validity Flashcards
what does validity mean?
-refers to if the test in question measures precisely what it aims to measure
-data must be accurate and present some truth compared to others outside the study
what does internal validity mean?
-if researcher has measure what they intended to measure
-effects observed due to manipulation of IV rather than extraneous and confounding variables
how do demand characteristics affect validity ?
-lower validity
-dont act naturally
-act in a way they think is expected
how do EVs affect validity?
-may have unwanted factors impacting results such as the Ps mood
what are confounding varibales
-factors that definitely impact results such as how much practice a participant got before a maths test
what is meant by external validity?
-factors outside of an investigation generalising to other settings, populations and eras
whats meant by ecological validity
-findings generalised from one setting to others (everyday life)
what is meant by temporal validity
-whether findings are true overtime
-free from historical bias
what is meant by population validity?
-if findings from the studies Ps can be generalised to target population
whats meant by mundane realism
-refers to how an experiment mirrors the real world
-(simulated task env is realistic = similar to task env in real world)
-so more ecologically valid
what can affect internal validity
-Evs, confounding variables
-e.g gender, demand characterstics,mood on the day,age, time of day, appearance
what can affect external validity
-population val
-temporal val
-eco val
-mundane realism
whats one way of assessing validity? (face validity)
-whether a test,scale or measure appears on the face of it to measure what its meant to measure
-looking at measuring instrument or asking experts to check it to see if it makes sense.
what is another way of assessing validity?(concurrent validity)
-when results obtained in a new test are close too or match those obtained on another recognized and well established test
what can we say about correlational data
-if co efficient is 0.80 or more than the new test, it has high concurrent validity
one way of improving validity(traingualtion)
-compare results of 2 or more studies of the same thing to see if they are in agreement - demonstrates the validity of the individuals results.
(use diff methods)
another way of improving validity (interpretive validity)
-the match between the researchers judgement and the actual Ps perspective (simply if researcher interprets the behavior correctly.
why do in depth interviews have low validity?
-lower interpretive validity
-alot to analyse so may misunderstand
solution to low valdiity in indpeth interviews
-two researchers
-reduce bais
-more training
-triangualtion
how can u improve validity in experiments
-use control grp
-good comaprisons could be cocnluded if change in DV was due to Iv
-standardistaion also reduces particpant reactivity and investiagtor bias
how else can u improve validity in experiements
-use of single blind and double blind
-reduce demand characterstics
-double blind means that it reduces DC and invetsgator effects
how do u improve validity in questionnaires
-incorporate a lie scale
-to assess the consistency of a respondents response and control effects of social desirbaility
-validity enhanced by ensuring data is anonymous
how to improve validity in observations
-covert observations have higher eco validity as researcher is undetected
-more likely to present natural behaviour
-behaviour categories that are too broad could badly impact the validity
