2.4.4 Reliability and Validity Flashcards
what is another word for reliability?
consistency
what does validity refer to?
legitimacy and accuracy of results
a study lacking reliability will also lack….
validity
what does internal reliability refer to?
how consistently the method measures within itself
in controlled research what needs to be standardised?
instructions, procedures, apparatus
in observations you can check for inter-rater reliability, what does this involve?
2 or more observers watching the same behaviour and comparing results
in questionnaires you need to make sure questions are unambiguous what does this mean?
interpreted in the same way by everyone
what are scales such as the likert scale? (strongly agree to disagree)
subjective, dependant upon interpretations
(internal reliability) when would final results be distorted, give an example:
when methods of measurement aren’t standardised
e.g. ruler consisting of variable centimetres
internal reliability can be checked by the split half method, what is this?
correlating results of half the items with the other half (odd numbers with evens on test) and gaining a high positive correlation coefficient
what does external reliability refer to? what should methods of measurement give?
how consistently a method measures over time when repeated
similar scores when repeated on same participants in similar conditions
give an example of when external reliability would be lacking:
if a ruler measured an object at different lengths each time it was used
external reliability can be checked by the test re-test method, what is this?
correlating results of the test conducted on one occasion with results of the test conducted on later occasion and gaining a high positive correlation coefficient
what does validity of a method of measurement refer to?
whether it measures what it is supposed to measure
what does internal validity refer to?
whether results were really due to the variables being measured
what does external validity refer to?
whether results can be generalised if conducted in different environments with different participants
what is another term for face validity?
content validity
what does face/content validity involve?
examining the content of the test to see if it looks like it measures what it is supposed to measure
in checking face validity what don’t you need?
a professional
in checking content validity who do you ask?
a professional/expert
e.g. in checking face/content validity of a intelligence test what are you looking for?
seeing if it actually measures intelligence or just general knowledge
what does concurrent validity involve?
comparing a new method/test with already established one claiming to measure the same variables and gaining a high positive correlation between the results of the two
what are 2 types of internal validity?
face/content validity
construct validity
what are 3 types of external validity?
concurrent validity
predictive validity
ecological validity
what is construct validity?
whether the test or method can be used to support the underlying theory concerning the variable supposed to be measured
why is construct validity needed?
to ensure you are really testing your hypothesis
construct validity - if theory suggests offspring of intelligent parents raised in a stimulating environment should be intelligent, what should be used to confirm it?
an IQ test
what does predictive validity refer to?
whether the test will predict future performance indicated by its results
what does ecological validity refer to?
whether the method measures behaviour that’s representative of naturally occurring behaviour
ecological validity is also the extent to which findings can be?
generalised
what needs to be considered if the findings are said to be generalizable?
if the environment is generalizable/natural
if the participants are representative of the target population
testing for ecological validity involves checking the task for mundane realism, what does this mean?
if the task would be done in the real world
what extraneous variables can affect the internal validity of a piece of research?
demand characteristics social desirability researcher bias individual differences (independent measures) order effects (repeated measures)
what are demand characteristics, why is it an issue for validity and how can this issue be overcome?
people change behaviour due to being observed
then measuring fake behaviour
single blind - not telling them the aim
what is social desirability, why is it an issue for validity and how can this issue be overcome?
changing your answer to look better
then measuring fake answers
overcome by ensuring anonymity
what is researcher bias, why is it an issue for validity and how can this issue be overcome?
researcher is expectant of an outcome so they measure the outcome rather than reality
overcome by having multiple researchers or double blind
what are individual differences, why is it an issue for validity and how can this issue be overcome?
when each participant is in one condition, individual differences between groups
overcome by repeated measures or matched pairs
what are order effects, why is it an issue for validity and how can this issue be overcome?
practice or fatigue effects in repeated measures
overcome by counterbalancing or matched pairs
