Threats to Internal Validity

when extraneous variables make it impossible to determine that the effects on the DV were a result of your IVs

variables other than manipulated variables (IV) with potential effects on behavior

the test does not measure or manipulate the construct it is supposed to measure (construct validity) or does not cover the construct conpletely enough (content validity)

either has no comparison/control group or has a poor one and puts people in groups without random assignment (quasi)

must not have either ceiling or floor effect or an imprecise measures

any changes due to the differential selection procedure used in placing subjects in various groups resulting in pre existing differences between groups including volunteering

any changes in subject performance due to the fact that subject participated in more than one experiment (is your IV within subject)

any change that occurs in the criteria used for measuring DV (observer or equipment)

Any event other than IV that occur between pre and post measure of DV problem any time multiple measures are taken

any of the many conditions internal to the individual (subject) that change as a function of passage of time; may be biological or psychological

any change due to a differential subject loss from the various comparison groups

Any change that can be attributed to subject motives biases or explanations

any change in the performance that can be attributed to the experimenter