Reliability and Validity. Flashcards
In order for an experiment to be reliable we tend not to trust one single _______. What do we do?
Measure.
We tend to repeat the experiment several times then take an average.
If the experiment only has 2 trials, what can we say about the reliability?
That the reliability is low.
How would we increase/improve the reliability?
Take a series of trials.
Do 20 trials instead of 2.
Name this:
- How dependable is your measure?
- Is it consistent over time?
- Would you get the same measure if you do it again?
The reliability of our measurement.
Apart from increasing the number of trials, how else can we improve reliability?
By controlling extraneous variables.
If we don’t control extraneous variables, it becomes ____________.
Confounding.
So name the 2 main ways to Improve Reliability.
- Increase the numbers of trials (and find the average).
2. Control Extraneous Variables.
To ensure the measure of something is reliable, take several ______ of the measure for each person, under ________ conditions and then find the _______ measure.
Readings. Controlled, Average.
More trials = _______ reliability.
Greater.
How do we ensure the reliability of measurements is high?
- Take repeated readings of variable and use average value
- Control Extraneous Variables.
What is meant by Validity?
To what extent is your measure and appropriate measure of what you are trying to measure?
Today, ______ is still an area of considerable debate.
Validity.
Validity is about how ______ your measure is.
Appropriate.
Name this.
- Are the measures appropriate?
- Are they measuring what they think they are measuring?
Validity.
Validity has 2 main distinctions. Name these.
- Internal Validity
- External Validity
What is validity very relevant to?
Design considerations of an experiment.
Why is validity important?
The experiment needs to be a proper, fair test that is representative of real life.
The experiment must be _____ for the data to be useful.
Valid.
What does Internal Validity refer to?
- Was there a real effect? (eg. not fluke, no mistakes, right test used etc.)
- Were the changes seen in the DV solely due to manipulation of the IV? (were all extraneous variables controlled etc.)
Internal Validity measures if there was any ____ to ________.
Threats, Validity.
What does External Validity mean?
How generalizable are the results…
- To Different Populations
- To Different Locations
- To Different Times
- To Different Measurements?
In order to have external validity for different populations, what must we have?
A varied sample. (different ages, genders etc.)
Explain what is meant by Ecological Validity?
Are our results valid in different locations eg. do the same study in 2 different places.
If you have the same results/affect from your experiment in two different places, what can be said?
Your experiment has a high ecological validity.
Explain what we mean in terms external validity, by the results being the same at different times?
This means- would the results from an experiment in the past (1970) still be relevant/valid today?
_______ validity takes into account if we can use a different ________ and still get the same results.
External, Measurements.
We can use different ______ to measure the same thing eg. aggression.
Techniques/Measurements
