Deck 5: Creating New Variables

Question 1

Which of the following Compute numerical expressions correctly calculates the white percent of a population? (WHITE is the number of persons white; NONWHITE is the number of persons not white.)

A) WHITE * 100
B) WHITE / NONWHITE * 100
C) WHITE / ( WHITE + NONWHITE ) * 100
D) 100 * ( WHITE - NONWHITE ) / WHITE

Answer 1

WHITE / ( WHITE + NONWHITE ) * 100

Question 2

A variable has valid codes of 1, 2, 3, and 4. You recode it so that cases with 1s are coded as 4s, 2s are coded as 3s, 3s are coded as 2s, and 4s are coded as 1s. What have you accomplished?

A) You have caused all the cases to be coded as 1s.
B) You have collapsed the coding.
C) You have reversed the coding.
D) You have split the file into four groups.

Answer 2

You have reversed the coding.

Question 3

The variable GENDER is coded 0 for males and 1 for females. EDUC is a person's years of schooling and SPEDUC is their spouse's years of schooling. The following Select If condition would limit the analysis to which persons? GENDER = 1 and EDUC > SPEDUC

A) men with more education than their wives
B) women who are married
C) women and persons with more education than their spouses
D) women with more education than their husbands

Answer 3

women with more education than their husbands

Question 4

Which of the following z-scores is closest to the mean?

A) −100
B) −1
C) 5
D) 100

Question 5

A data set contains information on 100 married couples. One of the variables is the husband's age; another is the wife's age. To create a new variable which is the difference between the husband's age and the wife's age, you would use which SPSS procedure?

A) Compute
B) Recode
C) Scale
D) Split File

Question 6

To limit an analysis to just a subset of the cases in a data set which meet a particular condition, you would use which SPSS procedure?

A) Compute
B) Recode
C) Select Cases
D) Split File

Question 7

Which of the following Compute numeric expressions takes the value of CARS, adds 3 to it, then multiplies the sum by 2, and then takes that product and adds 5 to it?

A) CARS + 3 * 2 + 5
B) CARS + 3 * ( 2 + 5 )
C) ( CARS + 3 ) * ( 2 + 5 )
D) ( CARS + 3 ) * 2) + 5

Question 8

Which of the following is correct about the order of arithmetic operations used by SPSS?

A) division is done before multiplication
B) multiplication is done before division
C) multiplication and division have equal priority
D) division is not permitted within SPSS

Question 9

What is the lowest level of measurement for which it is permissible to calculate Z-scores?

A) nominal
B) ordinal
C) interval/ratio

Question 10

A researcher is planning to create an index consisting of five variables: varA, varB, varC, varD, and varE. Based on the following output from the Reliability Analysis procedure, what would you recommend?
 
 

A) Drop varA.
B) Drop varB.
C) Create the index using all five variables.
D) There is no way to use all or some of these items to create an index with acceptable reliability.

Question 11

To divide the cases in a data set into separate subgroups in order to perform a particular statistical procedure on each subgroup, you would use which SPSS procedure?

A) Compute
B) Recode
C) Select Cases
D) Split File

Question 12

A survey includes 10 political statements. For each statement, respondents are asked to indicate their extent of agreement using a scale from 0 to 5. Half the statements are very liberal in nature, half are very conservative. The researcher wants to add up the scores on the 10 items to obtain an overall index of political liberalism. Before adding up the 10 scores, the researcher should do what with the conservative statement items?

A) convert them into z-scores using Descriptives
B) delete them using Select If
C) double their weight using Compute
D) reverse the coding using Recode

Question 13

Which of the following Select If conditions would limit the analysis to persons with less than 12 years of schooling who have 3 or more children?

A) SCHOOLING < 12 and CHILDREN > 3
B) SCHOOLING < 12 or CHILDREN > 3
C) SCHOOLING < 12 and CHILDREN >=3
D) SCHOOLING < 12 or CHILDREN >= 3

Question 14

Which of the following Select If conditions would limit the analysis to cases between the ages of 30 and 49?

A) AGE = 30 or AGE = 49
B) AGE <= 30 and AGE >=49
C) AGE >= 30 and AGE <=49
D) AGE >= 30 or AGE <=49

Question 15

You want to do a separate analysis for each of the following three groups and only for these three groups: persons with just 8 years of schooling, persons with just 12 years of schooling, and persons with just 16 years of schooling. What should you do?

A) set Select Cases to "SCHOOLING = 8 or SCHOOLING = 12 or SCHOOLING = 16"
B) set Select Cases to "SCHOOLING = 8 or SCHOOLING = 12 or SCHOOLING = 16" and put SCHOOLING into the Groups list on Split File
C) put SCHOOLING into the Groups list on Split File
D) Compute a new variable whose numeric expression is "SCHOOLING = 8 or SCHOOLING = 12 or SCHOOLING = 16"

Question 16

When all the values on a variable have been converted into z-scores, the mean

A) becomes 0
B) becomes 1
C) becomes 100
D) is the same as before the values were converted

Question 17

Which of the following is correct about the order of arithmetic operations used by SPSS?

A) addition is done before subtraction
B) anything within parentheses is done last
C) multiplication is done before addition
D) multiplication is done before division

Question 18

To collapse a variable with nine categories into a variable with just three categories while not losing any valid cases, you would use which SPSS procedure?

A) Compute
B) Recode
C) Select Cases
D) Split File

Question 19

You want to do a separate analysis for each region but only include persons born before 1950 in the analysis. What should you do?

A) set Select Cases to "REGION and YEAROFBIRTH < 1950"
B) put both REGION and YEAROFBIRTH into the Groups list on Split File
C) put REGION into the Groups list on Split File and set Select Cases to "YEAROFBIRTH < 1950"
D) put "YEAROFBIRTH < 1950" into the Groups list on Split File and set Select Cases to "REGION"

Question 20

To combine categories of a variable so that you end up with a variable with fewer categories, you would use which SPSS procedure?

A) Recode
B) Scale
C) Select Cases
D) Split File

Question 21

Which value of Cronbach's alpha indicates the greatest reliability?

A) .01
B) .05
C) .50
D) .95

Question 22

When would you use Recode and when would you use Compute?

Question 23

Why are z-scores useful?

Question 24

What are the two most common reasons for recoding a variable?

Question 25

Cronbach's alpha is a way to assess the _____ of an index. For an index to be considered "acceptable" for research purposes in the social sciences, alpha values should be _____.

A) reliability … below 0.25
B) reliability … 0.70 or higher
C) validity … below 0.25
D) validity … 0.70 or higher
(Note: * indicates

Question 26

"A set of items all assumed to be of approximately equal importance in measuring a single underlying concept that are then combined to yield a single overall score on the underlying concept" defines what the text calls a(n)

A) dimension
B) index
C) numeric variable
D) real variable

Question 27

You want to reverse the coding scheme for a variable that is currently coded 1 through 5. The new codes will still go from 1 through 5. Indicate below what the new values are for each of the old values.

Question 28

Write out a Select If condition that would limit the analysis to cases older than 59. (Variable: AGE)
_____________________________________

Question 29

A researcher uses the Compute procedure to create two new variables. The two new variables are called NEW1 and NEW2. The first case in the data set has a value of 10 on VARA and 2 on VARB. What values will that first case have on the two new variables?

A) if NEW1 = VARA − 8 / VARB then for the 1^st case NEW1 = _____
B) if NEW2 = VARA / 5 * VARB then for the 1^st case NEW2 = _____

Question 30

Write out a Select If condition that would limit the analysis to persons with a value of 1 on GROUP or a value of 3 on COHORT. (Variables: GROUP, COHORT)
_____________________________________

Question 31

When would you use Select Cases and when would you use Split File?

Question 32

You want to Compute a new variable that multiplies a person's hourly wage by their hours worked per week by their weeks worked per year. Write out the numeric expression that would accomplish this. (Variables: hourlyWage, hoursPerWeek, weeksPerYear)
_____________________________________

Question 33

You want to Compute a new variable that is the number of years older the firstborn child is than the lastborn child. Write out the numeric expression that would accomplish this. (Variables: FIRSTBORN, LASTBORN)
_____________________________________

Question 34

You want to Compute a new variable that adds 5 to the average of TEST1, TEST2, and TEST3. The new variable should be computed only if valid scores exist for all three tests. Write out the numeric expression that would accomplish this.
_____________________________________

Question 35

You want to collapse the coding scheme for a variable that is currently coded 1 through 5. The first two categories of the variable get combined and the last two categories of the variable get combined. The new codes will go from just 1 to 3. Indicate below what the new values are for each of the old values.

Question 36

Why would an analyst want to know the value of Cronbach's alpha?

Question 37

Write out a Select If condition that would limit the analysis to persons under the age of 30 who have been fired from their jobs 3 or more times. (Variables: AGE, TIMESFIRED)
_____________________________________

Question 38

The variable college grade point average was recorded for a sample of graduating college students. An analyst calculates the mean and standard deviation for the variable and transforms each person's score into a z-score. The first case has a z-score of 3. What does this z-score tell us about the first person's GPA?

Question 39

How can missing data be handled when constructing an index?

Question 40

A researcher uses the Compute procedure to create two new variables. The two new variables are called NEW1 and NEW2. The first case in the data set has a value of 10 on VARA and 2 on VARB. What values will that first case have on the two new variables?

A) if NEW1 = (VARA - 8) / VARB then for the 1^st case NEW1 = _____
B) if NEW2 = VARA - 8 / VARB then for the 1^st case NEW2 = _____

Question 41

You are computing a new variable called CRIMEALL. It is the average of four variables already in your data set: CRIMESA, CRIMESB, CRIMESC, and CRIMESD. Write the numeric expression (formula) for a compute command that averages these four variables but only if at least three of the CRIME variables being averaged have valid values.
_____________________________________

Question 42

A variable has a mean of 100 and a standard deviation of 10. Calculate the Z-scores for the first two cases and the raw scores for the last two cases.

Question 43

You are computing a new variable called CRIMEALL. It is the average of four variables already in your data set: CRIMESA, CRIMESB, CRIMESC, and CRIMESD. Write the numeric expression (formula) for a compute command that averages these four variables as long as at least one of the CRIME variables being averaged has a valid value.
_____________________________________

Question 44

You are creating an index based on four variables in your data set: VAR1, VAR2, VAR3, and VAR4. One of the cases in your data set has valid values for only VAR1 and VAR2. It has missing data for VAR3 and VAR4. For each of the following compute commands, indicate if SPSS would report a valid value or a missing value for this case on the index.

A) __________ index = SUM(VAR1,VAR2,VAR3,VAR4)
B) __________ index = SUM.4(VAR1,VAR2,VAR3,VAR4)

Question 45

You are computing a new variable called CRIMEALL. It is the average of four variables already in your data set: CRIMESA, CRIMESB, CRIMESC, and CRIMESD. Write the numeric expression (formula) for a compute command that averages these four variables but only if all four of the CRIME variables being averaged have valid values.
_____________________________________

Question 46

A variable has a mean of 1000 and a standard deviation of 200. All 500 values for this variable are converted into z-scores. What is now the value of the mean? What is now the value of the standard deviation?

A) mean: __________
B) standard deviation: __________

Question 47

A variable has a mean of 25 and a standard deviation of 5. Calculate the Z-scores for the first two cases and the raw scores for the last two cases.