20 Questions
Match the following regression models with their characteristics:
Probit regression model = Dependent variable is a binary variable Logit regression model = Predicted values are between 0 and 1 Linear regression model = Predicted values can be any real number Logistic regression model = Uses the logistic function for predictions
Match the following terms with their meanings:
Cumulative logistic distribution = Used in logit models for cumulative probabilities Pecking order hypothesis = Corporations should use cheapest financing methods first Asymptotes = 0 and 1 are asymptotes in logistic regression models Exponential function = Used in logit models to calculate probabilities
Match the following researchers with their study topic:
Helwege and Liang = Study of the pecking order hypothesis Myers = 'Pecking order hypothesis' proposer Probit and Logit models = Nonlinear regression models for binary variables Firm financing theory = Suggests using cheapest financing sources first
Match the following decision rules with their corresponding descriptions:
If d < dL = There probably is evidence of positive autocorrelation. If d > dU = There probably is no evidence of positive autocorrelation. If dL < d < dU = No definite conclusion about positive autocorrelation may be made. If 4-dL < d < 4 = There probably is evidence of negative autocorrelation.
Match the following ranges of d values with their interpretations:
0 ≤ d < dL = Positive autocorrelation likely. dU < d < 4 = Negative autocorrelation likely. 4-dU < d < 4-dL = No definite conclusion about negative autocorrelation. 2 = No evidence of positive or negative (first-order) autocorrelation.
Match the following limits with their definitions:
dL = Lower limit established for detecting autocorrelation. dU = Upper limit established for detecting autocorrelation. 0 = The minimum value the d statistic can take. 4 = The maximum value the d statistic can take.
Match the following statements with their correct descriptions:
Closer to 0 = Greater evidence of positive autocorrelation. Closer to 4 = Greater evidence of negative autocorrelation. dU < d < 4-dU = No definite conclusion about positive or negative autocorrelation. 4-dU< d < 4-dL = No definite conclusion about negative autocorrelation.
Match the following terms with their definitions:
Explained Sum of Squares (ESS) = Sum of squares explained by the chosen model Residual Sum of Squares (RSS) = Sum of squares not explained by the model Total Sum of Squares (TSS) = The total sum of squares in the data set Degrees of Freedom (df) = Number of independent values that can vary in the calculation of a statistic
Match the following statistics with their descriptions:
R-square = Proportion of the variance in the dependent variable that is predictable from the independent variables Adj R-square = Adjusted R-square value that penalizes excessive use of independent variables in regression models Root MSE = Root Mean Square Error, a measure of the differences between values predicted by a model and the actual values F-statistic = A ratio indicating whether there is a significant relationship between dependent and independent variables in regression analysis
Match the following terms with their meanings:
ANOVA = Analysis of Variance, a statistical method used to analyze differences among group means Regression Coefficients = Values that represent the change in the dependent variable for a one-unit change in the independent variable Hypothesis Testing = Statistical method to make inferences about a population based on sample data Confidence Interval = Range of values within which the true population parameter is likely to fall
Match the following concepts with their explanations:
Model Residual = The difference between observed and predicted values in a regression model Overall Significance Test = Test to determine if all slope coefficients are simultaneously equal to zero in a regression model Observed Data Points = Actual data values collected from observations or experiments Mean Squared Error (Ms) = Average of squared differences between predicted and actual values in a regression model
Match the following concepts with their description:
Fitted values from regression = Estimated probabilities for yi = 1 for each observation i Slope estimates in linear probability model = Change in the probability that the dependent variable will equal 1 for a one-unit change in a given explanatory variable Truncation in the model = Resulting in too many observations with estimated probabilities being exactly zero or one Logit or probit model = Usually used for binary dependent variables instead of linear probability model
Match the following variables with their interpretation:
Market capitalisation (x2i) = Measured in millions of dollars Coefficient -0.3 in the model = Constant term in the probability estimation Coefficient 0.012 in the model = Effect on probability of dividend payment per unit change in market capitalisation Probability of dividend payment = Modeled as a function of market capitalisation in the example
Match the following statements with their implication:
Many estimated probabilities are exactly zero or one = Result of process of truncation Firm's probability of paying a dividend is not exactly zero or one = Reason why linear probability model is not adequate Different kind of model needed for binary dependent variables = Use of logit or probit specification Small firms vs. large firms dividend payment certainty = Uncertainty regarding definite dividend payment based on firm size
Match the probability model with its description:
Linear Probability Model (LPM) = Simplest way of dealing with binary dependent variables Logit Model = Probability model based on logistic function Probit Model = Probability model based on cumulative distribution function Tobit Model = Model used for censored regression
Match the approach with its description:
Linear Probability Model (LPM) = Can be estimated by ordinary least squares (OLS) Logit Model = Model for binary response based on logistic function Probit Model = Model for binary response based on cumulative distribution function Tobit Model = Model used when dependent variable has a censoring point
Match the variable with its role in the linear probability model:
Pi = Probability of an event occurring Yi = Dependent variable representing outcomes (0s and 1s) x2i, x3i, … xki = Explanatory variables in the model ui = Error term in the model
Match the estimation method with the type of model:
OLS = Used to estimate a linear regression model for binary dependent variables Logit Model = Estimated using maximum likelihood estimation Probit Model = Estimated using maximum likelihood estimation Tobit Model = Estimated using a censored regression approach
Match the linear probability model with its characteristics:
Simplest way of dealing with binary dependent variables = Linear Probability Model (LPM) Based on assumption that Pi is linearly related to explanatory variables = Linear Probability Model (LPM) Utilizes OLS for estimation = Linear Probability Model (LPM) Dependent variable represented by a series of zeros and ones = Linear Probability Model (LPM)
Match the types of variables that can be included in the linear probability model:
Quantitative variables or dummies or both = Explanatory variables in the model Binary dependent variables only = Continuous dependent variables only = Categorical variables only =
Test your knowledge on dummy dependent variable models and different approaches to developing a probability model including the linear probability model, logit model, probit model, and tobit model. Learn about estimating models using ordinary least squares (OLS).
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free
