Exam 2 Flashcards
What are the three steps to conceptualizing a variable?
- Define the concept
- Think of different dimensions of the concept
- Think of indicators of each dimension
Can you conceptualize a variable such as height?
No, conceptualization is just for abstract variables like generosity, marital happiness, “race” in census date, “age” in Chinese society
A(n) _____ is an observation that reflects a variable, such as “have you ever been slapped by a parent?”
An indicator is an observation that reflects the variable, it is the third step in conceptualizing a variable
A(n) ______ is a specifiable aspect of a concept.
A dimension is a specifiable aspect of a concept. Step two in conceptualizing. For example, Two dimensions of “child abuse” would be physical abuse and psychological abuse.
What is operationalization?
Operationalization is the process of developing specific research procedures in order to make empirical observations of a concept.
List the four levels of measurement
Nominal
Ordinal
Interval
Ratio
What are nominal measures?
Nominal measures measure variables that only have categories, such as sex, race, religion, college major, hair color, city, birthplace.
What is the level of measurement if the topic is college major?
Nominal Measures
What are ordinal measures?
Ordinal measures measure variables with attributes we can logically rank order, such as satisfaction, religiosity, social class, conservatism
What is the level of measurement if the topic is generosity?
Ordinal Measures
What are interval measures?
Interval measures are used when the distance between attributes means something. We can calculate how much different. Numbers are used. Examples are temperature, height, weight.
What is the level of measurement if you are studying weight?
Interval Measures
What are ratio measures?
Ratio measures are used when the attributes of a variable are based on a True Zero point, when zero means zero. Examples would be number of times married, number of siblings, number of asian friends
If you are wanting to know how many times a person has been divorced, what level of measurement would you use?
Ratio Measures
True or false: with data, we can always change from a Higher level variable (Ratio) to a Lower level (Nominal)
True
What is the highest level of measurement?
Ratio
What is the lowest level of measurement?
Nominal
Rank the levels of measurement from highest to lowest.
Ratio
Interval
Ordinal
Nominal
Rank the levels of measurement from lowest to highest.
Nominal
Ordinal
Interval
Ratio
What measure would you use to measure “sex” (gender)?
Nominal measures
What measure would you use to measure race?
Nominal measures
What are the three criteria of measurement quality?
- Precision
- Reliability
- Validity
What are the two most important criteria in measurement quality?
Reliability and Validity
If an experimenter repeats an experiment 3 times and gets conflicting results each time, their measurements are not what?
Reliable, to create reliable measures, a researcher can test-retest, split-half, or use already established measures
Which is the most precise? Which is the least precise when asking about the phenomena of underage drinking?
A. What is your age? ____ (most precise)
B. Are you underage [Yes] [No] (least precise)
C. What is your age group? [10-19], [20-29] etc.
B would be the most precise
Which is the most precise? Which is the least precise when asking about the phenomena of drinking in 5 different age groups?
A. What is your age? ____ (most precise)
B. Are you underage [Yes] [No] (least precise)
C. What is your age group? [10-19], [20-29] etc.
C would be the most precise
How can a researcher create reliable measures?
Using:
- Test-Retest
- Split-Half
- Established Measures
Explain the Test-Retest method
Repeat the measures several time to see if we get the same results each time, you must use different random samples each time
Explain the Split-Half method
Split-half method: Used only when we have several indicators. Randomly split indicators into two sets, pick 3/6 and put it into set one, pick other 3/6 and put into set two and then test. Give the two sets of the questionnaire with two random samples, calculate the average, they should be similar or close to another, result one and result two should be the same as if you put the two groups together
CES-D score
Self-esteem scale
Zick Rubin’s measurement of “romantic love”
Are all examples of what?
Using established measures
The extent to which an empirical measurement adequately reflects the real meaning of the variable under consideration is called what?
Validity
What is the procedure in Split-Half method?
i. Test the indicators of a variable, get the results
ii. Randomly split the indicators into two sets
iii. Measure both sets of indicators
iv. The results should be the same as those in (1)
What are the five general principles in research ethics?
Requires openness, honesty, and truth
Forbid the falsification of results; condemn plagiarism
Informed consent
Do not harm the subjects (both physically and psychologically)
Protect the anonymity of the subjects
Explain the importance of debriefing?
The importance of debriefing – must disclose the real purpose and why you lied in the first place. Debriefing can compensate what we lied about. Interview subjects after the experiment. To discover any problems that were generated during the study. Also, to see if these problems can be corrected
Which measure would you use to measure age?
A ratio measure because it has a true zero point.
When variables are abstract, you need ______ questions to answer the question well
Many
Give two reasons why conceptualization is important.
- Most of what we study in social science are not concrete enough for use to observe directly, we need to define the problem and specify your variables
- The same things could mean differently in different time periods, in different sub-groups, and in different cultures
What is the first most important thing you must do before measuring a variable?
When measuring a variable, we first need to identify what level of measurement are the attributes of the variable
When your research question has several indicators, which measurement method would be best?
Split-Half
What is the criteria for a topic to be appropriate enough to conduct an experiment?
- Projects with limited and well-defined concepts
- Projects that are exploratory rather than descriptive
- Studies of small-group interaction
IV is the ______ DV is the ______
IV is the CAUSE
DV is the EFFECT
DV is the ______ IV is the _______
DV is the EFFECT
IV is the CAUSE
Experiments involve taking an ______ then observing the __________ of that action.
Experiments involve taking an action then observing the consequences of that ction.
In both pre-test and post-test, the subjects are measured in terms of the what?
Dependent variable
Which group is given a stimulus: experimental or control?
Experimental group is given a stimilus
Post-testing occurs after what?
Post-testing occurs after the stimulus is administered.
What is the purpose of a double-blind experiment?
To reduce bias
What is a double blind experiment?
Neither the subjects nor the experimenters know which group is the control group.
Most statistics assume which method of assigning subjects to E/C groups?
Randomization, most statistics assume randomization instead of matching
Which method of assigning subjects to E/C groups works best for small groups?
Matching works best for small groups
How does a researcher assign subjects to E+C groups?
Randomization or matching
What are the three pre-experimental designs and what is the common problem in all?
One-shot case study
One-group pre-test / post-test design
Static-group comparison
Common problem: validity
What are the features and problems of a one-shot case study?
One shot case study:
features: only one group, (E), only post-test
problems: we cannot be sure of the factors effecting results
What are the features and problems of a one group pre-test post-test design?
One group pre-test post-test design:
features: only one group (E), both pre and post test
problems: we cannot be sure of the factors effecting results
What are the features and problems of a static-group comparison?
Static-group comparison:
features: two groups, only post-test
problem: we cannot be sure of the factors effecting results
What is the difference between internal and external validity?
Internal validity is when the changes in experiments are not caused by the stimulus, but by other factors.
External validity is the problem of being unable to generalize results to the “real world”
What are the four sources of internal invalidity?
- Testing
- History
- Instrumentation
- Maturation
Three researchers all conceptualize “marital happiness” differently, which leads to different results, making the experiments invalid. What type of validity is this an example of?
Invalidity due to instrumentation, issues of conceptualization and operationalization
A researcher is conducting a test on how children pay attention in class. She conducts her first set of tests and then her second set of tests a year later. Her results are invalid, why?
Invalidity due to maturation, people are always growing and changing which can affect behavior
A researcher is conducting an experiment on people’s perceptions of Asians and while waiting for their experiment, there is breaking news of an asian man robbing a bank. The research would be invalid, why?
Invalidity due to history, historical events may occur during an experiment, changing subjects’ behaviors and attitudes
A subject feels one way during pre-test and changes her answers on post-test on purpose. The results would be invalid, why?
Invalidity due to Testing, the process of pre and post testing influences people’s behavior.
What does HSIRB stand for?
Human Subjects Institutional Review Board
A subject has an extremely high test score when you are pre-testing to see how study sessions affect grades. The results might be invalid, why?
Invalidity due to Statistical Regression, when subjects are in extreme positions to start with, it is difficult to assure the effect of the stimulus.
You select groups in a haphazard way for an experiment. Your results would be invalid, why?
Invalidity due to Selection Bias, this happens when subjects are not appropriately assigned to E and C groups.
Three of your subjects ditch your experiment. Your results would be invalid, why?
Invalidity due to Experimental Mortality, some subjects drop out of the project before the experiment is over.
Your subject feels deprived and it causes them to give up the experiment, your results would be invalid why?
Invalidity due to Demoralization; feelings of deprivation may cause the subjects’ giving up.
What is the strength of experimental methods?
We can isolate and control the stimulus
Experiments can be replicated
What is the major weakness of experimental method?
Artificiality and generalizability
The Soloman four-group design includes what two testing designs?
Static design and classical design
Would it be appropriate to use an experiment to study a large group of students?
No, experiments work best with small groups
