A QUICK AND DIRTY GUIDE TO VALIDITY:

Here's some of the basics to get you up to speed with validity. For a more detailed look at the topic click here A more in depth look at validity.

VALIDITY:

The best approximation to the truth of a given proposition, inference or conclusion.

GOOD THINGS TO REMEMBER:

Measures , samples, and designs DON'T have validity -- they may lead to valid conclusions, but in and of themselves they are not valid
There are four types of validity and they build on each other
Validity is something to be argued, not proven
Validity can be thought of as a set of standards by which you judge things

FOUR SUBDIVISIONS OF VALIDITY:

CONCLUSION VALIDITY: This type of validity asks the question -- Is there a relationship between two variables in a given study? In Susie's project we would ask: Is there a relationship between participation in parenting classes and parent involvement ?
INTERNAL VALIDITY: Internal validity refers to the question -- Assuming that there is a relationship between variables, is the relationship a causal one? This would lead us to ask: Do parenting classes cause parent involvement?
CONSTRUCT VALIDITY: Construct validity looks at the question -- Assuming there is a relationship in this study, and it is causal, did our program reflect our idea of the construct of program and did our measure reflect the idea of our construct? Or, in our example, Did the parenting classes, or treatment, reflect our construct of parenting classes and did what we measured as "parent involvement" reflect our construct of "parent involvement"?
EXTERNAL VALIDITY: External validity is concerned with GENERALIZIBILITY and asks -- Assuming that there is a relationship, and it is causal, and our constructs are adequately represented, can we generalize to other situations and people? Here, we would ask: Would parenting classes increase parent involvement in other classrooms with other parents?

THREATS TO VALIDITY:

Threats to validity raise the what about? and what if? questions that people often ask in terms of a research study. They are also related to the "how come you didn't think of this?" questions that committee members ask at a defense. More generally, threats to validity are the possible reasons that the inferences made from a research project may be wrong. These threats come in many shapes and sizes; threats to construct validity are the focus of the previous page (click here to return to the previous page: Threats to Construct Validity.