Policy Evaluation Analytics Final Exam

Part I. Multiple Choice: Place an “X” by the choice that best answers the question or completes the statement. Be precise in placement of your choice.
1 The Central Tendency Theorem (aka “Law of Large Numbers”) is the principle that given a distribution of a parameter:
* Is foundational to all statistical analysis
* Distribution of a measured variables should be normally or near normally distributed X
* There is a center point of all observed values in a data set
* That measured values of a data set will tend to gather around an arithmetic average of the data observed values
* All of the above
2 A normal distribution is a data set comprised of:
* An arithmetic mean as a central tendency
* Is symmetrical in shape;
* Has a medium and mode equal to the sample mean
* Has a calculable standard deviation of the mean
* All of the above- X
3 In a right skewed (aka “Positively Skewed) distribution the characteristics of the central tendency are:
* Mean, mode and median are all equal
* Mean is less than the Median and Mode
* Mean is greater than the Median and the Mode - X
* None of the above
4. In a left skewed (aka Negatively Skewed) distribution, the characteristics of
the central tendency is:
* Mean, mode and median are all the same, or equal
* Mean is greater than the mode and median
* Mean is less than the Median, and Median is less than the Mode.- X
(d) None of the above.
4 A data set of a measured variable that is normally distributed is important because it presents empirical rules that describe the percentage of the observed values that fall within 3 standard deviations of the mean, which are:
(a) .60-.87-.95
(b) .67-.88-.99
(x) .68-.95-.997 X
(d) .68-.94-.996
(e) None of the above.
5 Correlation between variables is:
(a) The relationship between ordinal variables in a statistical study.
(b) The strength of the relationship between every combination of 2 quantifiable variables in a study.
(c) The strength and direction of the linear relationship between every combination of 2 quantifiable variables in a statistical study. X
(e) None of the above.
8 Variables in a statistical study can be correlated within the range of:
(a) 0.0 to +1.0
(b) 0.0 to –1.0.
(x) 0.0 to +/- 1.0 X
(d) 0.0 to +/- any whole number.
1 Regression Analysis is:
* A means by which correlated relationships between variables is determined and quantified X
* A means to determine whether the relationship between two quantifiable variables are positive or negative;
* A means to test the hypothesis of cause –effect relationship between the quantifiable variables in a statistical study.
* None of the above.
2 In Regression analysis, the study variables are cast in a regression equation where the variables can be:
* Latent and objective
* Responsive or explanatory
* Are dependent or independent
* All of the above X
3 In a Regression equation there are how many dependent variables and how many independent variables;
* Not more than 2 dependent and not more than 6 independent
* One dependent that must be a latent variable, and any number of independent variables.
* One dependent and any number of independent variables limited only by the idea of “parsimony”. X
* None of the above.
4 In Regression Analysis the prime measure of the causative effect of the regression model (equation) is R2, which measures:
* The degree of proof that a causal relationship exists in the regression equation;
* Measures the degree of causative effect of the dependent variable on the right hand side of the regression equation;
* Measures the amount of variation of the dependent variable that is explained by the right side of the regression equation. X
* None of the above.
5 Assume a statistical study of “Satisfaction at Work” includes eight variables. One combination of variables “Engagement at Work” and ‘Enjoyment at Work” were correlated on the Pearson Correlation Matrix at + .92. What is the problem with this correlation measure “r” for the study?
* The correlation, r = .92 suggests too close a correlation between the two variables and presents a possible problem of multicollinearity; X
* There is no problem so long as the measured correlation, r, is correct;
* Both a and b;
* None of the above.
14. Probability is defined as:
* The predicted value of a particular outcome of an event;
* The possible outcome of an event or experiment;
* The likelihood that a particular event will occur expressed as a ratio of the observed frequency of an outcome to the total frequency of possible outcomes; X
* None of the above.
15. Given one roll of two honest die where the raw sample space is 36, the probability of any particular outcome given the sum of the two die when rolled within the sample space is:
* 1 out of 36, or .027; X
* (1 out of 6 or .16666) x 2 = 0.3332;
* 4/36 or 1/9 = 0.111
* None of the above.
16. The probability of all possible outcomes in a single game of chance is equal to:
* 100;
* The Sum of all the possible outcomes divided by the number of total observations;
* P (0) = 1.0 X
* None of the above.
17. The probability of all possible outcomes in 5 independent games of chance, such as investing in common stocks on the New York Stock Exchange, is equal to
(a) The sum of each independent outcome, the total of which must be 1.0.
(b) Independent games of chance do not have to have the same probabilities of outcomes, so the question cannot be answered;
(c) P (0) = P(A) x P(B) x P(C) x P(D) x P(E);
(d) None of the above. X
18. In Hypothesis Testing the “Null Hypothesis”, Ho, assumes and means:
* All observed results in regression analysis are the result of systemic causes;
* All observed results in the regression analysis are the result of random chance; X
* Has nothing to do with regression analysis;
* None of the above.
19.. In Hypothesis testing the “Alternative Hypothesis”, H(A), assumes and means:
* All observed results in regression analysis are the results of random chance;
* Observed results in the regression analysis are the results of systemic causes with random change having limited probability of effecting the results;
* Has nothing to do with regression analysis; X
* All of the above;
20. Statistical Significance means:
* A hypothesis tests the results of which rejects the Alternative Hypothesis; X
* A hypothesis tests the results of which rejects the Null Hypothesis;
* A t-test that yields a probability value for the observed value of the variable under study that is so extreme that the observed value is unlikely to have been caused by chance.
* None of the above.
1 Type I and Type II Error can exist in hypothesis testing. These types of Errors can be thought of as:
* Type I error is when the truce condition is H(A) and the Null is accepted as true; And, Type II error is when the true condition is H(o) and the Alternative, H(A), is accepted.
* Type I error is when the true condition is H(O), yet the Null is rejected in favor of H(A) and Type II error is when the true condition is H(A) and the Null is accepted as true; X
* Neither (a) or (b);
* Both (a) and (b).
2 Preconditions to inferential statistical analysis are:
(a) Having an SRS population with a normal or near normal distribution, which yields descriptive statistics; X
(b) Correlated variables and a regression equation with a sufficient R2 measure;
(c) A regression model that yields statistical s...