Experimental designs, reliability and validity

Solution Preview

EXPERIMENTAL DESIGNS

VALIDITY
Validity means 'does the experiment measure what it is supposed to measure?' There are several different types of validity to consider when conducting an experiment.

FACE-CONTENT VALIDITY - this examines the content of the test to see if it looks like it measures what it is supposed to measure.

CONCURRENT VALIDITY - this involves comparing a new method or test with an already well established one that claims to measure the same variables. If the test is valid then a high positive correlation should occur.

CONSTRUCT VALIDITY - if a test or method has construct validity it means that the results can be used to support the underlying theoretical constructs concerning the variable that it is supposed to be measured.

PREDICTIVE VALIDITY - this refers to if the test will predict future performance indicated by the results.

ECOLOGICAL VALIDITY - if a method has ecological validity it means that the results/behaviour is representative of a naturally occurring behaviour. This is the main critique of laboratory experiments, lab experiments are often criticised for being 'artificial' and producing behaviours/results that would not be found in the natural environment.

Lab experiments and operationalized variables may lack ecological validity; the only method of control is to use a non-laboratory environment (or 'field') experiment instead. This however is not always possible.

RELIABILITY

INTERNAL RELIABILITY - this refers to how often a method measures within itself. If methods were not standardized they would give distorted final scores, e.g. internal reliability would be lacking if;

A ruler consisted of variable centimeters
If an IQ test had half easy and half difficult questions (most would ...