Threats+to+validity-handout - Threats to Internal Validity...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Threats to Internal Validity * Internal Validity deals with whether or not the experimental treatments made a difference in this one, specific instance. Did the manipulation of the independent variable really cause what we see at the end of the study? Campbell & Stanley (1963) identified 7 key issues (and their interactions) that threaten the internal validity of a study: Selection The results of your study might be due to a pre-existing difference between groups in the study. For example, you want to study the effectiveness of new method for teaching math. You visit a local school and select two classes to try out your experiment with. In the first class you teach math using a traditional approach, but in the second class you teach using your new method. In the end you give both classes a test to measure how much they know, and those in the new method condition score better. BUT, what you did not know is that the class that got the new method was an advanced placement math course and the other class was a remedial math course. So, in the end you do not know whether it was a difference in teaching methods or some pre-existing difference in math ability that created the difference at the end. Statistical Regression to the Mean This in an effect that happens when dealing with people at the extremes of a population. Because our measures are not perfect, repeated testing will result in a move towards the average of the population. For example, you want to test a new method of teaching math to remedial students. In order to select your subjects you give them a pre-test and select the lowest 10% of scores. You then try to teach them using your new method, followed by a post-test. It turns out that the average of your sample goes up. BUT: Because of imperfections in measurement, we would expect this to happen regardless of what happened to them. So, we don’t know if the increase was due to the teaching method or if it was just a statistical artifact created by imperfect measurement. Maturation Biological and psychological process of subjects will change with the passage of time. Subjects will get hungry, fatigued, etc. during the course of an experiment, and these changes can create a confound that masks the true effect of an experimental manipulation. For example, at the start of your experimental you have a group of subjects tracing a pattern on a sheet of paper. The experimental condition receives loud bursts of noise in the background, while the control group is kept in a quiet room. After several
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 3

Threats+to+validity-handout - Threats to Internal Validity...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online