W2 - Science Flashcards
What are the 2 main considerations to the scientific research method?
- External validity (naturalness)
- Internal validity (control)
What is external validity (naturalness)?
It refers to the extent of the results of the study can generalise to other settings.
The more natural the study, the higher the external validity.
What is internal validity (control)?
It refers to the degree of which the study can accurately establish a causal relationship between independent variable and dependent variables.
Is a smaller / larger sample size better?
Depends on how representative the sample is.
If a small sample can represent the large population well, it is better than a large sample that is biased.
In general, large samples are more representative.
What is naturalistic observation?
Observe the behaviour of individuals in their natural environment.
High external validity.
Advantage of naturalistic observation?
Able to observe individuals natural behaviour without intervention.
3 disadvantages of naturalistic observation?
- We have no control, and hence low internal validity
- Behaviour that we want to study might not occur
- Observer bias
How to overcome observer bias?
Overcome by making the observer blind –> Record down individual behaviours without knowing what the hypotheses are / behaviour we want to study.
3 advantages of survey method?
*most used method
- Surveys can fit the questions we want to ask
- Low cost
- Quick and fast
4 disadvantages of survey method?
- Social desirability bias - Respondents tend to give socially approved responses
- Wording effects - Different results may be obtained depending on how the question is worded
- Response set tendency - Responses are given in a habitual way, and respondents don’t answer each question independently
- Low response rates
How to overcome social desirability bias?
- Anonymous respondents - Tend to produce more accurate responses
- Explain the importance of accurate responses
How to overcome wording effect?
Ask the same question in different ways to see if we get different responses + so that errors can cancel out each other
How to overcome response set tendency?
Reliability
How to overcome low response rates?
- Reciprocity - Give them something in return for completing the survey
- Explain the importance of the survey
- Let them know a survey is coming, so that they can anticipate it
What does reliability assess?
Reliability assess random errors
What tests are more reliable?
Longer tests of similar content are more reliable.
Theoretically half of them are going to be above the true score, while half will be under, which in turn cancels out to form a more reliable measurement.
How are reliability and validity related?
Reliability places an upper limit on validity.
Hence, if we cannot measure something reliably, we won’t have accurate measurements.
How should we assess the validity of a measure?
Face validity - On the surface, does it appear to be measuring the construct?
Content - Subject matter experts
Criterion-Related Validity
What is criterion-related validity?
Refers to whether the measure is related to, or predicts, a criterion variable (dependent variable)
What are the 2 properties of correlation?
Magnitude - strength of the relationship
Direction - positive/negative direction
