External validity
Experimental research is a rigorous process that is significantly used in the testing of the cause and the effects of relationships of variables. The experimental research involves a plethora of processes that include testing research hypothesis, measuring and description of variables, designing and interpretation of research. The findings obtained in the experiment help in the drawing of conclusion, and are categorized into internal validity (the cogent conclusion) and external validity (the generalized inference). The external validity of research findings is an indispensable facet of any psychological research that should, by all means, be tested through replication before any generalized conclusion is drawn.
Discussion
External Validity: Definition
External validity (trivially called generalizability) can be defined as the ability of research findings to be generalized across population, settings and time. Tayie (2005) accentuates that analysis of variables, instrumentation and subject selection can sternly affect the generalizability of findings. In relation to this, the utmost rudimentary issue in external validity is generalization, which, in its basic form, means the application of the conclusion of a study to other somewhat related contexts.
Generalization
Generalization across participants is one of the major types of generalization whose overarching idea lies in the relationship that coexists among conceptual variables. Similarly, applying results from an experiment to another group of participants- a group that is dissimilar and more encompassing than those used in the initial experiment-and holding assumptions that the results obtained in a population will generalize to a different population used in the experiment (Stangor, 2011). In the same light, generalization across setting, another type of generalization, elaborates application of findings from an experiment to a situation or condition that dissent from that of the original experiment. Moreover, it refines the limitation of results of an experiment to some vital issues, which encompasses settings, experimenters and measured variables used in research, attributed to the exclusive nature of any experiment (Stangor, 2011).
Replication: Types of Replication
In validation of experiments and empirical tests, Tayie (2005) asserts that replication, a process of repeating a research, can be used. One of the basic types of replication is exact replication, which entails all aspects, similar or close to the original studies, and it contains necessities that are of higher degree of standardization. The main purpose of this replication is to counter check whether similar or different results can be obtained involving the same experiment (Stangor, 2011). For instance, a study can be structured to illustrate how external motivation levels influences employee performance in a company. An exact replication of the study can be undertaken to institute the validity of the initial findings.
Conceptual replication, a study based on original research but uses different operational procedures to better appraise the true relationship between the involved variables, both independent and dependent (Mitchell & Jolley, 2010). For instance, a study can be designed to institute how the performance of high-school students is affected by punishment. The researcher can use positive punishment as the mode of punishment. The same study can be conceptually replicated by using a negative punishment.
Constructive replication, on the other hand, involves the addition of new factors during a complete start over of the study in a bid to test the same hypothesis similar to first experiment (Tayie, 2005). For instance, children are in most cases involved in wrong deeds while in group. A study can be design in this respect to affirm the level of wrongdoing of the children because of peer influence. The same study can be constructively replicated by including a small group of rebellious children to the initial group.
Participant replication involves repeating an experiment with a new set of participants. Technically, this is a constructive replication whose main intent is to ascertain whether similar findings can be obtained by a different set of participants (Stangor, 2011). For instance the testing of people from different cultures, experimented through testing whether the reception of children exposed to video games are the same or different for Chinese children as they are for U.S. children.
Relationship between External Validity and Replication
According to Van Der Staay, Arndt and Nordquist (2009), the assessment of the generalizability of study findings should have its basis on the integral part of the model building process. Moreover, the assessment of the external validity is a pragmatic process that is viably achieved through replication. In light with this, the analytical approaches to external validity propose a varied series of theoretically motivated replications, while the robustness approach reckons on the replication processes across various settings, outcome variables, treatments and units.
Conclusion
Concisely, external validity is a generalization process that is attained through the process of replication. The replication processes aid in the realization of better results of a given research, and they extend the scope of previously performed studies, to allow the generality of results. Similarly, the generalization process is categorized in accordance to the participants and settings or the conditions relevant to the experiment, and assist in the limitation of the findings to some vital variables.
References
Mitchell, L. M. & Jolley, M. J. (2010). Research Design Explained (7th ed.). Belmont, CA:
Wadsworth Cengage Learning.
Stangor, C. (2011). Research Methods for the Behavioral Sciences (4th ed.). Belmont, CA:
Wadsworth Cengage Learning.
Tayie, S. (2005). Research Methods and Writing Research Proposals. Cairo: Center for
Advancement of Postgraduate Studies and Research in Engineering Sciences, Faculty
of Engineering – Cairo University.
Van Der Staay, F. J., Arndt, S. S., & Nordquist, E. R. (2009). Evaluation of animal models of
neurobehavioral disorders. Behavioral and Brain Functions. 5(11), 1-23.
doi:10.1186/1744-9081-5-11.
