Deck 15: Practical Issues in Data Collection and Coding

Question 1

When data entry has been completed, which of the following should be checked to ensure there have been no errors?

A) Maximum values
B) Minimum values
C) Missing values
D) All of the above

Answer 1

All of the above

Question 2

Dummy coding refers to:

A) Using the mean value for missing data
B) Using one and zero to refer to having a property or not having that property
C) If one does not know the actual value of a data point, the researcher adds what the researcher expects the value to be
D) Placing a zero in a data set for missing values

Answer 2

Using one and zero to refer to having a property or not having that property

Question 3

A value label:

A) Gives the mean value to a variable
B) Is the maximum value associated with a variable
C) A number associated with a level of a nominal variable
D) Is the same as a variable label

Answer 3

A number associated with a level of a nominal variable

Question 4

The researcher collected data on religious affiliation. The question had four categories of religion. The mean was calculated to be 2.34. How is the best way to interpret the mean?

A) Most people were Christian
B) Respondents were over two times more likely to be Christian
C) The variable is probably normally distributed
D) It has little meaning and should be ignored

Question 5

The researcher used exploratory factor analysis. The findings showed that the factor loadings of five questions were .72, -.70, .90, -.80, and .38. The researcher should look more closely and think of removing which of the questions.

A) The one that had a loading of .90
B) The one that had a loading of -.80
C) The one that had a loading of .38
D) All loadings are strong and all questions should be kept

Question 6

If a researcher summates the scales of a construct that has five items, this means that the researcher would:

A) Sum the high and low questions within a construct
B) Sum each of the items within a construct
C) Sum the standard deviations within the five items
D) Sum the five items and divide by the standard deviation

Question 7

Factor analysis is used to:

A) Determine internal consistency
B) Verify Cronbach's alpha
C) Reduce (simplify) the number of items measuring a construct
D) Determine results for a study with two or more independent variables and a normally distributed dependent variable.

Question 8

Factor loadings have values of:

A) Always more than 1
B) Usually more than 1, but sometimes slightly below 1
C) Between -1 and 1 including zero
D) The correlation squared

Question 9

Using Table 15.3 in the text what 12.7 percent of the variance could be attributed to items related to:

A) Low pleasure from doing mathematics
B) Difficulty with learning mathematics
C) Math achievement
D) Was the omitted item due to a low factor loading

Question 10

It is desirable to pilot test instruments prior to using them for data collection.

Question 11

Raw data does not need to be checked before it is entered into a computer program.

Question 12

When coding data it is best to use letters for nominal variables. For example, in statistical programs one should input an "M" for male and an "F" for female.

Question 13

"Mutually exclusive" means that the respondent can logically see that two categories can apply to answer the question.

Question 14

Coding rules should not apply consistently across individuals.

Question 15

If one assigns a "1" to a person who says they are religious and zero to a person who indicates they are not religious, this is an example of dummy coding.

Question 16

The researcher assigns a "1" to republican, "2" to democrat, and "3" to independent; this is an example of a variable label.

Question 17

A codebook is a printed record of the names and labels for each variable and value label.

Question 18

The researcher was interested in looking at religious affiliation. She had five different religions in her study. The mean religion was 2.6. This value has little or no meaning.

Question 19

Cronbach's alpha is used to rate internal validity.

Question 20

Exploratory factor analysis is used to rate internal consistency.

Question 21

Factor loadings less than .30 are considered low.

Question 22

Questions with negative factor loadings should be eliminated from an instrument.

Question 23

Why is it important to compute basic descriptive statistics?

Question 24

What is the most important reason for reducing the number of measures in a survey?

Question 25

When is it appropriate to compute an overall alpha?