Answer the following multiple choice questions and check whether you have understood the concepts seen in the class. Each question has only one correct answer. If you think that there is more than one correct answer, choose the best answer. In such cases, please feel free to send your comments to your instructor.

Question: In science we think of causality in Ans: deterministic terms. Ans: probabilistic terms. Ans: necessary and sufficient terms. Ans: predictive terms. Correct_Answer: 2 Question: There is a strong correlation between the number of storks nesting in town and the number of births. This is an example of Ans: a strong causal connection. Ans: a necessary correlation. Ans: a spurious correlation. Ans: a confusion of temporal order. Correct_Answer: 3 Question: A necessary cause is a cause which Ans: by itself can produce the effect. Ans: is highly correlated with the effect. Ans: must be present for the effect to occur. Ans: all of the above. Correct_Answer: 3 Question: A sufficient cause is a cause which Ans: by itself can produce the effect. Ans: is highly correlated with the effect. Ans: must be present for the effect to occur. Ans: all of the above. Correct_Answer: 1 Question: There must be a very strong association between two variables if we are to assume a causal connection between them. Ans: true Ans: false Correct_Answer: 2 Question: When attempting to determine causality, the issue of temporal order is rarely a problem in social science. Ans: true Ans: false Correct_Answer: 2 Question: When controlling for a third variable we are essentially Ans: holding constant the values of the dependent variable so that they are no longer free to vary. Ans: fixing the values of the independent variable so that they are no longer free to vary Ans: fixing the values of the control variable so that they are not free to vary Ans: eliminating the influence of the independent variable in the analysis of the bivariate relationship Correct_Answer: 3 Question: Introducing a control variable into a bivariate analysis will result in a weaker correlation between the dependent and independent variables Ans: always Ans: sometimes Ans: never Correct_Answer: 2 Question: Partial gamma (or lambda) Ans: compares and contrasts the gammas (lambdas) from each partial table. Ans: measures the association between X and Y controlling for a third variable Z. Ans: is always smaller than the bivariate gamma (lambda) Ans: is the sum of the individual gammas (lambdas) computed from each partial table. Correct_Answer: 2 Question: The gamma (or lambda) of a bivariate table is 0.45. After controlling for a third variable, partial gamma (lambda) is also 0.45. What can we say about the original relationship? Ans: it was spurious Ans: it was suppressed. Ans: it was chain relationship. Ans: not enough information to say anything. Correct_Answer: 4 Question: The lambda for a bivariate table is 0.62. The lambdas computed from three partial tables representing three categories of the control variable are .04, .01, and .02. This is evidence of Ans: a spurious relationship. Ans: a suppressed relationship. Ans: an interaction. Ans: a chain relationship. Correct_Answer: 1 Question: The gamma is 0.4. Two partial gammas of 0.5 and 0.56 are computed., This elaboration outcome is evidence Ans: for doubting the zero-order relationship between X and Y Ans: that the control variable is suppressing the "true" strength of the relationship between X and Y. Ans: that X and Y are only related due to their joint relationship with the control variable. Ans: that the control variable exerts its influence on Y before X does. Correct_Answer: 2 Question: When selecting a control variable to elaborate a bivariate relationship, the main selection criterion should be Ans: the variable that results in the strongest partial measures. Ans: the variable that does not destroy the zero-order relationship Ans: the variable that is the most theoretically meaningful in terms of its connection with X and Y Ans: the variable that essentially replicates the zero-order relationship since we already know such a relationship exists. Correct_Answer: 3 Question: A problem with bivariate elaboration when controlling for more than one variable is that Ans: samples must be quite large Ans: it is not uncommon to find empty table cells Ans: the interpretation of multiple tables tends to get messy Ans: all the above. Correct_Answer: 4 Question: An intervening variable is one that precedes both the independent and dependent variables in time Ans: true Ans: false Correct_Answer: 2 Question: Partial gamma or lambda is an index of the strength of association between two variables controlling for the effects of one or more other variables. Ans: true Ans: false Correct_Answer: 1 Question: Suppose that we wish to assess the relationship between X and Y. Would Z, which shows no relationship with X or Y be appropriate as a control variable? Ans: Yes, but only if Z is interval Ans: Yes, but only if Z is nominal. Ans: No, it is irrelevant. Ans: No, it will bias the results. Correct_Answer: 3 Question: If we control for a third variable Z, and find that the original X->Y relationship does not change, then Ans: the relationship between X and Y is probably spurious. Ans: the original relationship is spurious. Ans: there is no relationship between X, Y and Z. Ans: the original relationship is non-spurious. Correct_Answer: 4 Question: We find that the relationship between social class and mobility orientation is mediated by the variable cultural background. The third variable is Ans: the control variable Ans: the spurious variable Ans: an intervening variable. Ans: a suppressor variable. Correct_Answer: 3 Question: Suppose we determine that there is a relationship between sex and voting. Moreover, we find that "level of interest" shows that the less a group is interested in the election, the greater will be the amount of deliberate non-voting among women as compared to women. Level of interest is: Ans: a conditional variable Ans: an intervening variable Ans: a spurious variable Ans: a suppressor variable. Correct_Answer: 1 Question: A real or non-spurious relationship Ans: can be inferred from the time sequence of the variables Ans: cannot be inferred Ans: can be inferred from the size of the partial association Ans: is unknown in the social sciences. Correct_Answer: 1 FOOTER: