The difference between external validity and internal validity
External validity is a measure that determines whether or not the results of the sample group, achieved during an experiment, can be applied to a wider population outside of the study (Shuttleworth, n.d.). Contrary to internal validity, which ensures the logic and reliability of the experiment’s design, external validity aims to establish whether or not the study results can be used to make generalized conclusions.
McLeod (2013) states that external validity “refers to the extent to which the results of a study can be generalized to other settings (ecological validity), other people (population validity) and over time (historical validity)” (para. 2). On the other hand, internal validity measures if the effects observed were due to the independent variable studied (McLeod, 2013).
The ways to ensure external validity in research
Kraus (2013) argues that many research studies have limited external validity by design – for instance, due to being set in laboratories and not in real-life circumstances. However, researchers can ensure that their findings can be extended to different contexts by choosing appropriate sampling methods or imitating real-life conditions.
The question of external validity is one of the most pressing issues in many of the studies conducted today, which is why some researchers devise tools for measuring their research’s external validity as part of the study. Nevertheless, some common methods of ensuring external validity still exist. For instance, researchers can make sure that the conclusions achieved due to the study can be extended to a wider population by ensuring that the chosen sample group mirrors the general population correctly (Lund Research Ltd, 2012). On the other hand, external ecological validity can be reached by imitating real-life circumstances or conducting studies in various contexts to determine the variability of results due to ecological circumstances (Lund Research Ltd, 2012).
External validity issues in the example studies
Del Campo-Avila, J., Conejo, R., Triguero, F., & Morales-Bueno, R. (2015). Mining web-based educational systems to predict student learning achievements. International Journal of Interactive Multimedia and Artificial Intelligence, 3(2), 49-54.
In this study, the researchers aimed to determine if the analysis of the student’s continual assessment results can predict their final course grade. The study focused on university-level students of the subject “Principles in Informatics,” assessed in two consecutive courses. The researchers did not provide any assessment of the study’s external validity or justification of the sampling methods used, which is why it is unclear whether their conclusions can be extended to a wider student population. Moreover, the researchers did not address the study’s ecological or contextual validity by evaluating external circumstances that could have affected the results. Overall, although the authors make an indirect reference to longitudinal validity by discussing the possible areas for the future development of web-based education programs, the study’s external validity remains relatively low, as most of the components remain unaddressed and unevaluated.
Onorato, M. (2013). Transformational leadership style in the education sector: An empirical study of corporate managers and educational leaders. Academy of Educational Leadership Journal, 17(1), 33-47.
This article describes a study of educational leadership practices in educational institutions of New York, NY. The researchers’ goal was “to explore educational school leaders, and the need for transformational leadership style within the educational sector” (Onorato, 2013, p. 33). The study addressed educational leaders in all education levels by analyzing leadership styles and practices of 45 principles from elementary, middle, and high schools. The authors somewhat address the notion of external validity by following the sampling guidelines for random selection. Nevertheless, the article does not explain the possible influence of geographical area on schools’ practices, thus not providing an evaluation of the study’s ecological validity. Therefore, only one of the components of external validity is referred to in the study, whereas other concepts are not measured in any way; hence, the overall treatment of external validity in the study is inadequate to justify the results’ generalization.
References
Del Campo-Avila, J., Conejo, R., Triguero, F., & Morales-Bueno, R. (2015). Mining web-based educational systems to predict student learning achievements. International Journal of Interactive Multimedia and Artificial Intelligence, 3(2), 49-54.
Kraus, M. (2013). External validity and believability in experiments. Berkeley Science Review. Web.
Lund Research Ltd. (2012). External validity.Laerd Dissertation. Web.
McLeod, S. (2013). What is validity? Web.
Onorato, M. (2013). Transformational leadership style in the education sector: An empirical study of corporate managers and educational leaders. Academy of Educational Leadership Journal, 17(1), 33-47.
Shuttleworth, M. (n.d.). External validity. Web.