A threat to validity is any factor, other than the one(s) the researcher intends to study, that can provide an alternative, plausible explanation for the observed results (i.e. the value of the Dependent Variable). These factors are sometimes called Confounding Factors. The more threats to validity that are present, the less confidence a researcher can have that the Independent Variable(s) actually caused the observed results. During the process of study design, a researcher should try to identify as many potential threats to validity and reduce or eliminate them as they design the study.
While the goal is to reduce the presence of validity threats, even the most carefully designed study will still have some threats to validity. Often times by eliminating or reducing one threat, the researcher introduces a different threat. The study design process involves making deliberate decisions about which of the competing validity threats is less problematic. The write-up of a study should always describe the threats to validly, both those that the study design was able to address and those that were unaddressed and still remain.
Researchers identify different types of validity threats. There are at least the following three (some researchers identify more):
Section 1 of a course uses a new approach for learning a new concept. Section 2 uses the current approach for learning the new concept. The students in Section 1 score 25% better on the corresponding exam questions than students in Section 2.
Does this result show that the pedagogical approach is beneficial? Take a moment to think about the potential validity threats in this study design.
.
.
.
.
.
.
.
.
.
.
.
.
.
Keep scrolling :-)
.
.
.
.
.
.
.
.
.
Almost there!
.
.
.
.
.
.
.
.