Introduction
Over many years, psychologists have invested a lot of time and effort to establish and maximize the validity of test scores. The main reason for the maximization of validity was to ensure that tools used for psychological assessments measure what they are designed or intended to measure. The existence of test tools that provide scores that have high validity ensures that psychologists have reliable and valid instruments that enhance the process of testing (Borsboom, Mellenbergh & Heerden, 2004).
Therefore, based on accuracy, psychologists can draw accurate inferences from the scientific test processes and apply the findings to make accurate decisions. The following paper explores validity in relation to the process-focused (PF) model and the traditional model of validity, the importance of PF, and the process of implementation of the PF.
PF Model and Traditional Model of Validity
The PF and traditional models are critical in establishing test score validity in the psychological processes. The major point of difference between the traditional and PF model lies in the various areas of psychological assessment. The first key difference between the traditional model and the PF model relates to the process of conceptualizing validity. In the traditional model, the validity of a score is determined based on the outcome.
According to Bornstein (2011), the scope of the test score is based on a correlation with a related theoretical variable. On the other hand, the extent of respondents’ engagement is applied invalidity conceptualization in PF model. Bornstein (2010) noted that the primary evidence of validity in PF model is based on the extent to which a test score is modified through the manipulation of variables in a theoretically related process. In addition to the differences in the conceptualization process, the traditional and PF models differ in the research process. In the traditional method, criterion correlation is applied in the assessment of the predictor (Bornstein, 2011).
The PF model provides a shift from the correlation to the application of experimental approaches that manipulate variables. The result is the moderation of the test score and criterion relationships. The PF model applies the empirical process that enables psychologists to arrive at definite inferences in relation to the aspect being studied. Unlike the traditional methods, the PF model moderates the variables that are being tested and manipulate outliers in order to reserve them.
Bornstein (2011) noted that the key difference between the two models is that the PF places emphasis on the process. Furthermore, it shifts from the correlation used in the traditional method to the use of experimentation and manipulation of variables (Bornstein, 2011). Therefore, the researcher can identify the potential bias in the test and enhance the assessment process.
The other points of differences between the two models include the test development goals and test development challenges. In the traditional approach, the test scores are maximized through outcome correlation, while in the PF, the test development goals are based on demonstration of impacts of the processes that are theoretically related. Despite the differences, a common feature of the two models is that on their own, the models do not have the ability to produce a comprehensive picture of the test score. However, the PF model has the ability to produce information that is unique to the traditional model.
Importance of PF Model of Validity
The PF model provides unique information than the traditional model. According to Bornstein (2011), the validity is conceptualized in relation to the extent to which the respondents engage in the predictable test process. Unlike the traditional models, the PF emphasizes on the process; hence, inferences drawn are unique. Furthermore, the PF model is an inclusive process. In the process of psychological test scores, variables that are considered extraneous are reserved.
The traditional model considers the extraneous variables as problematic and depicts them as sources of interference with psychological tests. However, PF model reserves the variables and confounds them by providing a framework for manipulation and analysis of the tests using all the variables. The procedure for dealing with extraneous variables enables the PF model to achieve accurate test score that are crucial in drawing conclusions and making right decisions.
According to Bornstein (2011), continued reliance on correlation methods in the process of quantification of validity is a source of problems in psychological tests. The common fundamental problem is that in the social sciences such as psychology, there is a high degree of things or variables correlating. As such, the traditional model ends up correlating everything. The PF model bridges the disconnect that is present in the unidirectional psychometricians presentation of the test score validity. To avoid the correlation of everything, PF model applies the experimental manipulations in the process of the of respondent’s test process.
Therefore, PF model provides bidirectional validation strategy via application of experiments. Furthermore, Bornstein (2011) noted that the PF has a unique feature that links the process of testing to other fields of psychology. Thus, the PF model enhances in-depth understanding of test score validity.
Psychological Variable and Test
The psychological variable that will be assessed is the prediction of behavior for adolescents. The variable involves the test of the personality of the adolescents. The most relevant tool to measure the personality variable for adolescents is the 16PF Adolescent Personality Questionnaire. 16PF Adolescent Personality Questionnaire is an instrument commonly applied to assess normal personality range of the adolescents (Cattell & Mead, 2008). According to Cattell and Mead (2008), the instrument is designed to measure the structural elements of adolescents’ personality. The 16PF Adolescent Personality Questionnaire is a multi -level instrument, the different levels of the tool are integrated to provide an in-depth evaluation, and prediction of accurate behavior of the individual being tested (Cattell & Mead, 2008).
Using PF Model’s four Steps to Assess the Construct and Psychological Test
The first step as outlined by Bornstein (2011) entails deconstruction of the instrument for the assessment. The step entails the identification of the processes that influence the construct and the test score. The deconstruction involves the identification of the various variables that relate to adolescent personality by use of the 16PF Adolescent Personality Questionnaire. Therefore, the 16PF Adolescent Personality Questionnaire will be deconstructed into the sixteen scales that make up the elements of personality such as the reasoning, dominance, social boldness, self-reliance and tension. The context variables that modify the scales in relation to prediction of adolescents’ behavior and the overall construct of 16PF Adolescent Personality Questionnaire will be identified.
Operationalization and the evaluation of the processes will be the second step. The process will entail testing of the outcomes empirically. The values obtained for the different test scales that relate to the personality of the adolescents as outlined in the 16PF Adolescent Personality Questionnaire will then be manipulated. The extraneous variables will be reserved for the manipulations process. The hypothesis that relate to the adolescent personality will be described in details.
Through experimentation, the validity will be determined in the process of the 16PF psychological testing. The third step of assessing the construct and the psychological test will entail interpretation of the outcome achieved in the second step. The interpretation process will also entail detailed evaluation of the outcomes. The key to the evaluation will involve identification of the factors that denote the validity and the possible limitations. For instance, there are limitations that may affect 16PF Adolescent Personality Questionnaire in the prediction of the adolescents’ behavior such as the hit rates, the misses, the false positives and false negatives.
The final step will entail generalization of the evaluation. The generalization will entail testing the construct under different populations and contexts to determine whether similar patterns will be obtained. The validity of a test involves the degree in which a test or test instrument achieves what it is designed to measure (Borsboom et al., 2004). Therefore, the evaluation will help in establishing whether the tool measured what it was designed to measure. Variations in the tests in the different pupations and contexts will point to unreliability and lack of validity. An instrument that does not achieve accurate results signifies that it cannot be relied on for validity (Cattell & Mead, 2008). Therefore, validity and reliability of the test will be deemed to have been achieved if the results will be replicated in the different populations and contexts.
References
Bornstein, R. F. (2011). Toward a process-focused model of test score validity: Improving psychological assessment in science and practice. Psychological Assessment, 23 (2), 532-544.
Borsboom, D., Mellenbergh, G. J., & Heerden, J. (2004). The concept of validity. Psychological Review, 111(1), 1061–1071.
Cattell, H., & Mead, A. (2008). The Sixteen-Personality Factor Questionnaire (16PF). London: Sage Publishers.