There are many sources of internal invalidity. Three are listed below. Identify the source of invalidity in the following table and match it to the descriptions.

Description	Instrumentation	Selection Bias	Compensation
The teachers at the preschool feel sorry for the kids who are not in the experimental reading program, so they read extra books to them and help them with phonemes. This confounds the results.
The experimenter used a measure for the posttest that differed from the measure he had used for the pretest, making it difficult to interpret the results.
The researcher decides to assign the students in her class to the intervention group (watching a movie on poverty) and then chooses a different sociology class to be the control. She finds that her class changes their spending behavior, whereas the control class does not, but she then is not sure if it was actually the intervention or something else that caused the change.

The classical experiment, if coupled with proper subject selection and assignment, addresses most problems of internal invalidity. Does this experimental design guard against the problem of instrumentation? If so, how? If not, why not?

Yes, good experimental design avoids this problem by using the same instrument for both the pretest and the posttest.

No, any effect of instrumentation is going to affect the experimental group more than the control group.

Yes, the group affected by instrumentation just becomes the control group.

No, it is better not to use tests.

You submit a paper detailing an experimental study you have just done on parental responses to night terrors in children. The study was conducted shortly after 9/11. Your critics have argued that your study results are due to the specific time that you did your study.

This is an example of (a threat to external validity / a threat to internal validity / compensation / testing).