Test Developer Profiles

Chapter 2

Dr. Ana Muñoz-Sandoval, Ed.D.

Tests Developed:

Woodcock-Johnson-Revised (Spanish language translation/adaptation)

Woodcock-Muñoz Language Survey

Woodcock Language Proficiency Battery-Revised

Dr. Ana Muñoz-Sandoval received her Ed.D. degree in 1992, specializing in Intercultural/International Education, from the University of Southern California. She is currently the Associate Director of Measurement/Learning/ Consultants, LLC, with responsibility for the translation and development of tests into Spanish and other languages.

Dr. Muñoz-Sandoval was born in Chile in 1947 and grew up in Argentina. Her early education was completed in Mendoza, Argentina, where she received both nursing and teaching credentials before coming to the United States in 1970. Since that time she has also resided in Italy, Nepal, Pakistan, and South Africa. She was a student of German from 1979 to 1980 at the Tribhuvan University in Kathmandu, Nepal, and the Goethe Institute in Poona, India. She received a B.A. degree in Anthropology in 1982 and an M.S. degree in Student Personnel Administration in 1984 from the State University College at Buffalo, New York.

Dr. Muñoz-Sandoval's training as an educator in both the United States and Argentina, her knowledge of second-language learning and human cognition, and her experience in teaching Spanish as a second language at the university level contributed to the selection of her dissertation topic: Influences of Memory and Auditory Processing in Foreign Language Learning in Two Cultures. The results of that study indicated that while memory was the most important quantitative predictor in second-language learning, there appeared to be other, noncognitive factors having an even greater impact. To investigate other reasons for student success or failure, qualitative research procedures were applied to the question. Dr. Muñoz-Sandoval received the Delta Epsilon dissertation award from the University of Southern California in 1992 for her dissertation.

Since the completion of her doctoral studies, Dr. Muñoz-Sandoval has directed the translation and developmental work for several batteries. For almost a decade, her efforts have concentrated on the translation/adaptation of the Woodcock-Johnson Psycho-Educational Battery-Revised (WJ-R) into parallel Spanish forms that were published in 1996 (Batería Woodcock-Muñoz: Pruebas de habilidad cognitiva-Revisada, and Batería Woodcock-Muñoz: Pruebas de aprovechamiento-Revisada). Two related batteries in Spanish were also developed during that period, the Woodcock-Muñoz Language Survey (1993b) and the Woodcock Language Proficiency Battery-Revised (1995).

Two important features of the developmental work underlying these batteries are the "consensus translation" approach (1992) and the "equated U.S. norms" procedure applied to the problem of interpreting results from non-English-language tests used in the United States.

A critical goal for a test translation is the exact communication of intent, involving subtle language nuances, of directions or time content. Toward this end, a strategy of "consensus translation" was adopted. In the case of Spanish-language tests, the initial translation of items and directions involves several Spanish-speaking professionals in generating items and conducting the translation of test instructions. In adhering to a consensus translation approach, the output from each translator is reviewed by at least two other translators. The "corrections" suggested by these reviewers are not automatically accepted, but are passed back and forth among reviewers and the original translator until a consensus is reached. Other translators are consulted, if necessary, until a consensus is achieved.

During early stages of test development, special attention is also directed toward designing items and test instructions that will be deemed appropriate across the Spanish-speaking world. Thus, professionals from several regions of the Spanish-speaking world, not just one, are involved in developing item banks and in preparing test instructions. Each of the cooperating professionals also critically reviews the text of the test and answer keys for possible Spanish-language problems based on regional perspectives.

The second set of issues that must be considered by developers of a Spanish-language test, especially if it is to be used in the United States, relates to norming. A frequently asked question is, "Why not base your norms on data from a sample of Spanish-speaking children and adults residing in the United States?" We believe that test norms, no matter what the source, are valid only to the extent that they provide meaningful information with which to compare an individual's test performance. At this time, it would be difficult, if not impossible, to assemble a norming sample of Spanish-speaking subjects in the United States that would be accepted by the majority of professionals as representative enough for their purposes. In contrast, the general population of the United States is the group that is the standard reference for norms on tests used in the United States. Through calibration and equating research, it is possible to relate performance on tests in the Spanish language to the distribution of like abilities in the general U.S. population (Woodcock & Muñoz-Sandoval, 1993a, 1994). For example, the sample chosen to calibrate and equate the Spanish-language items in the batteries named above were drawn from both inside and outside the United States. These data were gathered in Argentina, Costa Rica, Mexico, Peru, Puerto Rico, Spain, and the United States. Calibration-equating data for each test were typically obtained from approximately 2,000 native Spanish-speaking subjects. The Spanish-language ability of these subjects ranged from two years of age to university graduate student.

Dr. Muñoz-Sandoval is presently completing development of the Bilingual Verbal Ability Tests (BVAT) with colleagues Jim Cummins, Criselda Alvarado, and Mary Ruef. The core of the BVAT is comprised of three oral language tests drawn from the WJ-R. In the past three years, these authors have worked with sixty bilingual professionals in developing and translating these three tests into fifteen other languages.

The BVAT procedure has been developed to help estimate the total linguistic repertoire, or verbal ability, of a bilingual person. The bilingual person brings to the classroom or workplace a base of linguistic knowledge that is more than can be measured in English alone or in their first language alone. The BVAT takes into account total language development by assessing the linguistic and conceptual knowledge that individuals possess regardless of language source. The BVAT will be useful in many aspects of educational decision making, particularly decisions regarding special education intervention and placement of bilingual students.