validity Flashcards
what is validity
refers to whether a measure is actually measuring what is intends to be measuring
what is face validity
this is a way of measuring validity on the face of it (at first sight)
e.g. questions of a questionaire
what is concurrent validity
comparing new scores gained and comparing them to old established tests, if correlation co efficient is greater than 0.8 then it has high internal validity
what is internal validity
when the effects observed in a study are due to the manipulation of the independent variable and no other one
how can concurrent validity be improved
remove any questions that seem irrelevant or ambiguous
how can face validity be improved
getting an expert in the field to examine all the questions and remove or reword questions that are ambiguous
what is external validity
the extent to which results can be generalised to other settings, populations, and time periods
what is ecological validity
generalise findings to other settings like every day life
what is population validity
extent to which sample can be generalised to the target population
what is temporal validity
findings from a certain period can be generalised to other eras and time periods
How to improve internal validity
Reduce investigator effects - double blind procedure, not allow participant or researcher from knowing aim of the study
Reduce demand characteristics- hide the purpose of the study from participants , covert observation single blind
Use lab setting to reduce confounding variables
How is external validity assessed
Meta analysis, comparing findings from diff research w same hypothesis, if consistent then high ecological validity e.g. van izjendoorn cultural finding to strange situation
Consider the setting, natural setting has higher ecological validity
Consider the task does it have mundane realism
Asses whether participants were behaving as natural as possible, ensure demand characteristics are kept to a minimum
Improving external validity
Reduce demand characteristics Double blind or single blind
Improve ecological validity - covert observation and field experiment would be better