evaluation of research Flashcards
memorise and understand
reliability
the measures used in an experiment must be consistent within themselves and over time
validity
the extent to which the results of a study reflect what the measuring instrument says its measuring
validity vs reliability
validity: the extent to which something measures what it intends to measure
reliability: the extent to which a one would get the same result if the same measure were to be given to the same person again under the same circumstances
2 ways to measure reliability
internal consistency
test-retest reliability
internal consistancy
evidence is provided by the split-half method. wthin the test/measures the variable should be tested more than once and yield a similar result each time
example of internal consistency
in a questionnaire to assess someones happiness, their should be more than one question that related to happiness and on every question it should yield a similar result from the participant
test-retest reliability
how people perform on a test at one time is similar how they perform on it at a later time.
example of validity
in a questionnaire to asses someones personality. the personality type a person gets is the same as when they retake the test several weeks later
4 ways of measuring validity
face validity
construct validity
concurrent validity
predictive validity
face validity
does the test appear to be measuring what it claims? experts in the field judge whether the material in a measure (i.e. a survey) is appropriate
construct validity
do the test items relate to the aspects/constructs of the theory being tested?
concurrent validity
does the new test/scale that we are using as a measure yield the same results as a scale/test that we already know well?
predictive validity
the extent to which the measure can predict other attributes thought to be related to the constructs being tested
conclusion
a decision or judgement about the meaningfulness of a research result
what does a conclusion address
supports of a hypothesis
final analysis of results
sample size vs population
sources of error and their improvements