Social Stats Final Questions

The level of measurement for “degree of satisfaction with your cell phony company” (very satisfied; somewhat satisfied; not at all satisfied) is…

Ordinal

Imagine a business corporation where the vast majority of employees receive relatively low salaries, but a very small group of employees has extremely high salaries.  The distribution of salaries is thus skewed to the right.  If you were a bargaining agent for the employees, you could calculate the _______________ to illustrate a lower pay level, on average, for most employees.  But if you were the employer, you would calculate the _______________ to illustrate a higher average pay level.

median; mean

Working with a sample of people from an Atlanta neighborhood, Dr. Jones calculates the mean and standard deviation for the variable “years of schooling.”  She finds the mean for the sample is 14 years of schooling, and the standard deviation for the sample is 2 years of schooling.  Given this information, which of the following interpretations of the sample's standard deviation is correct?

The average number of years by which cases differed from the sample mean is 2 years.

The spread (or variability) in a set of values is lower in a _____________ distribution than in a ____________ distribution.  

leptokurtic; mesokurtic

A _______________ is a characteristic of a sample, whereas a _______________ is a characteristic of a population.

statistic; parameter

The Empirical Rule is a description of…

The area underneath a normal curve

Following the rough version of the Empirical Rule, we can say that about 68 percent of values fall within one standard deviation above and below the mean of values, and about 95 percent of values fall within two standard deviations above and below the mean of values.

True

A z-score tells us…

How many standard deviation units a given value falls from the mean.

If a case has a z-score of -2.34, then the case has a value on Y that is 2.34 standard deviations above the mean.

False

If the z-score for a case is 3.28, then the case has a value on Y that is…

Uncommon because it is relatively far away from the mean.

1. In statistical inference, the _________ distribution "bridges the gap" between the _________ distribution and the _________ distribution.

sampling, sample, population

A sampling distribution shows the values that a particular statistic (such as the mean) can take across repeated random samplings of a given size n from the same population.  In other words, the sampling distribution is the distribution of a statistic rather than the distribution of individual y values on a variable.

True

The Central Limit Theorem describes the shape, center, and spread of the ____________.

Sampling distribution

The shape of the sampling distribution becomes more normal as n approaches infinity (where "n" refers to the size of the samples, not the number of samples).

True

Four plots are presented below. The plot at the top is a distribution for a population where the population mean is 60 and the population standard deviation is 18. The other plots are

(i)    a single random sample of 500 cases from this population

(ii)   a sampling distribution of 500 sample means from random samples of size n = 18

(iii)  a sampling distribution of 500 sample means from random samples of size n = 81

Determine which plot is which.

(i) is Plot B; (ii) is Plot C; and (iii) is Plot A

Confidence intervals are an example of ______________.

Inferential statistics

The level of confidence and the size of the sample both affect the precision of the confidence interval estimate.

True

Increasing the level of confidence will increase the width of the confidence interval, thus making the interval less precise.

true

Smaller samples result in larger standard errors, which in turn make the interval wider and less precise.

True

A random sample of 500 working-class households has a sample mean annual household income of $65,000.  An interval estimate of the population parameter at the 95% confidence level is $64,800 to $65,200. Which of the following interpretations of the interval is correct? 

if random samples of 500 homes were repeatedly selected, and we calculated 95% confidence intervals around each sample mean income, then 95 percent of the time the intervals we compute will contain the true population mean

Which of the following statements is true?

C. Both A and B are true

When the p-value is less than or equal to the alpha level, we reject the null hypothesis.

True

When the p-value is less than the alpha level, then the TS_sample will be greater than the TS_critical and therefore will fall into the rejection region.

true

Which of the following statements is true?

C. Both A and B are true

When the null hypothesis value falls within the confidence interval, this tells you to reject the null hypothesis.

False

In bivariate analysis, y (the dependent variable) is the hypothesized cause of x (the independent variable).

False

Which question needs to be addressed in order to bolster a causal argument?

All of the above questions need to be addressed

Which of the following statements about statistical significance is false?

If the relationship between two variables is "statistically significant," then by definition it is also "substantively significant," i.e., it has theoretical or practical importance.

If you are using a confidence interval of the difference-in-means to test the null hypothesis that the population difference between groups is zero, then which of the following would be the correct thing to do?

c. Both a and b would be the correct thing to do

Which of the following statements about Student's t-distribution is true?

d. All of the above are true.

Which of the following statements about contingency tables (i.e., crosstabulations) is FALSE?

Contingency tables are best used when the variables have a large number of categories because then the pattern of the relationship is easier to interpret.

If the independent variable is placed on the columns of a contingency table, then we  need to calculate ____________ percentages to describe the pattern of the association.

column

Which of the following statements about chi-square are TRUE?

d. both a and c are true

Which of the following contingency tables has more degrees of freedom?

4×3 table

The larger the dimensions of the contingency table (i.e., the more rows and columns), the more opportunities there are to balance out a fluke sampling event with other cell frequencies that are “free” or independent of it.

true

Which of the following statements about measures of strength is FALSE?

We prefer such measures to be based on chi-squared because then the interpretation will be clear and intuitive.

Which of the following statements about lambda is FALSE?

Lambda is most appropriate when both variables are ordinal.

If the mode on y is the same across every category of x, then lambda …

c. both a and b are correct

Which of the following statements about gamma is FALSE?

Gamma is not a PRE-based measure of strength.

Professor Jones calculated a gamma of -.74 and asked you for advice on how to interpret it.  Which one of the following statements would be appropriate when advising Professor Jones?

This gamma indicates that knowing the order of a pair of cases on x will reduce error in predicting their order on y by 74 percent.

Using the General Social Survey, Professor Twining calculated a gamma of .47 (p < .05) for a crosstabulation between subjective social class identification (measured as low, middle, high) and level of education (measured as less than high school, high school graduate, college graduate, etc.).  She concluded that education has a moderate and statistically significant effect on social class identity.  She then wondered whether the strength of this association varied by the age of the respondent.  Specifically, she hypothesized that level of education would have a much stronger effect on social class identity among elderly Americans than for young Americans.  She found the gamma for elderly Americans equaled .48 (p < .05) and the gamma for young Americans equaled .46  (p < .05).  In this example, Professor Twining investigated the possibility of ____________________ but what she found instead was ____________________ .

interaction; replication

Professor Grey hypothesized that those people who read public health warnings about the risks of smoking cigarettes will be more likely to quit smoking.  He collected a large random sample of people who are current and former smokers, and he discovered that yes, indeed, those people who said they had read the warnings were significantly more likely to say they had quit. Professor Grey was tempted to conclude that a causal relationship exists between reading warnings and quitting smoking.  However, before reaching that conclusion, he controlled for prior level of education, and discovered that the original association between reading warnings and quitting smoking was reduced to a weak and insignificant effect.  He diagrammed the three-variable relationship as follows:

In this example, Professor Grey finds that the apparent relationship between reading warnings and quitting smoking was  __________________.

spurious

The following diagram is from our in-class example on the social psychology of homelessness. The diagram presents a causal model explaining why those who are homeless for longer periods of time have higher levels of psychological depression.   The causal logic of the model can be summarized like this:  As the length of homelessness increases, levels of social support and sense of mastery both decrease, which in turn increases the level of psychological depression.    

In this diagram, how many variables are endogenous variables?  

3

Refer to the diagram in question #3 above.  Which variable has an indirect effect on depression?

length of homelessness

Refer to the diagram above. Which variables are intervening variables?

social support and sense of mastery

Which one of the following statements is TRUE?

The loess curve allows us to check whether the linearity assumption is met.

Which of the following correlation coefficients indicates the strongest linear relationship between variables?

-.803

If r = -.60, then r-squared = .36.  Given this information, which of the following statements is correct? 

the two variables have a negative association

If r = 0 and b = 0, then it’s still possible for a strong relationship between variables to exist; for example, the relationship may be strongly curvilinear.

true

Which of the following statements is TRUE?

All of the above statements are true