Unit 3 - Chapter 7 - Validity Flashcards
What is the ultimate goal of research?
To develop a theory that explains the relationships found among variables
What is validity?
An indication of accuracy in terms of the extent to which a research conclusion corresponds with reality.
What 4 types of validity must be considered in designing and evaluating research?
- Internal validity
- Construct validity
- External validity
- Statistical validity
What is internal validity?
- Internal validity is the most fundamental type of validity as it centres the logic of the relationship between the independent and dependent variable
- An experiment has internal validity if there is reason to believe that there is a cause-effect relationship between the independent and dependent variable
- Did the independent variable cause the dependent variable to change?
What is “confounding”?
- Confounding is a type of error
- Confounding happens when the effects of two variables cannot be separated, resulting in confusion over which variable changed
- Example —} Group A is given a test to measure the effect of loud music on typing speed on Monday. Group B is given a test to measure the effect of soft music on typing speed on Tuesday. If there is a difference between the results of Groups A and B, is it due to the different levels of noise, or that the test happened on different days? The two variables, levels of noise and different days of the week are confounded
What is a subject variable?
A difference between subjects that cannot be controlled, but can only be selected. Example is gender … people cannot be assigned to one sex or another, but must be chosen from preexisting groups. This is the same with race and IQ.
LOOK THIS UP AND COME UP WITH MORE EXAMPLES … NOT CLEAR
What is construct validity?
- The extent to which the results support the theory behind the research.
- Does the test measure what it is supposed to measure?
- If you want to measure intelligence, you don’t survey people’s shoe size. A survey that collects shoe sizes can’t tell you anything about intelligence. Would another theory come up with the same results?
Construct validity example
Your hypothesis is that anxiety is conducive to learning. You select participants that bite their fingernails and then test to see how fast they can learn to write by holding a pencil in their toes.
PAGE 175 - LOOK FOR ANOTHER EXAMPLE - THIS IS CONFUSING
What is a manipulation check?
- Aspects of an experiment designed to make certain that variables have changed in the way that was intended
LOOK THIS ONE UP AS WELL
What is external validity?
- How well the findings of an experiment generalize to other situations or populations
- If I want to repeat the experiment, will it produce the same results for me that it produces for the researchers?
What is ecological validity?
- The extent to which an experimental situation mimics a real world situation.
- Can results obtained in a laboratory setting be reproduced in a natural setting, for example, your home?
- This type of validity is related to external validity, but is not the same
- The example given is of an experiment conducted in 1949 that measured people’s level of discomfort with saying “bad” words, such as “bitch” or “rape”. If the experiment was conducted today, it is likely that people would not be as uncomfortable saying the words today as they were in 1949.
What is statistical validity?
- It is similar to internal validity
- Was the observed relationship between the independent variable and the dependent variable a true cause-effect relationship, or was the result accidental, and caused by pure chance?
- To establish statistical validity, appropriate sampling and measurement techniques must be used
Examples of experiments and their variables
Researchers want to determine if different levels of light influence people’s ability to learn a list of words.
- A group of participants is randomly assigned to Group A or Group B
- Group A is the control group - they will be given a list of words to learn in ordinary light.
- Group B is the experimental group - they will be given a list of words to learn in a semi-darkened room.
- Other variables are kept constant for both groups - same room, same day, same chairs, same number of words on the list, etc. You’ve controlled for as much as you possibly can to the point where the only difference between the two groups is the amount of light in the room.
- The independent variable (what are you changing?) - the level of light
- The dependent variable (what is it going to influence?) - the number of words that the participants learn
- If this experiment has internal validity, there will be a cause-effect relationship between the level of light and the number of words learned. For example, when the light is quite low (measured in candles)
What are the problems with the Festinger expreiment done by Michael and Natsumi?
Michael and Natsumi didn’t control a number of variables.
- Their first mistake was that they didn’t run the $50 and the $1 group concurrently. When they couldn’t get people to stay and talk to the next group of students, they split the experiment in half. First they did the $50 group and completed that and then they started on the $1 group. Both groups should have run at the same time. Since they were different circumstances and settings, the two groups did not run under the same conditions
- Mistake #2 - The attitude survey that students were asked to complete wasn’t administered correctly, so the answers are useless. By tacking the survey onto the final exam, some students may not have recorded their true feelings and instead indicated what they think the professor wanted to hear. Also, not everyone completed the survey, so the data set was incomplete. The two people who they tracked down and got them to complete the survey, did not share the same circumstances as those who did the surveys
