Story, Dale 10/19/2014
POLS 3310
REVIEW QUESTIONS FOR TEST 2
 Briefly define or otherwise show that you know the meaning of these levels of measurement. Give one example of each. Explain why it is an example.
 Interval
 Ordinal
 Nominal
 The four levels of measurement, ranked in terms of their level of sophistication, are nominal, ordinal, interval, and ratio. What is the level of measurement for each of the following measures? If the level of measurement is higher than nominal, give an example of the measure as it might be found at each lower level of measurement (for ratio level measurement you do not need to give an example of interval level measurement).
 Age, in years
 Religion, in six different categories according to how strict the religion is
 Type of government, in three categories: polyarchy, authoritarian, and totalitarian
 Voting results, in percentages according to party
 Political ideology, in five different categories ranging from far right to far left
 Give examples for nominal, ordinal, and interval level measures for these variables:
 Degree (or type) of political participation
 Government expenditures
 Describe and define the two types of measurement error—cite examples of and tests for these types of error.
What sample size would you need for a random sample if the population consisted of 10,000 sampling units, the population variance was 1.6, and the standard error of the sample was 0.04?
 Given these data for approval ratings for President Obama in ten regions of the nation. The population mean is 45. Using the random numbers table in FN & N (p. 500 ff), select a simple random sample of 4 sampling units. Calculate the sample mean, the standard error of the sample mean, and the 95% confidence interval (using plus or minus 1.96) around the sample mean. The data for the ten regions are: 49, 43, 45, 50, 44, 45, 41, 46, 47, 40.
 Calculate the sample mean, the standard error, and the 95% confidence interval around the sample mean from the following date on regional approval ratings for a president. This is a sample of 4 from a population of 10. The data are 43, 47, 48, 42
 Example of multiple choice question: The standard error of the sample mean for the above data (rounded to the nearest, single decimal point) is: (a) 4.8; (b) 1.5; (c) 6.2; (d) 0.6; or (e) none of the above.
 Define and identify the following. Cite examples as appropriate.
 Convenience/accidental sample
 Purposive/judgmental sample
 Quota sample
 Simple random sample
 Systematic sample
 Cluster sample
 Compare and contrast the three means of administrating surveys.
 Given these results of two Likert scaled items, which of these items has the highest “discriminatory power” (DP)? Use the DP index. Show all your work.

Group 
N 
1 
2 
3 
4 
5 
Item 1 
Highest 25% 
10 
0 
1 
2 
3 
4 

Lowest 25% 
10 
4 
3 
2 
1 
0 








Item 2 
Highest 25% 
10 
1 
1 
3 
2 
3 

Lowest 25% 
10 
3 
2 
2 
3 
0 








 Given this sample of ten respondents, calculate the mean and variance for each of these approval ratings for two presidents.

President 
President 

A 
B 
1 
78 
79 
2 
85 
83 
3 
90 
85 
4 
70 
77 
5 
75 
75 
6 
92 
90 
7 
82 
81 
8 
65 
75 
9 
68 
70 
10 
95 
85 



 If presidential approval ratings are normally distributed with a mean of 60 and a standard deviation of 8:
 What are the Z or standard scores for the following values of presidential approval rating.
 60
 80
 56
 68
 48
 What percentage of the cases:
 are less than 44 and greater than 68
 are less than 60
 are between 52 and 60
 are between 68 and 76
 are greater than 76.