The extent of performing the relationship between the variables elucidating the large population is referred to as generalization and the validity of generalized scientific inferences is referred to as external validity. The cause effect relationship, if able to be generalized from specific inferences of particular population or group of study to other population settings then, it is assumed to be externally valid. The external validity is highest in the case of casual inferences as they can be applied in diverse situations across space and time. The major loss associated with the external validity is that the qualitative relationship between cause and the effect is made from the small sample of limited geography, which may not be applied as it to other demographic locations.
2. Threats Associated With External Validity
The facts and figures of generalization may be a major threat to external validity as they might not always be correct in its explanation. When certain results associated with specific demography, at definite time and the place and generalization is made that the inference of that study is applicable to other demographic location, time and some other group or population then the threats would be of three types;
The argument of the critics may be that the time of study was peculiar or the people or population or the sample taken for the study was not suitable to be generalized for other set of populations or that the area under study was so remote that it couldn’t represent the other locations. The generalization may be wrong when the independent factors or cause of time, place and people are strongly related to other factors or variables and in such situation the external validity threats to the independent variables. The following threats cover almost all the possible threats faced by the external validity;
2.1. Capacity-Treatment-Interaction is observed when the independent variable interacts with other features. For instance, the therapies for quitting smoking habit would be applicable to the volunteers but not exact path would be desirable for non- volunteers. [linkunit]
2.2. Circumstances like treatment administration, timing, scope, location; treatment conditions etc that are specific to certain situations decrease the generalization.
2.3. Effects of Pre-Test are seen on generalization activity in a situation when the pre-test explains the actual relationship between the cause and effects.
2.4. Effects of Post-Test affects generality when the inference of cause-effect relationship could be made with the post-test only.
2.5. Reactivity can also limit the generalization of inferences, if the effects are produced as the effect of original situation and as a result are inapplicable on the other situations.
3. How External Validity Could be Improved
There are certain ways to improve the phenomenon of external validity;
• If the sample model is followed up for making the general selection that is random selection reducing the chances of biased or no-random sampling in a certain population.
• If the participation of respondents is maintained with the minimum dropout rate.
• By making the application of proximity theory to the maximum level for accounting the similarities in groups of people, their place, time and even methodologies being followed up.
• By replication of study with several locations, different groups of people that are randomly selected and in the different time periods; this would reduce the chances of critics to prove the study wrong as it wou8ld be gone through a self check.
• Trochim, William M. The Research Methods Knowledge Base, 2nd Edition.
• Mitchell, M. & Jolley, J. (2001). Research Design Explained (4th Ed) New York:Harcourt.
•Shadish, W., Cook, T., & Campbell, D. (2002). Experimental and Quasi-Experimental Designs for Generalized Causal InferenceBoston:Houghton Mifflin.
• Brewer, M. (2000). Research Design and Issues of Validity. In Reis, H. & Judd, C. (eds) Handbook of Research Methods in Social and Personality Psychology. Cambridge:Cambridge University Press.