Deck 6: Measurement, Validity, and Reliability

Question 1

How do variables and constants differ?

Answer 1

Variables differ from case to case, but a constant is the same for our entire sample.

Question 2

What is operationalization?

Answer 2

The process of taking concepts and more defining them in a way that makes them measurable.

Question 3

Why do we code survey answers into numbers?

Answer 3

We do that, because the statistical software packages will not run tests on words.

Question 4

What is the difference between mutually exclusive and exhaustive?

Question 5

How do discrete and continuous variables differ?

Question 6

What are the two types of criterion validity?

Question 7

Which type of reliability becomes an issue only when we have multiple researchers working on a project together?

Question 8

What is the difference between interval-level data and ratio level data?

Question 9

______ is the process of defining concepts to make them measurable for research.

A) Sampling
B) Operationalization
C) Coding
D) Scaling

Question 10

A _______, such as victim harm, is a specifiable aspect of a concept.

A) Dimension
B) Code
C) Scale
D) Hypothesis

Question 11

Robert is interested in researching the impact of drug court participation on later recidivism. In this case, drug court participation is considered the ________________.

A) Control variable
B) Dependent variable
C) Independent variable
D) not enough information to tell

Question 12

Sam is interested in researching the factors associated with juvenile delinquency. We think that poor school performance will be associated with increased delinquent behavior, but we are also concerned about whether delinquent peer associations might also impact delinquency levels, so we want to take that into consideration as well. In this case, peer associations are considered ______.

A) A control variable
B) A dependent variable
C) An independent variable
D) not enough information to tell

Question 13

What is the process of assigning numerical values to answer options on our surveys?

A) Operationalization
B) Measurement
C) Analyzing
D) Coding

Question 14

A _______ survey item has only one possible answer option per respondent.

A) Open-ended
B) Coded
C) Mutually exclusive
D) Controlled

Question 15

A variable with a finite number of values is known as __________.

A) Mutually exclusive
B) Close-ended
C) Exhaustive
D) Discrete

Question 16

What level of measurement is a variable asking students to tell us how many college credits they have completed?

A) Nominal
B) Ordinal
C) Interval
D) Ratio

Question 17

What level of measurement is a variable asking each competitive eater what rank he/she got in the last chicken-wing eating competition?

A) Nominal
B) Ordinal
C) Interval
D) Ratio

Question 18

Adding up a number of responses to survey items gives us a(n) _________.

A) Exhaustive response set
B) Set of codes
C) Continuous variable
D) Index

Question 19

The extent to which our measure is capturing the full range and meaning of the concept in question is called __________.

A) Content validity
B) Concurrent validity
C) Inter-rater reliability
D) Criterion validity

Question 20

The extent to which our measure works well when compared to other items is called __________.

A) Concurrent validity
B) Content validity
C) Inter-rater reliability
D) Criterion validity

Question 21

The stability or consistency of a measurement is known as __________

A) Reliability
B) Criterion validity
C) Face validity
D) Content validity

Question 22

The highest levels of measurement, and the most informative, are ________ and ____________.

A) Nominal and Ordinal
B) Ordinal and Ratio
C) Nominal and Interval
D) Interval and Ratio

Question 23

If I distributed surveys to only a group of surgeons, the question of what job they have would be considered a variable.

Question 24

If we suspect that those who drop out of a court-recommended anger management program will be more likely to be arrested within a year, arrest is considered the dependent variable.

Question 25

In research we study how the dependent variable impacts the independent variable.

Question 26

It is easier to code close-ended survey questions than open-ended ones.

Question 27

As long as there is an answer option for every possible respondent, our survey item is mutually exclusive.

Question 28

A question "what is your hair color" would have ordinal-level response options.

Question 29

Likert scales are generally on a scale of 0 to 100.

Question 30

Validity refers to the stability or consistency of a measure.

Question 31

Reliability refers to the stability or consistency of a measure.

Question 32

Ensuring that all the researchers conducting data collection were trained in the same manner is an example of face validity.

Question 33

Predictive validity assessments, such as the SATs and GREs, aim to predict future performance.

Question 34

Once a measure is determined to be reliable, it is also always found to be valid.

Question 35

Discuss the four different levels of measurement and the types of information we can get from each.

Question 36

How can using multiple data collection methods help to test for and enhance validity and reliability. Provide at least one example of how we can do this.

Question 37

A __________ is a rather vague abstract that might cause different people to conjure different meanings.

Question 38

We need to ___________ concepts to make them measurable.

Question 39

A _________ is an attribute that does not vary at all from case to case.

Question 40

Having an answer option available on a close-ended survey question for all possible survey respondents means our item is __________ .

Question 41

Having an one, and only one, answer option available per respondent on a close-ended survey question means our item is __________ .

Question 42

________ level of measurement simply codes the answer options and provides us with the least amount of detail.

Question 43

_______ refers to the extent to which we are accurately measuring what we intend to measure.

Question 44

The _________ method involves including several items on a survey that measure the same construct and testing half of them against the other half for reliability.