reliability and validity

• reliable - gives the same results multiple times
• inter-rater reliability - two or more agreeing on same results
• valid - accurate results
• ecological validity - extent that it reflects real life
• standardising research - creating specific procedures which are followed each time
• operationalising - defining the variable
• internal validity - validity inside the research - extent that the results are due to iv
• external validity - does it apply outside the research
• mundane realism - is the tasks in the study true to real life
• population validity - is the sample used true to real life
• temporal validity - is the research true to life in todays world
• test retest - if you get the same results after testing twice
• face validity - if a test appears on the face of it to measure what it wants to measure - determined by eyeballing from an expert
• concurrent validity - a particular test is demonstrated when results obtained are close to another well established test
• ways of assessing validity - concurrent validity - face validity
• improving reliability questionnaires - rewrite q's - check for ambiguity - replace open q's with closed
• improving reliability interviews - same interviewer - ensure they are properly trained - structured interviews
• improving reliability observations - use behaviour categories
• improving reliability experiments - standardise procedures
• improving validity experiments - control group - standardise - double blind
• improving validity - questionnaires - complete lie scale first - test for social desirability bias
• improving validity observations - run covert observations - prevents demand characteristics - behaviour categories
• improving validity interviews - triangulation