does not matter how significant it is. Moreover, if it is not significant, then it does not matter how close to being significant it is. Therefore, if the 0.05 level is being used, then probability values of 0.049 and 0.001 are treated identically. Similarly, probability values of 0.06 and 0.34 are treated identically. | 55 | |
| 1584696098 | What are the benefits of using log functions? | Convenient percentage/elasticity interpretation Slope coefficients of logged variables are invariant to rescalings Taking logs often eliminates/mitigates problems with outliers Taking logs often helps to secure normality and homoscedasticity | 56 | |
| 1584696099 | When should log functions not be used? | Variables measured in units such as years and percentage points should not be logged Logs must not be used if variables take on zero or negative values | 57 | |
| 1584698638 | What is the equation for adjusted R2? |  | 58 | |
| 1584713440 | How can you tell if a model has omitted variable bias? | occurs when a model is created which incorrectly leaves out one or more important causal factors. The "bias" is created when the model compensates for the missing factor by over- or underestimating the effect of one of the other factors. | 59 | |
| 1586244699 | What is the marginal effect of a log-log model? | (%∆y) / (%∆x) | 60 | |
| 1586244700 | What is the marginal effect og a log-lin model? | (%∆y) / (∆x) | 61 | |
| 1586244701 | What is the marginal effect of a lin-log model? | (∆x) / (%∆y) | 62 | |
| 1586244702 | What is the interpretation of the log-log model ln(y) = -0,85ln(Xi) ? Where y = sales and x= price? | A 1% increase in price yields a 0,85% decrease in sales | 63 | |
| 1586244703 | What is the interpretation of the log-lin model ln(y) = 0,08X, where y=wage and x= yrs education? | An additional year of education yields an 8% increase in wage | 64 | |
| 1586244704 | What is the interpretation of the lin-log model y=65,32ln(x), where y= consumption spending and x=income in $ ? | A 1% increase in X yields a 65,32/100 change in y, or $0,6532 | 65 | |
| 1586967455 | What is the interpretation of the lin-log model y=36,42ln(x), where y= test score and x=income ? | A 1% increase in income yields a 0,3642 increase in test scores | 66 | |
| 1586967456 | What is the interpretation of the log-lin model ln(y) = 0,0128X, where y=earnings and x= age? | A 1 year increase in age yields a 1,28% increase in earnings | 67 | |
| 1586967457 | What is the interpretation of the log-log model ln(y) = 0,0554ln(Xi) ? Where y = test scores and x= income? | A 1% increase in income yields a 0,0554% increase in test scores | 68 | |
| 1586967458 | Explain how to interpret Lin-log models | A 1% increase in X gives a 0,01BETA increase in Y | 69 | |
| 1586967459 | Explain how to interpret log-lin models | A 1 unit increase in X yields a 100*BETA % increase in Y | 70 | |
| 1586967460 | Explain how to interpret log-log models | A 1% increase in X yields a BETA %increase in Y | 71 | |
| 1586967461 | What is important to note about linearity and parameters/variables? | We assume linearity in the PARAMETERS, not the variables | 72 | |
| 1586974695 | What is the formal equation for the confidence interval? | β ̂ - (t_crit)*(seβ ̂ ) ≤ β ≤ β + (t_crit) * (seβ ̂ ) | 73 | |
| 1586974696 | When using T-test when do you reject the null? | When Tcal > tTcrit | 74 | |
| 1586974697 | What is the equation for the F-statistic? | F = (R2/K)/((1-R2)/(n-k-1)) | 75 | |
| 1586974698 | What does the F-statistic meassure? | We are testing whether ANY of the B's are statistically signifikant | 76 | |
| 1586974699 | What is variance inflation factor? | The variance inflation factor tells you how much a parameter is inflated, compared to if the variable had been completely uncorrelated to the other variables in the model. It means that there is some degree of multicollinearity in the model. | 77 | |
| 1586974700 | What would it mean if the variance inflation factor of a predictor variable were 5.27 ? | This means that the standard error for the coefficient of that predictor variable is 2.3 times (√5.27 = 2.3) as large as it would be if that predictor variable were uncorrelated with the other predictor variables. | 78 | |
| 1586976958 | What is the rule of thumb for the max size of VIF? | Below 10 | 79 | |
| 1586976959 | If you have multicollinearity (high VIF) in a model, what can you do? | 1) Drop the offending variable if it makes sense. 2) Get more data if possible 3) Live with it and explain the issue. | 80 | |
| 1844430226 | True or False: The heteroskedasticity-robust standard errors are always bigger than the usual standard errors | False. While most often true, it is not a given fact. | 81 | |
| 1897390532 | If there is heteroskedasticity present, can an OLS estimator still be unbiased and efficient? | The estimator can still be unbiased (1-4), but the estimator will be asymtotically inefficient in the presence of heteroskedasticity. | 82 | |
| 1904556233 | Whats is the difference between having MLR 1-4, MLR 1-6 and MLR 1-5+large sample? | MLR 1-4: Unbiased and consistent. MLR 1-6 is Exact tests MLR1-5+"large sample" means asymtotically valid tests | 83 | |
| 1904556234 | What is MLR 6? | Normality in the error term | 84 | |
| 1904561649 | What are the four causes of endogeniety? | Wrong functional form. (^2 or similar) Omitted variable bias. Simultaneity Measurement error | 85 | |
| 1904561650 | Which 2 things needs to be fulfilled in order to have an omitted variable bias? | An effect: B≠0 and a correlation (X1, X2) ≠ 0 | 86 | |
| 1904569406 | When carrying out hypothesis tests, which two interpretations are important? And what should always be mentioned before the null hypothesis? | Both the statitic and the economic 1) Assumptions 2) Test-size t(375)=xx.x (show how it's calculated) + significance level 3) Conclusion 4) A nice drawing of the standard normal | 87 | |
| 1904577419 | How is the correlation calculated? | COV(A,B)/ (σa*σb) | 88 | |
| 1904639821 | Breush pagan and graphic tests are coming for the examn | ... | 89 | |
| 1906294551 | What is the first order condition for OLS? | SUM(Yi-Bohat-B1ihatX1i-B2ihatX2i-...-BjihatXji) = 0 | 90 | |
| 2096510629 | Which moment is SLR4? | The first moment | 91 | |
| 2096510650 | Which moment is SLR5? | The 2nd moment | 92 | |
| 2096510981 | What is the mathematical equation for MLR 5? |  | 93 | |
| 2096513282 | What is the mathematical equation for the variance of the OLS estimator? |  | 94 | |
| 2096514097 | The true variance is unknown, but how can it be estimated and what is the formula? (for a simple model with 1 regressor) |  | 95 | |
| 2096517526 | What is the matrix formulation of the OLS estimator? |  | 96 | |
| 2096524148 | How can you relate two SLR models compared to a single MLR model with aux regress? | 1) You regress X1 on X2 2) Then regress Y on the errors from this regression | 97 | |
| 2096527200 | What are causes for violations of MLR4? | 1) Incorrect functional form 2) Omitted variables 3) Simultaneity (x causes y, y causes x) 4) Measurement error (attenuation bias) | 98 | |
| 2096537809 | What is a general formula for omitted variable bias? |  | 99 | |
| 2096538323 | What is the formula for homoskedasticity? |  | 100 | |
| 2096542355 | What are the needed steps for hypothesis testing? |  | 101 | |
| 2096547485 | What is the equation for inconsistency (asymptotic bias) in the OLS? (simple) |  | 102 | |
| 2096552979 | When is the OLS asymtotically efficient? | Under MLR1-5 + a large sample | 103 | |
| 2096553814 | What is the equation for the LM test statistic? |  | 104 | |
| 2096554460 | How do you carry out an LM test? |  | 105 | |
| 2097430993 | What is the equation for the chow test with two groups? |  | 106 | |
| 2097431313 | What is the general form of the chow test? |  | 107 | |
| 2097432745 | How does the variance look with heteroskedasticity? |  | 108 | |
| 2097433388 | What are the implications if the errors are heteroscedastic? |  | 109 | |
| 2097434087 | What is the equation for the robust variance? (White 1980) |  | 110 | |
| 2097447354 | How does robust OLS and weighted least squares perform in small samples? | WLS performs fine, but Robust OLS needs a large sample to be asymtotically valie. | 111 | |
| 2097451710 | What is the equation for the Breusch-Pagan test and what test is used? |  | 112 | |
| 2098567714 | What is the null and alternative hypothesis for the Breusch Pagan test? | Ho= Homoskedasticity Ha= Heteroskedasticity | 113 | |
| 2098569820 | How is the classic (big) white test formulated? | The white test, tests that u^2 is uncorrelated with 1) the explanatory variables, 2) their squared terms and 3) the cross products. Hence, for k=3 there will be 9 regressors. It will test against a LM test n*R2 | 114 | |
| 2098575419 | How is the simplified White test formulated? | The simplified white test, tests that u^2 is uncorrelated with 1) the explanatory variables, 2) their squared terms | 115 | |
| 2098578136 | What is the null and alternatively hypothesis in the white test? |  | 116 | |
| 2098586206 | How do you apply weighted least squares? | If the type of heteroscedasticity is a known function of the explanatory variables "h(x)", by dividing all parameters (and the error term) with the square root of h(x) | 117 | |
| 2098590748 | What is the formal equation for weighted least squares? |  | 118 | |
| 2098591166 | What is the error term in WLS, and what is it's properties? (both regarding expectations and variance) |  | 119 | |
| 2098592512 | What is the intercept of WLS? Is WLS unbiased, efficient and BLUE? Which tests are valid using WLS? |  | 120 | |
| 2098599375 | A |  | 121 | |
| 2098599565 | A |  | 122 | |
| 2098600956 | Described in words, what does Feasible GLS do? | Feasible GLS tries to model the heterescedasticity and thereafter correct for it. | 123 | |
| 2098605341 | What is the procedure for FGLS? |  | 124 | |
| 2098647670 | What are the two main assumptions(requirements) an instrument "Z" has to satisfy? |  | 125 | |
| 2098650249 | What is the equation for the IV estimator? |  | 126 | |
| 2098651575 | What are generally the main issues with using the IV-estimator? |  | 127 | |
| 2098654191 | What happens if you have a "weak" instrument (z and x are not highly correlated)? | A low correlation will greatly inflate the variance of the IV estimator. | 128 | |
| 2098655858 | What is the difference between the variance of the of the OLS estimator and the IV estimator? |  | 129 | |
| 2098662940 | What are the requirements for the instruments in a IV regression with more than one instrument? |  | 130 | |
| 2098664032 | What is the matrix notation for the IV estimator? |  | 131 | |
| 2098666902 | IV Q1 - answer next slide |  | 132 | |
| 2098667071 | IV Q1 - answer | C | 133 | |
| 2098674754 | What are the 2 steps for 2SLS? |  | 134 | |
| 2098687870 | What does the dif-in-dif estimator measure? | The dif-in-dif estimator measures the difference between a policy change over time. | 135 | |
| 2102602150 | What are the assumptions for the first differences model and which assumptions are different from the OLS? |  | 136 | |
| 2102605743 | What is the equation for the first difference estimator? |  | 137 | |
| 2102606377 | How does OLS estimates compare to FD estimates and which should you use? |  | 138 | |
| 2102609188 | What is the equation for the fixed effects estimator? |  | 139 |
