Psychometric Toolbox: un paquete Excel disenado para cursos en medida y psicometria.
The Psychometric Toolbox (PT) is designed as a user-friendly, noncommercial package intended to be used for instructional purposes in introductory courses of educational and psychological measurement, psychometrics and statistics. While excellent programs of this type already exist--for example, TAP (Brooks & Johanson, 2003), which focuses on classical test theory, and IRT-Lab (Penfield, 2003), which specializes in Item Response Theory - the main contributions of PT are its wide scope and versatility, which make it possible to carry out Guttman analysis, classical item and test analysis, factor analysis, and IRT analysis. Furthermore, a variety of procedures for scoring and norming are available. Finally, PT includes a powerful data preprocessor which allows many statistical treatments, sampling procedures, and data transformations to be performed.
The PT package is organized in six separate modules: (1) Data preprocessing, (2) Guttman scaling, (3) Classical item analysis, (4) Item factor analysis, (5) Scoring and Norming, and (6) IRT analysis. Because they have been designed for instructional use, these modules are intended (a) to be very easy to use, (b) to provide clear results, and (c) to make extensive use of graphical displays. As well as the basic functions expected in programs of this type, some not-so-usual features are now provided as example. The preprocessor module allows the user to replace missing values, perform split-half test partitions, randomly split the sample for cross-validation purposes, perform Jacknife re-sampling, and create automatically parallel forms. The item factor analysis computes the omega coefficient with its confidence intervals and displays one and two canonical solutions in order to detect violations to unidimensionality (e.g. Ferrando & Lorenzo-Seva, 2013). The Scoring and Norming program provides confidence intervals for the transformed scores and creates the test normative table on the basis of the user's specifications.
PT is a collection of Excel macros and does not need a specific installation (as long as Microsoft Excel is already installed in the computer). We developed them in Microsoft Office Excel 2010, and the programming language used is Visual Basic of Applications. The functions implemented are made available to the user in a specific ribbon, and the contents of the document must be enabled by the user if the implemented macros are to be executed. We check our modules before publishing them on the web page, so that we are sure that they are free of any malicious code. With the datasets that we have analyzed so far, we have found no limitations of disk or memory space. With regards to the accuracy and correctness of the PT results, we have checked them to those which can also be obtained from commercial programs (SPSS-descriptives, SPSS-reliability, SPSS-factor, and BILOG-MG3) and found all of them to be accurate at least to the second decimal place.
Limitations and future extensions
At present, the PT package is only intended for binary and gradedresponse items analyzed on a scale-by-scale basis and fitted with dominance models. In the near future, we plan to extend the package capabilities so that (a) other items formats (mainly multiple-choice) could be assessed, and (b) other models, including multidimensional models could be also fitted. Improvements in the functioning of the existing modules as well as the output they provide are also on the way.
Availability and accompanying documentation
The program and accompanying material are available free of charge from the website:
The documentation consists of: (a) a summary of the functioning of each module with the procedures available, (b) a complete user's guide, which is intended for instructors and includes a theoretical background and detailed step-by-step analyses of empirical data, and (c) the data-files used as examples in the guide.
(Manuscript received: 7 December 2016; accepted: 13 February 2017)
Brooks, G. P., & Johanson, G. A. (2003). TAP: Test Analysis Program. Applied Psychological Measurement, 27, 303-304.
Ferrando, P. J., & Lorenzo-Seva, U. (2013). Unrestricted item factor analysis and some relations with item response theory. Technical Report. Department of Psychology, Universitat Rovira i Virgili, Tarragona. http://psico.fcep.urv.es/utilitats/factor.
Penfield, R. D. (2003). IRT-Lab: Software for research and pedagogy in item response theory. Applied Psychological Measurement, 27, 301-302.
Pere J. Ferrando *, Antoni Masip-Cabrera, David Navarro-Gonzalez, and Urbano Lorenzo-Seva
Research Centre for Behavioral Assessment
'Universitat Rovira i Virgili', Spain
* Acknowledgments: The development of PT was supported by a grant from the Catalan Ministry of Universities, Research and the Information Society (2014 SGR 73), a grant from the Spanish Ministry of Economy and Competitiveness (PSI2014-52884-P), and a grant from Rovira I Virgili University-ICE (A01/16). Correspondence should be sent to: Pere J. Ferrando. Universidad 'Rovira i Virgili'. Facultad de Psicologia. Carretera Valls s/n. 43007 Tarragona (SPAIN). Tel: +34 977 558079. E-mail: firstname.lastname@example.org